Laz,
the G2 is on the HVPS, they are all separate circuits for each CRT, but they get their main supply from the potted section of the HVPS.

There are several recommendations that I will make to correct some problems on these boards, and there's a point where I'll recommend that you replace the board (I've already done that on this thread). For the sake of safety and not having part in you damaging your high voltage supply, neck board or your CRT, I'll rather stick to my initial recommendation (."At this point I'd recommend that you replace that board.").

That board has many "Critical Safety Components" on it, and they play a very important part for tube protection and X-radiation safety. I've mentioned this earlier on this thread, and I've also mentioned that we should not get into these circuits on this thread.

Originally posted by sly

The 2nd thing is the air direction the fans blows from the bottom of the PJ to the top. This works fine if the PJ is right side up but if it is on the ceiling would it not be better to revers the fans so the air goes from top to bottom aka (the air goes up). So it it not fighting the convection??

Or if I am to keep the air flow the same would a remote blower blowing into the PJ be the best. This has a lot of advantages one the noise but the best is I could put a air filter on the blower so my PJ never gets dusty.

Any thoughts on the matter??

Sly

I asked the Electrohome engineers the same question and they said to leave it alone, the fan pressure far exceeds convection. Their advice makes sense because that three-fan plenum puts pressure to the HDM and focus boards and to the three tube covers and neck boards also. In tests we did here at E-Tech we found that the two LVPS fans contribute most of the noise and it is due to the fan motor bearings resonating the sheet metal; our Fan Noise Mod handles that very effectively with no change in fan voltage or air flow; a similar mod to the plenum give no detectible improvement, much to our surprise.

Problem solved

It was due to the G2 on the red and green drive were set to 80% and the blue was at 40%.

If i do a master reset should the RBG G2's set to the default or are they independent of the reset.

Just a quick one when all is set properly one should see no raster or image on the screen with no input connected right?

sorry for all these questions but this is my first CRT projector experience.

Mike sorry to be a pain.

Laz

Originally posted by Aksor
Problem solved


If i do a master reset should the RBG G2's set to the default or are they independent of the reset.

Just a quick one when all is set properly one should see no raster or image on the screen with no input connected right?

sorry for all these questions but this is my first CRT projector experience.

Mike sorry to be a pain.

Laz

Yep, always do a full reset. No problem with the questions. Just trying to keep you from doing harm to yourself or your projector.

Guys!

Having to set G2 to 80 means game over for such tubes; they are nearly gone; have a look here and be sure to check the filament voltage across P14: http://www.etechvideo.com/techtip1.htm

Hi !

Mr. Mike Parker.


I have read this tread with greatest interest.

I have one question in favor of you.

To enhance detailness in 9500LC, I think the video switcher ic(SD5401)
should be tweaked .
high frequncy signal is losing in this part, I guess.

When i bypass this circuit, the detailness is much improved.

But i can get OSDif bypassed.

I need your advice on this.

In Seoul, Korea, we have 5 HDTV (ATSC channel).
so most of my Source is HD 1080I.

Jaehong Lee,
you're right, and that is the weakest link in the video chain. As you mentioned, that switch selects between the internal video and the main video source. The other sections of that switch is for the DC restore circuits. The modification for this is not easy...

This is one of the five tweaks to that board that I have in my personal (private) bag of tricks.

well I did it after looking at the air flow
it seems to have 150CFM of fans air is sucked in the sides top and over card cage.
into a air box where is is blown over the 3 CRT's and the 2 deflection cards.

So I got a 300CFM blower at Home Depot that you can use a dimer on if you want to. Very quiet. Ran a 4in tube.
into the air box and added some baffles to even out the air flow. added a air filter to the blower ( the main reason for this)

The new air flow it in is the same for the CRT's and the 2 deflection cards but now it also feeds the card cage
so the air is now reversed for the card cage.
so far I seem to have a lot more air flow with a lot less sound..

I will post some pics later

Sly, yes, please post pics and a description. Thanks!

Cary

Mike:
I have the blanking out on bright scenes problem. what is the cure.
I also have the problem of vertical squeeze mod, can someone please help to shown pictures of the resistor to be changed.

My Marquee is of build Aug 1994.
Regard.

Originally posted by mp20748
The gathering next week (5-17-2003) will be in Chevy Chase Maryland, not Bethesda. It will be right off of 410 (east west highway). Shoot me an email if interested (mp@wisdom-technologies.com). We plan to start around 3PM.

Hey Cary, Don't you wish we were closer to Maryland? Be a hell of a road trip, huh? :D

Hmmmm...Lemme look at the map....

Cary

If your projector blooms and blanks out on bright scenes, you'll need to have your neck boards upgraded. That involves about 10 components on each board, and you should have all three done at once.

Guys,
I'll do what I can to make it as if you were there. This thread will greatly pick up after the gathering. Plus, wait till I get back from my visit with TIM later this month, we're gonna make you very proud to own a Marquee:D

I thought I would share a convergence trick I use.

if you want to get that last bit out of the projector
converge on video and use your own patterns.


I been doing this so long I forgot their was a other way ;)

you will be amazed at how much the image can vary from video to the internal patterns. And it is more so at the higher resolutions.

I use a 1 pixel crosshatch that I have made. It also has some markers on the edge if I am going to match it to a another projector.

Don't be shy sly. Post your pattern!

I wouldn't expect that the external source would be different. It uses the external sync signals so the deflection works exactly the same. It must be a difference between the delays in the external RGB vs the delays in the internal RGB. I wouldn't have thought it would be that different but I'll sure try an external pattern.

Cary

I don't have them here and some were at a strange rez.
I have remade them many times on the spot.
just draw some lines


And yes there is the delays in the Red vs Green vs Blue

But also the lines are finer so it is easer to do.
The bult in lines are kinda thick.

Ok, thanks sly. I'll just do it in Paint or something. I didn't think of the resolution really making a difference but if we want to keep it at 1 pixel thickness in both X and Y then I should probably match my desktop rez (1440x720).

Cary

There is a guy who sells some great hirez test patterns, who's stuff I reviewed on Techweenies.com. I forget offhand the name, but there is a link on the home theater page.

If I remember right, he has 640x480 patterns available for free, but you have to pay a small sum for the higher resolution ones.

Mongo

http://www.fi.muni.cz/~kas/screentest/

here is a free one with source..

Sly, do you happen to have an i386 executable in another form than an RPM package?

Here's one that I find useful periodically.

--Bill

And here's another.

--Bill

How bout used or rebuilt tubes for a marquee 8500, non-LC. I need a green one badly. Where's a good place to look? I did a google search and found nothing.
Marc

VDC. Video Display Corporation. Search on here and you'll get names/number/etc. $495 plus $100 core charge (refunded when you send yours back) plus freight. These are for a rebuilt which are as good as new!

The gathering went as planned, but we cancelled showing the transcanner. That will still happen, and I'll make that happen in another week or so (next gathering), because I'm really excited about this project. Plus some who wanted to attend could not make it, so we're doing it again.


The projector setup and viewing was hindered by video card failure, but we later (had to go buy another card, etc - the downsides of PC's:rolleyes: ). Once we got that back up, it went rather well. We did an overview of the technology, and I had a chance to show what to look for when a projector is properly setup (mini seminar), and hopefully we'll do this again with screen shots in a couple of weeks.

The 8500LC though it was not fully setup, was a sight to see once the card problem was fixed. I wanted to show how bright, clean and sharp this projector would get before clipping. I also wanted to show that there was NO noise in the image. The image was displayed with the HTPC as source, it had my MP-1 mod, but it was in unity gain (700mv output). I had bought my scope to show what the IRE window should look like with proper scoped calibrations, and I also wanted to show that the projector was performing on a true 700mv signal (not boosted), I'll have to show this at the next gathering.

Since I've announced that I'll no longer be doing tours or projector "calibrations", I've been swamped with emails. I'll still do a few now and then, but would rather focus my time on other things. But some of you are persistant, so I'll be making a trip to New York in late June or early July. Contact me if you'd want me to stop by, if you live anywhere between Virginia and New York. I may even make a couple of trips and extend into August, but I'll mostly be concerned with showing you how to maintain your own projector, and making sure that your projector is performing properly before you throw money into ("calibration"). I've been asked many times if the projector should have been evaluated before calibration, and my answer is absolutely, because a perfectly working projector could have a few glitches when it arrives to it's final location. And this happens with the brand new projectors, so it's very important to do a good eval. first.

The Marquee is easy to setup, and I've been recommending that the owner install and do the preliminary setup themselves to become familiar with the technology (if HT is a hobby). During this process, the owner would also discover if there's any problems with the projector, and could seek the help from the forum or others here on the forum to correct any problems before a "calibration" is scheduled. This is something that I have promised on this thread, so we'll also be covering setup and calibrations, but only from a perspective of making sure the owner of the projector makes the right decision before seeking a "calibration" or "setup". This may sound unusual, but I'm also responding to many emails that I've received on this matter. I may not be an "expert" on this, but with over 30 years experience in electronics and over 20 years experience with CRT projectors, to include all aspects (stacks, simulators, video walls, immersive, etc.) of display setups. So yes, I do think I'm qualified to say something on this matter.

Mike, Whether you think you're an expert or not, I'm sure that I'm not the only one who's still all ears! I think it's safe to say we think your qualified on the matter as well! ;)

Mike,
As someone who attended your event on Saturday, I can attest to it's usefullness. It reassured me that it's worth it to persist through this process because, as "imperfect" as you say the set up was, I thought the picture was mindblowing. I didn't have the slightest longing for a simple plasma screen after I saw what that machine could do.
Marc

I also would like to thank Mike for a very informative and enjoyable afternoon. My friend Jack and myself both got a lot out of the presentation and it made us both want to rush home and give our PJ's a good viewing.

For those who have not yet me thim, Mike Parker is a really good guy who is full of interesting information and is very generous in sharing it with others who share the CRT hobby.

I'd also like to thank Mike Parker for sharing his wealth of knowlege! I greatly enjoyed the Saturday gathering.

JSDDS

Does anyone know a source, or good used parts seller/list, where I might find a green tube for the 8500. A good used tube, or I take it a rebuilt from VDC are my options, I take it?
Marc

Here we go. Now for those of you who like bending the rules, and putting 8 cylinder engines on motorcycles, 200 watt amps on 6" speakers, etc., etc.

Here is how you "Hot Rod" the video chain. As you may already know, with the exception of three coupling caps, the entire video chain in the Marquee is directly coupled (DC) from the BNC connectors to the CRT's.

The front end of the VIM is where you'd find these coupling caps. There is one on each input (RGB). Direct coupling is better, but it has it's drawbacks, and that is that if the DC offset is off on the source (input) device, it will effect the offset in the video chain of the projector. When the DC level (offset) is off, you'll then effect the blanking level (this is the reference for black). And from there you can forget about the great black level feature of this CRT.

So what they do is add these coupling caps and make the front end AC coupled, meaning that the caps will block DC, and only pass the AC signal. But it's not that simple, because once you AC couple you'll have to get and keep the DC offset at zero, so they add a circuit called a "DC Restore Circuit" and this circuit would keep the video front end DC level to zero, and if the following circuits are designed properly, the entire video chain would have a balanced (0) DC level. Therefore with proper DC level, we will have proper blanking level and black levels. So our goal is to not effect this feature in our quest to improve the projector.

To enhance the performance of the video chain, change the value of these three caps (C33, C49 and C65). The value is 470N or 0.47 ufd. If you increase the capacitance the the picture would get hotter, but don't use electrolytics (especially polarized electrolytics). If you use electrolytics, you'll force DC reference into the video chain whenever there is bright scenes. The DC restore circuit will correct this, but it may not do it fast enough because of longer charge on the higher value cap. For instance, if you double the value of the coupling caps (0.94 ufd), you'll may also be putting the circuit outside of the speed of the DC restore circuit, because there is also a cap in that circuit that holds a charge. So what you may experience by doing one thing, you would more likely be effecting something else. So the "hot rod" effect of changing out C33, C49 and C65 would not be a miracle improvement, though this works well in high end audio (there's no need for DC restoration). It could be a problem in video.

I'm of the opinion that you cannot drastically improve on the performance of a Marquee, and that's mostly based on the fact that the video chain has a very wide bandwidth, and the components that were used (upgrades) were very high performance devices. Still there's room for improvements, but I would think that the end results would be more subtle than drastic.

The video chain in a marquee is very well designed. And this means of Hot rodding could cause effect to the video chains DC offset level. You'll experience a brighter and falsely more dynamic picture, but pay close attention to the black bars from the test disk.

Now, here's a nugget. You can improve on the video chain, by replacing that cap with the same value (0.47 ufd), go from a 50vdc to a very high quailty 100 vdc poly. The change would be subtle, but noticeable if you're using a high quality input source.

Will do! Thanks Mike! Any more while you're in a givin' mood? :)

Cary

Hey Mike, it looks like the video chain is also AC coupled on the neck board. Before hitting the tube's cathode pin, video goes through both a .1uf 50V and a 10uf 16V. Has anyone ever looked at upgrading these?

Cary

You guys are making me jealous! Before you go into retirement Mike, I wonder how many CRT enthusiasts here on the west coast would chip in to bring you here for one of your demos/seminars.

I would be more than willing to offer my home theater as the host site.

I have an 8000 now with an 8500LC ultra on the way. I'm sure we would make it worth your while and enjoyable.

Brad

Originally posted by jcmccorm
Hey Mike, it looks like the video chain is also AC coupled on the neck board. Before hitting the tube's cathode pin, video goes through both a .1uf 50V and a 10uf 16V. Has anyone ever looked at upgrading these?

Cary

Actually, if you look at the diagram, there are two output sections on that card (dual complimentary cascode output amplifiers) Q13 and Q14 are connected 'directly' to G1 on the CRT, while Q1 and Q22 are connected to the cathode of the CRT. They called this dual output a "mirror cascode amplifier" that drives both the cathode and G1 elements of the CRT's.


They put the I-sense and over current protect in the cathode drive (Q1/Q22) of the CRT, there we find the two caps that are a part of the "protect circuit", and that's why they are inclosed in the dashed lines on the diagram, meaning that it's a critical safety area ("all components within boundary are not service replaceable in the field")

That circuit -- I leave alone.

Mike. I like nuggets :) On the coupling caps, is there a recommendation as to type to replace with? tants, poly, etc...or is anything (of quality) as long as it's not an electrolytic sufficient?

Mike,
That I-SENS and OVER-I circuit on the neck board is pretty wierd to say the least. It looks like they are actually depending upon the resistance (ESR) of those capacitors to detect an over-current situation for the beam current. Replacing these would indeed be tricky. (and I guess, not advisable)

Cary

EDIT: But they *are* AC coupling caps in the video chain and therefore will be forever nagging me....

Mike, I am able to get hold of clc449 and max4224, but the clc449 is 4 times the price of max4224, is it worth while to use clc449

Originally posted by jcmccorm


EDIT: But they *are* AC coupling caps in the video chain and therefore will be forever nagging me....



That makes two of us. And I've thought about playing around with that circuit, but the thought of damaging a tube or crushing the X-Radiation protection circuit is my main concern.

If only a better quality cap could be used (same value), because the biggest problem I'm having with this is the size of the caps... One day we should try this, but first I'd have to either get my radiation meter fixed or get another one.

---

joekjtam,
I don't know, I'm still using both CLC449's and MAX4224's. But if there's no noticeable difference between the two, I'd say go with the Max, because it's cheaper and the CLC449 is no longer being manufactured (maybe that's why it cost so much more).

mike
thanks for your reply,
I have the problem of blank on bright scene and i am repairing my marquee 8000. will you have a post regarding this shortly.

Originally posted by joekjtam
mike
thanks for your reply,
I have the problem of blank on bright scene and i am repairing my marquee 8000. will you have a post regarding this shortly.

This problem was covered on this thread (much earlier). You have the very early version neck boards that were prone to this problem. The fix is to have the neck boards upgraded, or just replace them. You'll also need to replace U2, it should be a CLC409, you'll need to get that guy out of there and put in a CLC449 or MAX4224.

Originally posted by Jay Wilson
Mike. I like nuggets :) On the coupling caps, is there a recommendation as to type to replace with? tants, poly, etc...or is anything (of quality) as long as it's not an electrolytic sufficient?

I'm presently using a Digikey part number: P4671-ND (0.47 mfd 50vdc metal film caps).

Joekjtam

I am looking for some CLC449's. Where can you get them and how many/price do you have to purchase?

Walt

Mike,
I have read through the entire thread, you have advise that there is about 10 components that need to be changed to have the blank on bright scene to be cured. I do not have a comprenhensive idea of the 10 changes. Do you mean that by changing the clc409 to max 4224 will cure the problem.


Aspec
I travel to China Shumshen freqently and found that one or two of the ic company is still stocking the clc449. They are quite costy and is around US20 each. No limit on the nos. that need to be purchased but the price will drop when the quantity is large.

Originally posted by mp20748
I'm presently using a Digikey part number: P4671-ND (0.47 mfd 50vdc metal film caps).

Thanks Mike!

Originally posted by joekjtam
..........

I do not have a comprenhensive idea of the 10 changes. Do you mean that by changing the clc409 to max 4224 will cure the problem.



I did not provide a comprehensive procedure for the neck board upgrade. The reason is because the repair involves the "protect circuit", and much care and consideration should be giving to servicing this circuit, therefore I don't post mods or fixes for any circuit that's deals with "safety" or "protection" - I would suggest that you replace the board.

The MAX4224 will work in place of the CLC409 on the neck boards.

------


Concerning the chip upgrades...

This week I had contacted the manufacturer (National Semiconductor) of the LMH6702. That chip is supposed to be a direct replacement for the CLC449 (Comlinear) that were used in the Marquees, and is now discontinued.

IT WILL NOT FUNCTION PROPERLY IN THE VIM -- Do not use!

After trying this chip I had experienced distortion when text was being displayed, also it gave an enhanced brightness, which created a strange bloom or clipping.

I called and emailed their tech support, and was told to also send an email. I later received a return call from one of the application engineers. We discussed the problem that I was experiencing with the chip. I had also told him that I duplicated the same problem on a test board. The engineer told me that he would contact the designers of the chip and would also test it himself... Later in the week I got a call from the engineer informing me that he was also able to duplicate a simular problem with the LMH6702. His findings were that the chip was having a "parasitic Oscillation" In other words it was doing its own thing. We discussed a fix or cure for it, and he suggested that I modify the circuit with resistors and such. I told him that the chip is broke, it has a problem that needs attention, and why should a total redesign of the circuit be needed for an "exact replacement" chip. His response was that it works well in most applications, and that it had passed all of their test, however it does have this minor distortion issue.

I simply repeated that the chip is bad, and it should be pulled from distribution. So just in case you see this chip (LMH6702) as a direct replacement - it ain't!

Now concerning the CLC449, he had also informed me that it had a clipping problem, and that if it's used in a circuit, that circuit should as well be modified to correct a bandwidth problem they later found with the chip :rolleyes:

The Max4224 works well on all neck boards, but MAY not work properly in some versions of the VIMs (confirmed). I'll have to wait and check this further, but I have several other chips on order that I'll be testing. Plus, the engineer gave me a nugget on how to make these chips work when they won't directly cross. I just would not try it on the LMH6702 because that chip is broke, they can talk all the technical babble they want, it's still broke, and if you can't remove the top off the chip and repair it, it should not be used - period.

I have a very clean and excellent performing chip that I'm testing. After I check out the ones on order I'll post back on this...

Hi Brian

Here it is!

DH

I was fortunate enough to have Mike Parker himself come to my home this past weekend and spend some time tweaking my projector. I've met Mike before and he is a heck of a nice guy as well as someone who knows Marquee's inside and out.

He was able to get my Marquee 8110+ sharper and clearer that I have ever seen it before. Matter of fact, it is so good now that I need one of his mods to get rid of some of the grunge left in the source-end of my system.

Thanks again Mike for helping me out!

Mongo

Mongo, do you realize how much envy your post is generating? ;) Congrats on the newfound picture quality - I wish we could all do so well!

I replaced all 6 H1110 op amps on my VIM with the Maxim Max4224and the 3 H1110 on my neckboards.

After I turned on the projector the raster was full of video noise interference. So I removed the Max4224 from the neckboards and put the H1110 back and the noise problem is gone.

Question for Mike. I seem to have a vertical line down the middle of my raster that is a touch brighter than the raster background with no input signal present on all 3 tubes.

If i bring up an internal test pattern the line is gone. Is this normal or should I be concerned about it.

Laz

Mike

I experience similar problem of large background interference when i change the H1100 at the neck board to MAX4224, but the picture is sharper. What is the final replacement of the ic for the vim and the neckboard please.

Guys,
I've used the Max4224 in a several neck boards with no incidence. I've notice a problem when used on a VIM, but not all of the VIM's that I've used them in. I have received a different chip in to try out in the next day or so. I'll post back the results after testing.

The MAX4224 is a better chip, and the problem (noise) that you've experienced can be correctable by changing the value of the feedback resistors to taylor the bandwidth of the chip, and changing the bypass caps. It can be done, but I'll rather rule out other chips first before we start modifying the circuit. But as was stated, changing that chip will make a difference. But not in all cases.

------

The Transcoder is ready for testing :) :) :) :) :) :) :) :) :) :) !!!!!!

BUMP.

Transcoder:
I was hoping to test it today on a Motorola HDTV box, but that has been put on hold, maybe this weekend. I'm also waiting to send a test unit off to Michigan for evaluation, but backordered parts have delayed that, but that will happen next week.

So far, only one person other than myself has seen the unit in action here at my house on a D-JVC deck. I feel I've achieved my goal with this unit, and was able to overcome the common transcoder problems (black level, dark image, weak colors, etc.). And as an added bonus, it also has the Clarity performance of the MP-1 mod.

We'll be getting back to that "special" 8500LC that was at the gathering once the owner of the projector gets back from vacation. I'll continue where we left off and will have the transcoder there.

-------

Marquee:
I'm still testing the new chip, but have been very busy with some other issues, as we (my wife and I) are getting ready for court to file against a roofing company that improperly installed a roof 9 years ago that has caused severe structural (hidden) damage to our home. The good news on this is that the state has made them accept responsibility, so that eliminates who's at fault. It's amazing of what could happen from just a water leak. Much thanks to the help of Mark Haflich and Steve Bruzonsky for their legal advice, and to Dave R. for professional insurance insight. The support and resources from some forum members has been awesome!

So, I'll be in and out, around and about, but we will continue...

Mike

Are you sure the MAX 4224 is a faster chip than the H1100? I can't find the specs on the 1100 but I seem to remember that it was 800mhz. The MAX is a 600mhz chip. Am I wrong?

I know I had steaking issues with a 409 on the green tube and an 1100 on the vim. When the 409 was replaced with a 449, the streaking went away. This leads me to believe that one must have the fastest opamp at the end of the video chain or the same all the way through.

Walt

Originally posted by aspec2
Mike

Are you sure the MAX 4224 is a faster chip than the H1100? I can't find the specs on the 1100 but I seem to remember that it was 800mhz. The MAX is a 600mhz chip. Am I wrong?

I know I had steaking issues with a 409 on the green tube and an 1100 on the vim. When the 409 was replaced with a 449, the streaking went away. This leads me to believe that one must have the fastest opamp at the end of the video chain or the same all the way through.

Walt

The Max4224 is 600Mhz (-3db). The H1100 is 850Mhz (-3db) the CLC409 is 350Mhz (-3db).

Considering the CLC409's bandwidth (350Mhz), that should be more than required for the 120Mhz bandwidth of the projector, but somehow it wasn't. The H1100 (Harris HFA1100) is 850Mhz, but they switched to the CLC449 (1,2Ghz), and I asked about this change, and was told that the H1100 had a high frequency roll-off at certain high scan rates.

The Max4224 was a suggested cross or replacement, and so was the LMH6702 (720Mhz) that was supposed to be a direct replacement for the CLC449 (discontinued comlinear), but it's not. And why would they think that a 720Mhz chip is a direct replacement for a 1.2 Ghz chip :rolleyes:

I'm starting to look at this spec (Mhz/Ghz) thing as hype. Because none of the above really makes any technical sense...

So the specs seems to mean little of nothing

I'm starting to look at this spec (Mhz/Ghz) thing as hype. Because none of the above really makes any technical sense...
Interesting coincidence here... Just yesterday I pulled the complete specifications package for several IC's including both the National CLC chips and one I have been looking at as a possible replacement instead of the MAX4224, the Burr-Brown OPA658 (not the 685). The OPA658 loosely would be considered a 900Mhz chip.

When I looked at the operational parameters and charts, several things stood out. All the chips -3 db down point varies dramatically with gain, load (total load as well as whether it is purely resistive or has a capacitive component to it) as well as voltage output. Also, during rolloff it may be smooth and gradual or wavey with an irregular phase response relative to frequency. The various manufacturers do not always list the same parameters and graphs in the specs, so it's very difficult to compare apples to apples on paper.

Gain Flatness might be the most significant parameter, with the CLC449 at a max of .1 db from DC to 200Mhz, the CLC409 at 0 db from DC to 75Mhz (peaking) or .2 db from DC to 125Mhz (rolloff). The OPA658 clocks in at .1 db 0 to 135Mhz, so it may not really be a viable candidate in some circuits.

The CLC449 has that 1.1Ghz bandwidth at unity gain, non-inverting, with an output of .2v p-p. At 2v p-p it is only 500Mhz. With more gain, it drops even lower. The CLC409 shows its 350Mhz bandwidth only at 2v p-p.

The OPA658 doesn't have any information that plots bandwidth to output voltage but it does show the variance with gain. At +1 db of gain you have the 900Mhz. But at +2 db gain you're down to 680Mhz and at +5 down to 370Mhz.

The MAX4224 has basic specs that look similar to the OPA658, but it has a gain flatness of .1 db from DC to 300Mhz which probably makes it the best overall candidate.

I haven't looked at each location where CLC449's are used to determine in each case what the gain, output voltage and load might be, but I'd bet it's not consistent. And that just means you have to look at exactly what the parameters are for each circuit needing a sub to determine initially which part might be the best candidate -- not just the generic quoted bandwidth.

I Hope this is helpful. The only reason I started all this was because I needed a 449 sub in an 8 pin dip package for the loop-through switcher input boards. Only the Burr-Brown part seems to be available in that configuration.

--Bill

Bill,
the real strange one in the bunch is the LMH6702. This is a newly released chip from National Semiconductor. It is their direct replacement for the discontinued CLC449. I've waited 8 months for this chip to become available, and the minute I put it into service, I noticed that it was different. It produced a tad brighter image. I thought that maybe that was a plus, but when looking at the signal on a scope (20mhz). It was also oscillating (parasitic). I thought I was doing something wrong, so I tested the chip stand-alone -- less, but still some oscillation. So at this point I thought to call National and speak to an application engineer.

First I was surpised that he was totally unaware of this problem, and he wanted to make sure I was using a reasonable feedback resistor value, and grounding and bypass was also in order. We both agreed that my circuit was adequate. Now, why was I experiencing this problem:(

The application engineer said that he would look into the matter, and would also inform the designer of the chip. A couple of days later I got a call from the application engineer informing me that they were also able to duplicate the same thing that I was seeing, but not as bad. I could not believe that this chip failed a basic bench test. I asked if there was going to be a replacement or recall for the chip, he said no, because the designer (or actual manufacturer) of the chip claims that they could get it to work in their lab in a "more specific test" come on, it's an opamp, what test where they doing that could have possibly giving them these very impressive specs without them seeing the chip oscillate. In my opinion the chip is defective, but they say that I should try a redesign of the circuit, or should I say that they suggest to try and change several components around the chip to get it to work properly. This makes no sense to me, It's real simple, the chip is defective, period.

The manufacturer of this chip (LMH6702) should have those involved in the making and testing of this chip take a drug test.

Yea op-amps can be fickle at fast speeds.

have you looked at the slew rate of the chips
I find this to be a more useful number..
but how sensitive the chip is to output load is the real test.

The load will change from capacitance to inductive as you sweep the spectrum and then there is the resonances too.

Way back everyone in the computer world moved away from them and had the amps integrated into DAC

The circuit around the chip affects it so much. Just changing a brand of a resistor or cap can change things.
It is a manufacturing nightmare.

Mike it sounds like the LMH6702 might just be a lot faster?? What kind of changes do they want?


As for the spec (Mhz/Ghz) thing as hype
I would say yes it is..
for a 1Gig chip they will feed it a 1 GiG sine-wave
with the chip configured with NO FEEDBACK!!
and if they get some of that sine wave out -3dB
it is called a 1Gig chip ( I call it a attenuator)


I just looked at the spec for the Max part
I think my Gamma circuit has less ringing their test jig has looking at there plot.. :)

Mike, on that 6702 -- I've looked at its specs before and reviewed them again just now. I'm not particularly impressed by it especially with the gain flatness only out to 120Mhz, and it shows roll off well into the passband at every gain except +1. But it does have a killer slew rate of 3100 v/msec which is close to double all the others, though I'm not sure of the impact of that in this application.

I agree with you with your assessment of the chip -- it does seem to be broken, or at least mis-specified. I ran into to something similar a few years back with some wideband audio IC's where supposedly identically spec'd chips from two different manufacturers were behaving quite differently in the circuit. One was very quiet but didn't sound all that great, while the other sounded much better but had a really high noise floor, to the point of unusability. Noise was spec'd identically by the manufacturers.

It turned out that the reason for the high noise on the one brand was due to out of band oscillation which required a small compensation rc network in the feedback loop to keep it stable. Doesn't sound quite pin-for-pin compatible, does it? Well we finally learned that the spec sheets were in error and had (accidentally) omitted the need for additional external compensation. The chips were operating as intended but the docs were wrong. Sounds a bit familiar...

--Bill

Sly,
they (national) claims that the LMH6702 is a direct replacement for the CLC449. Compare the specs, how could it possible be a direct replacement. Plus it doesn't work.

Yes, op amps are very finicky, and the best replacement for the CLC449 would probably need to have a few components on the board changed. But let me assure you that the LMH6702 would not be a good candidate for this. This chip is just right on the edge of oscillating, and it seems to me that if it fails a basic bench model test, it'll definately fail a populated board test, where the chip would have to behave itself in a multi-signal noise bed. I know of and have used and tested much faster chips than the 6702, and none of them sheads tears during testing, some may need the feedback resistor values changed, or a more precise bypass cap, but none has done what the 6702 does.

Aside from the parasitic oscillation, it'll actually cause distortion of lines and bleeding of text in the video image (stuff I've never seen an op amp do).

Bill,
I think they made a mistake somewhere with this chip. It may be the wrong specs that applies to another chip, or maybe a mistake in design. I just hate it when they make the claim that it's a "Direct Replacement"...

Anyway, as you guys may already know, I'm already ahead of this, and do have a very good replacement.. ;)

...and the winner is? ...:)

Cary

Mike,

Is there another thread here at AVS about your component to RGBHV transcoder?

Or is this the best thread to learn about that new item?

Mike

I have changed the neckboard ic from h1100 to max4224 but noise at background occures. I have also purchased 3 nos. of CLC 449. Will the replacement completely removed my video noise problem and improve the image or is it better to change back to h1100.

regards

Joekjtam,
I've heard of this problem before, but I've yet to see it myself. I really want to see it because I'm sure I can make it go away without much hassle. If you have the CLC449's they will work, and would be better than the H1100's -- so go with the CLC449's.

Pocatello,
the transcoder is ready for testing, but I've been waiting on a few components that should arrive tomorrow, and then I will finish a duplicate of the unit I have. Once that's done the duplicate will be sent out for evaluation. And yes, that's when the new thread will start.

Cary,
I'm working on a Mike Parker modified VIM. It'll have that chip on it...

Mike,

I've been trying out one of the Marquee Switcher output cards that I changed from the CLC409 chips to MAX4224's. I have another to use as a baseline which has H1100's in it.

With the MAX4224's the picture is a bit sharper but it almost seems as though the output level is higher. The black level moves up a bit to low grey and the whites crush on things like certain facial reflections (which look like shiny skin), objects that reflect, etc. All detail above a certain level is just gone, however this crushing level is much lower than white overdrive on the projector and isn't corrected by lowering contrast. I'd say the 4224's have more gain than the others, but I don't think just gain would fully account for what I'm seeing. The card with the H1100's is a little softer but has no crushing at all.

When you changed to 4224's did you have to make any circuit changes at all? This is an extremely simple circuit, two bypass caps and a series resistor on each rail, 75 ohm to ground input, 75 ohm in series on output, and feedback consisting of one 120 ohm from output to inverting input, and a 120 ohm from that junction to ground. That's it. I added an external buffer amp to drive the cable to the pj just in case it was a current or load issue, but it had no effect on the visuals. This is the only active stage in the switcher for these tests.

If you have any suggestions I'll try them tonight and scope the H1100 and MAX4224 card outputs for comparisons.

Thanks.

--Bill

hmm what is the bypass cap value??

and post the scope pics..
also what is the voltage on the rails??
how close does the output of the chip get to a rail??





I would think you would need at lest 2 caps on each rail
a 0.1uF and a 0.01uF or something like that.

Bill,
the MAX is a much faster chip than the CLC409, so I would suggest that you first try changing the resisitors values to around 249 ohm or something near that. 120 ohm feedback/ground resistor is definately too low of a value for the speed of the Max4224. I would also replace the bypass caps and put one 4.7 mfd and one 0.1 mfd in parallel on each rail (+/-) of the chip.

Mike why the 120 to 249Ohm??
Is the 120ohm too inductive?
Or is it just to reduce the load on the amp..
If both are changed the gain is the same.
The only thing is the impedance is less.
So I guessing you want the resistors to be more capacitive and less inductive?

I would also think that a .1uF would be to big with the faster chip and something like a .01uF would be needed.
or maybe something even smaller yet.
The .1 is good for the 20Mhz range but the faster stuff it is too inductive and just has no capacitance.
The chip might have internal caps for the fast stuff..
and not need it??
The caps do seem to be getting better.
They just moved the cap analyzer to a different campus.
I would have loved to sweep a new .1uF and see where is becomes a non-cap. If I get some time I might drive over and try it and see how the newer .1 perform

Sly,
by changing the resistors to a higher resistor value, you'll lower the "parasitic oscillation" of the chip by lowering the chips operating bandwidth.

What Bill described is the same problem that I was having with the LMH6702 (exactly). If he's able to scope the chip he should see the oscillation, and as you know, the solution is to increase the resistor values. This also what the application engineer at national semiconductor had suggested.

I know the theory behind the 0.1 as bypass (oscillation), but with a higher capacitance cap network, I've gotten very good results (lower noise on the rails).

Mike thank for the clarification.

So I see where it is going.

So instead of adding a very small cap from the output to the (-) input.
You increase the resistors so you get the increase in capacitance for free. very clever.. saves a part too..
It is kinda complex because is also increases the capacitance from (-) to ground too. But the capacitance
of the pin its-self stays the same but the impedance is less. There is a lot of interaction going on to think about.
But I guess when all is done it is matching the capacitance of the (-) pin to the feedback loop as well as
making the (-) look more capacitance in general so it will get less crosstalk from the (+) pin..

I think I knocked a dust-bunny out of my brain that part has not been used for such long time:)

I'm sorry I haven't gotten back to this by now. I will try the resistor changes in the next couple of days. If I need to do further testing I'll open up and extra switcher box I have just to have a chassis and power as a place to work.

The bypasses on it are already 10mf and a .1 so that's not too far off from what Mike was suggesting. And the only thing in this box is input cards, about 1/2 of which are passive, so there's not much garbage sneaking in on the rails to be concerned about, so I guess self-oscillation would be the only issue.

Also, I got a bunch of the Burr-Brown OPA658 chips to try on the bridging input modules in place of 409's. Haven't tried those yet either, but if I was thinking I would have realized how incredibly simple it would be to convert those to passive inputs by disabling the passthrough and electronics altogether. I have a bunch of them so will try it both ways.

--Bill

My most recent challenge has been figuring out where some video signal noise that I see in certain types of scenes, test patterns and solid colors, is coming from.

To my surprise there were two different types of noise from two sources. One (the most obvious one) that looked like hazy fast and faint horizontal phase bars and beat patterns turned out to be the ATI video card. Many of the ATI cards seem to have varying degrees of noise that you may or may not notice depending on your pj/tv/monitor and viewing habits. Older 9500 Pro's (with a part number NOT ending in 110) may have it severely to the point that vertical lines may zig-zag slightly.

There's a more subtle noise that comes and goes (varies but doesn't completely disappear) and looks more like troughs of light rather than hazy bars. You might also call it very fat herringbone which randomly changes speed. The troughs are wider top to bottom and they pass up or down on parts of the screen (usually in the lower 1/2) on all colors, and are most visible to the eye on white or cyan. This one is coming from the projector itself (using the internally generate signals) and exists even when there's not a single cable connected to it. It's not what I'd call severe but I don't believe it should be there at all. It is the same when a signal is connected, i.e. the amount of noise in a white screen is the same whether it is internally generated or from an HTPC or other source. It seems to be more noticeable at higher vertical rates such as 72Hz, though it is there at all rates.

Possibly of no relationship, a lot of noise and fluttering can also be seen in each tube when it's at its 0 IRE screen level. Looking in the lens at the low level raster it is quite plain, but for the moment I'm attributing it to amp noise that is normal.

So my question is about what is normal on a 9501LC? If there absolutely shouldn't be a trace of anything, what might be a likely cause of this? Do any of the wires between horizontal, convergence, vertical or astigmatism boards and the tubes need specific positioning or are there any other areas that might be sensitive to radiated signals? Unless I have more caps opening up or leaking somewhere, this is about all I can think of that might cause what I'm seeing.

Any advice?

--Bill

Bill, that noise is normal (depending on the amount), or should I say it is there even on a new unit. It's also there on most CRT projectors, especially the high end units.

That noise has been on my agenda for the past two years. I have spent a many day tracking it down. It was a tough persuit, but I found it and for the most part, I can make it go away.

Last month at the gathering where we had the Marquee 8500LC. I was asking if anyone could see the noise in the image or on the test pattern. No one was able to spot that usually noticeable noise. Earlier this week I drove 65 miles to one of my most critical customers, who owns a very tightly tweaked 9500LC and a Digital Leeza. I removed his VIM module and today I returned and reinstalled that VIM. This customer has had that 9500 and Leeza for a year now, plus he notices everything in the image, so he was a perfect for this. After I reinstalled the VIM and without any calibration, there was a very noticeable difference, and this is the things that stood out:

- the noise was not noticeable in the image or the test pattern.

- the colors had more punch.

- the black level and shadow detail was the biggest improvement.

I removed his VIM, but had already done the rest of his projector...

I just want to second mike's comments about the VIM board upgrades. The picture was terrific before and even better now especially the black level and lack of noise.

Mike, I take it you're referring to the high amplitude noise, and not the background raster noise? Or are they one in the same as far as the mods are concerned.

So don't keep us in suspense! How will this new found knowledge relate to us out here? Is it something you plan to disclose in a future topic (where *have* all those wonderful topics gone, anyway?) or is it something you plan to do as a commercial venture?

I could make a guess as to what the fix is, but not precisely how to implement it correctly the first time.

Any more details?

--Bill

Bill,
there are four sources of the moise inside the projector. One is the power supply, though it's an industrial grade very fast, highly filtered switching power supply, it still puts some noise on the power supply rails (and this is true with all switching power supplies).

The second is the control module (main computer), as you know it has massive capacitor decoupling on its 5, 15 volt rails, and even some of it's data lines (SDL, SDA, Int and analog ramps). This is supposed to isolate (as much as possible) that computing noise from the rest of the projector.

Third, is the interstage noise. Each board or circuit can generate it's own waste onto the rails or data lines. Especially the high voltage power supply and the deflection module that has high voltage/frequency pulses.

Last is grounding and circuit board design. And that includes some of the components.

This noise will show up in the internal test patterns, especially the white field. And it will and can take up a small portion of the image detail. Normally it would appear as a haze or slight snow in the background of the image, and there's usually a little random motion (mosquito/bars, etc) in the image. So there's two levels as you indicated. One is hash and the other is mosquito/bars. The very high frequency noise is usually displayed as hash (snow, cloudiness = loss of black detail). While the other stuff is usually jumping around on the screen (mosquito) or rolling.

In my research I've talked to engineers at Christie Digital, Microenergy (makers of the low voltage power supply) and Engineers at Washington Labs (FCC compliance laboratory), plus I have a ton of my own experience with noise, the guys at Washington labs was excellent. The lab guys do the testing of equipment for UL and FCC compliance. They also make recommendations to the manufaturer on what to do if the device has an unacceptable noise level, so they were a wealth of help here, plus they always invite me to their seminars.

There are some special components that are being used, but none of them involve expensive or special caps. Fact is, I do swap out several caps in the projector, and I do that to rule out any potentially leaky caps, but I can accomplish the same results if the original caps are good.

You got PM!

Shortly my 9500LC will be up and running with all of MP's latest mods (let's call them improvements). Almost all the caps in my Marquee were replaced with higher quality caps by KBK prior to MP working his latest magic and my machine came with an Ultra VIM module.

My new HT is not yet finished. ARS sound treatment gets installed in two weeks. Then we have to recarpet, install 12 chairs, wall sconces, wood trim, the speakers including 2 massive in ceiling transmission line subs, the 4 way masking Stewart screen, and the projector. It will be about another month before we reinstall the projector.

According to MP the new caps on the control board do result in a measurable 2db noise lowering but the improvement according to MP is dwarfed by the other mods or parts improvements MP is performing. MP has my back up 9500LC running in his house. It has several of his latest mods. Despite it having 2 very worn tubes, it looked very very good when my wife and I visited MP and his wife Saturday, no visable noise, none. They are close friends of ours although I honestly believe this does not bias my assessment of MP and his work. Trust me campers, I am one of his harshest critics and we do disagree on several things CRT related.

MP has been working on a component to RGB transcoder. It makes the other out there look and act like toys. If you use a transcoder to feed your Marquee you will be in line for a very substantial improvement when he brings the transcoder to market. This baby will not be cheap because it is packed with quality parts and SOLVES what's wrong with the several transcoders out there. It is hand tuned. As Phil T likes to say, "stay tuned"!

THIS HAS BEEN AN UNPAID NONPOLITICAL MESSAGE.

I tried the MAX4224's with modified feedback values in the switcher output card. The closest I could get to 249 ohms with metal film was 220 ohms, so I used those.

As far as the white crushing I don't see much if any difference from the earlier test. If there is any improvement it is slight at best. I tested it driving the pj directly as well as through an Extron distribution amp, but the results were the same.

On a whim and since I had another extra output card I decided to bypass the output circuit altogether so the switching could still be utilized but the card just passed its input directly to its output passively. On the output jacks there now is a 2' cable set connecting to the Extron DA now acting (essentially) as the output driver for the switcher and driving the pj.

That was an unexpected surprise. At first glance the image looked softer with less edge detail, but as I watched for awhile I noticed that the detail within the image was greatly enhanced. Small facial colorations and details were very clear and there was a lot more visual depth to the image. And compression artifacts at the digital level were much clearer, but very separate from the images. It appears as though the output card I've been using as the reference with the H1100's adds a sharpness to the edges that isn't really there, while robbing the image of more subtle detail. I had always assumed the two (sharp edges and image detail) go pretty much hand in hand but this change says otherwise. The MAX4224 card makes the edges look sharper yet, but the loss of subtle detail even before crushing is quite obvious.

It would be outstanding to accomplish a modification to the output cards which is equivalent to this clarity, but I'm not sure if it's possible. I guess next is looking for oscillation on the 4224 and how it is affected by feedback changes.

Any other suggestions?

--Bill

Bill,
yes, the key is the background detail. That's where the resolution is.

We've been so used to looking for sharpness on the foreground and edges, that we never knew the background could look so good. But once you've experienced this, it'll be hard to tolerate anything different.

Concerning the "white crushing", you'll probably have to play around with different resistor values. Also, you should replace the 75 ohm resistors with a better quality resistor. If you want I could send you a batch of the right resistors to try (including the 75 ohm), but the 4224 should work great. And there's nothing worse than the CLC409's, so if you stay with it, you'll make that switcher something really worth owning.

Bill,
it sounds like the you have overshoot

too much boost in the high end also knowing as peaking
I needed to add a 1pF cap on the gamma circuit I did
form the output to the (-)
but you need to look at this with a scope.
A fast pulse generator is also a very nice thing to have
to feed it.
Without a look with a scope or even seeing the picture there is a lot of guessing going on.
As for the crushed(saturated) white that sounds like the voltage of the output is getting too close to a voltage rail and clipping.


Does the image seem more contrasty through the circuit??

you might need to reduce the gain a bit

A smaller resistor from the output to (-) vs the the resistor from (-) to ground will do that

also check the supply voltages are they +5V and -5V ??

Mike, thanks, I'll take you up on that offer. It will certainly help me zero in on the issues with that card. With this passive output in place I'm now looking for the chip/circuit that I can't "see", which in the analog world is like searching for the Holy Grail.

Sly, no there's no increase of contrast, if anything it's (visually)lower than before the fb resistors were changed. I haven't checked rail voltage or done any scoping on this yet. If rails were insufficient, though, I'd expect to see the same characteristics on both chips not just the 4224. How fast on the pulse? What freq and width?


Slightly OT, but there is a PBS HD production that is worth a look. I don't know the name of it, but it's about 1/2 hour in length and is a story with Ashley Judd as a reporter interviewing a 100 year old man on his birthday. The video on that is pretty excellent, though you do see some minor compression artifacts.

--Bill

I haven't been following this thread at all, but had to take a look.

Can I just say (and I mean this in the nicest of ways):

You guys are insane..;)

(Can I come play in your sandbox for a while?..:)

I'm so jealous!

Curt

Curt,
thanks for stopping by, It's good to have some of the technical elite to stop in. I was also expecting to hear from KBK by now, but with Sly around, it's almost like KBK has been posting.


Guys,
I have the fix to make the Max4224 work on the VIM. I'll post it later...

Mike

I have changed the h1100 to max4224 and found that there is ghosting at bright scene and background noise is severe i.e. bright background lines.

Is there any way to modify the curcuit to resolve the problem.

regards

Originally posted by mp20748
I was also expecting to hear from KBK by now, but with Sly around, it's almost like KBK has been posting.


You know, I thought that awhile back but bit my tongue :D

Originally posted by joekjtam
Mike

I have changed the h1100 to max4224 and found that there is ghosting at bright scene and background noise is severe i.e. bright background lines.

Is there any way to modify the curcuit to resolve the problem.

regards

Yep, replace R12 (261 ohm) with a 331 ohm, but it must be a metal film @ 1 %.

R12 is the feedback resistor for U2 (Max4224), so it's right near the chip.

"How fast on the pulse? What freq and width?"
the speed is not that important but the edge rate and the quality of the pulse is.

A nice fast edge is great for tuning.

I use a HP8131A 500MHz pule generator.
I had it running at 9nS around 111Mhz

But as long as you have something that will give a nice clean pulse with a fast edge it should work.

Then see what is wrong with the output.
Overshoot?
A small cap out to (-) should help that.
or if the edge is slow and rounded add a small cap from the (-) to ground.

And the power needs to be good.

If you are probing around the circuit itself a fet probe is needed and even that will affect it some.
I added a 1pF cap to adjust the circuit.
If you look at your probe how much capacitance does it have??
So you can see just probing will change the circuit.

Also you will need a in-line 75Ohm term to put on the scope to look at the output. Or a tee on the scope with a 75Ohm term.
The tee might cause some ringing on its own..

And a 50 to 75 ohm feed through for the pule generator.

you then should be able to bypass the amp and check the source it should be perfect. if not it will be hard to tune to

You can try it with video just use vertical stripes
But the video card might not have a clean pulse to start out with?

joekjtam,
have you replaced R12 on the neck boards?

The Max4224 produces a sharper test grid, but it has the distortion problem with the original 261 ohm resistor. When I changed that resistor, the distortion went away. We need to confirm that this is a fix with every version of the neck boards.

Sly, thanks for the information. Unfortunately I have nothing that produce a pulse that fine in that frequency range. I've have a look around to see what might be available.

I think the probes I have are 10pf so probably aren't adequate either.

Oh well, I guess it's vertical stripes and eyes!

--Bill

If you can term the signal to 75 ohms you still can look at the video out. you do not need a probe at all for that..
And if you compare the video going into it vs out you should be able to see what is going on.

The first thing is see what it is doing. then go from there.
if it is oscillation you will know right away.
or if it has massive overshoot. So you do need to look at it.

I would start with your video out of the PC
run the H sync to the external sync on the scope
Set it for a nice hi res setting and load the stripes.
then plug one of the colors into the scope with a 75 ohm term at the scope. A in-line type is the best but a tee with a 75 term plugged into it will work OK too. or just a 75 ohm resistor soldered into the tee keeping the leads as short as you can.

Now look at one of the stripes and note how the edge looks .
now plug that color into the amp and plug the output of the amp box into the same tee/in-line term and note what has changed.

mike

I have not changed the resistor yet as i am waiting for the method to use max4224 to the vim so that i can ask my tech people to do it for me at the same time.

Thanks for your help and assistance.

I changed only the 3 "RGB" chips in my switcher to the MAX4224 and it seems to be working OK. The picture is sharper. The two sync chips are different on my switcher and also use different feedback resistor values then the RGB lines. 121 Ohms on the CL409's and 511 Ohms on the CL430's (sync.)

Should I change the two 121 Ohm feedback resistors on the new MAX4224 chips?

What are you displaying with?

I never could get next to the look of it with the 120 ohm resistors in there, and tried pairs of 220's as a test. But it still didn't look right, especially not compared to subsituting the Extron amp as the outut. So I'm still in a 'get back to it' mode awaiting some parts from Mike before I do much more.

You shouldn't need to do anything with the sync circuits that I know of, just the RGB amps.

Also, on that switcher, make sure you have the passive input cards (they have a single input for each line with no bridging). The bridging cards are active and also have the 409 chips in them.

--Bill

300 to 500 ohms for the feedback/grounds should be about right.

Bill, can you shoot me your address again. I now have that VIM ready for you to check out, I'll also send the resistors, but it may be next week before I can get them out, I'm headed for the desert (Arizona) this weekend.

Mike, I take it that the actual value isn't that critical as long as it's uniform between feedback and grounds, and each stage? I'll try something in that range that I can find six of on hand.

I sent you address and other info in email.

--Bill

All of my input cards are passive. I will try some 330 ohm resistors. I only have standard resistors (not SM). If the 330 Ohm metal film resistors work I can order the SM parts.
I'll let you know how it turns out.

Mike,

I have change the R12 to my EH8000 neck board to 330R and the ghost / background noise have been better. Formerly there is 10 ghost to sharp object at bright scene and now it have change to 4 and not very noticeable during movie. Should I change R12 further to improve.

I have other problem with my EH8000.

1. banking on bright scene for red colour eg. during explosion when the whole scene is fill with red colour. I have check my neck boards. They are rev. 1 neck boards and the green one have many modification done including many small capacitors but the blue and red neck board does not have the modifications.

2. During the checkerboard test pattern, the middle portion is a little red tinted, but it does not show up in other testing such as grid or white screen. Is it only the problem with the pattern generator?

Mike

I have change the R12 to 330R.

The amplification factor of the ic seem to have changed and i have to adjust/lower G value to cope for the new ic. The result is very good. I have a much sharper picture.

Can you advise how to use max4224 at the vim.

Just a quick question. I may have found some CLC449 chips but I need the full part number, it should be a CLC449AJ?. Does anyone know what the ? should be?

thanks,

Andy

Andy

It should be an "E", the CLC449AJP is the 8 pin DIP package. The
CLC449AJE is the 8 pin SOIC package. The latter is the package
style used in my Marquee.

DH

I ordered some CLC449AJEs and am waiting for them to arrive. I should have some extras if anyone is interested.

Walt

Originally posted by joekjtam
Mike

I have change the R12 to 330R.

The amplification factor of the ic seem to have changed and i have to adjust/lower G value to cope for the new ic. The result is very good. I have a much sharper picture.

Can you advise how to use max4224 at the vim.

This option is only if the CLC449AJE's are not available. If you can get your hands on the CLC449's go that way. The Max4224's work well on some versions of the VIM's, but not on all, and would require some circuit changes. This could be a little mor complicated, unlike the neck boards single resistor change.

The IC's to really replace are the CLC409's. Some fail by smearing in the image, but without the smear, they are low performers. Replacement of the CLC409's will produce a sharper and more dynamic image.

So Mike, is it your opinion that the MAX4224's with the 330 feeback resistor is superior to the CLC449's? Or if 449's are already present, just leave them? If there's *any* improvement in noise or detail and no other 'expense' to deal with it seems like it would be a worthwhile change. Is there noticeable anything?

--Bill

Would someone enlighten me?

I have seen several stigmator boards that have shown signs of overheating of certain components. Usually the worst is in the red channel, R235 and R203 (sometimes almost ashes), with R135 not far behind. Then if overheating present at all in the blue channel, it's R635, R603 and R535. Every so often the Green might be affected with either R435 or R403 showing signs of heat. These are the same resistors electrically in each channel.

Is this due to generally overdriving the tubes and/or overcorrecting on stigmator adjustments, or is there another cause like defective tubes, other components or something? From their position in the circuits, it looks like stig drive outputs (correction) could be just set too high. (203, 403, 603 are -15 rail, 135, 335, 535 are Y stig output shunt to ground, 235, 435 and 635 are X stig output shunt to ground)

Are these the only components affected or could you say that if these are overheated then xxxx parts are probably bad too?

Thanks for any insight.

--Bill

Toto, do you think we'll be back in Kansas anytime soon?

"Anyone for some Marquee maintanance"

That depends on where you bought it, Hee Hee Hee.....

Chip S.

Not again! don't start

Bill,
I've not experienced that. I've never seen a burn on that board, or had a problem with the astig drifting from bad or weak components.

I've seen some Marquees that operated in high temperature installs, that the vertical, convergence, focus, neck and deflection modules showed burn on the components, to including the boards themselves, but not the astig board.

I'll keep my eye open for that.

-----

Can anyone help me find a source for mini RCA's (male). I've purchased some, but they did not fit the female plugs on the boards (the center pin was too large).

I'll need these to complete my VIM/Neck Board mods. I need to know the cost for these, or if I'm able to offer them.

Mike, to minimize the search, or at least keep from doing double work on everyone, what major places have you looked? Seems like a year or so ago when I got my Marquee, I figured out that someone had them. Someone here posted the size specs, and in thumbing through a catalog I saw 2 different vesions and one happened to be the right one after frther inspection. I never ordered, but I'll look for that catalog.

Jay,
there's a link on this thread. I made my purchase from that link.

Let me know what you find.

Thanks Mike. I guess I've just been lucky. :-)

[opinionizing mode ON <<--Bill]
In looking over the circuit more I'm pretty sure now that only a couple of things could cause these particular parts to overheat.

Other components on the board most notably leaky bypass caps on the rails probably could.

Bad astigmatism coils could, but I'd be hard pressed to imagine more than one of those having a problem at a time.

The most likely candidate would be from a machine that had been retubed and the flaring/centering rings way off adjustment. There would be a tendency to try and correct the poor astigmatism by over-adjustment electrically which would mean a lot of additional current demands and higher output correction voltages from that card. That demand may be beyond normal design parameters and cause rail series resistors and output shunt resistors to slowly cook.

Thinking back on it, this is exactly the state my Marquee was in when I received it over a year ago -- the stigmator card in it was fried very similar to this one. In fact, at one point it made a little pop and the red stigmator channel quit working altogether. That was probably when one of the rail resistors for red cracked in half.

--Bill

Mike,

There are actually two potentially useful links earlier in this thread regarding the rca mini plugs. Sly had sent this:I wet to work and grabbed a set of calipers
and the VIM's mini RCA's measured 6.35mm with a 1.7mm PIN

so it looks like CTP-174-6A-1
and
CTP-175-11 with 1.7mm pin option
is what we can use..

So Bob found them...
He's referring to http://www.connect-tech-products.com. In there, some of the plugs have a smaller pin diameter option, and the CTP-175-7a comes with the 1.7mm pin standard.

The other link is to http://www.excess-solutions.com/audio_connectors.htm which has a model ES519 that appears to be a mini, but I can't tell about the pin size. There aren't any other usable specs on that site.

Which one of these did you try?

--Bill

Cool!

I went with the Excess Solutions. They were the wrong ones. I'll get right on the Connect Tech's, they're what I need.

thanks

Yes I got some from Connect - Tech

the parts are cheap they gave me some samples I used on the gamma PCB

I talked to a Bruce Creedy (510) 656 8490
The local guy here.
The parts are so cheap I see no need to go to a junk shop.

My friend and i have marquee9500LC.
i have tweeked input board capacitor mod and taking out color corection board, which gave exllent result.

I am planning to exchange mini RCA cable with tefron shield silver cable.


i need at least 12 mini RCA plug ( male )., hopefully 30.

Is there any one send to me in Korea.

Of course i will cover all expense.

they are selling 1000 baisi. i don't need that much.

May be i can send good tefron shield silver cable or HD tapes which recorded good Korean ATSC or Japnese BS HD .

i hope some body help us.

Hi Guys,

I hope you don't mind me barging in here ;o)) I have two Marquee projectors. One being a 8110 which I converted to a 8500 (well almost, it's missing the stigmator stuff) and a 8500 Ultra. I plan on completely rebuilding one of them for HT use. The other will stay in reserve as a parts supply. Of course I'll be having questions here and there ;o)) I have already taken the Ultra completely apart. Doing so I took pictures of all parts, which I plan on showing on my website (as soon as the pictures are up, I'll release the URL ;o))

Of course, I have been reading avsforum up and down and in particular this thread with great interest. I searched the web for some infos on the CLC op amp and found the National Semiconductor part number LMH6732.

Mike: Have you experimented with these, yet? They seem (specwise at least) to be an interesting alternative.


Greets,
Reinhard

Reinhard

Your 8500 Ultra should already have the preferred op amp, the
CLC449. We have been replacing the older CLC409's with CLC449's
in the older marquee's (8000's). Then the 449s became hard to find
until recently. My pal Walt found a source, but he had to buy a whole
tube (95 of them). If you need some CLC449s PM me or Walt (his ID
is aspec2 on this forum), we have several left. For a good price too.
Mike P. had tested one of the LM chips, but his report on it was not
good. The MAX4224 was a good replacement and even though it has
lower specs than the CLC449 I found they worked well on the neck boards.

Good luck with your refurbish of the 8500. Indeed, why would anyone
pay some technician to do that when he can have all that fun himself.

DH

Hi Donald, all,

>Your 8500 Ultra should already have the preferred op amp, the
>CLC449

I haven't looked into the 8500 Ultra yet. I just took it apart ;o))

The Ultra tubes are more worn out than the 8110 tubes. The 8110
tubes have 2800 hours on them. Red is mint, green has a very slight
mask (you really have to look hard to see it. I would place it between
0-1 from the CRT burn website). The blue has a slightly stronger mask
(maybe like 1 from the CRT burn website). The 8110 came from a virtual
cave experiment. The 8500 ultra came from the development depo at Audi
and this machine has around 7000 hours on the tubes. The red has almost
no mask (maybe a 1), the blue has a slight mask (somewhere between 1-
2). Unfortunately the green tube has a very visible mask (3-4) in 4:3
format. To make matters worse it not only has one mask, but two (same
size but shifted to the right, when viewing the tube from the front).
However the two other tubes will go into a box as replacement ;o)) and
the 8110 tubes will hit the 8500 Ultra chassis.

>the 449s became hard to find until recently.

Yes, I read that. That's why I was looking for replacements. I missed
your post that you found a new source for the CLC449.

BTW: Where are the tube hours stored? Does the EEPROM on
the backplane board hold this info?

>Good luck with your refurbish of the 8500. Indeed, why would anyone
>pay some technician to do that when he can have all that fun himself

Hehehe. I *am* a technician. While not on CRTs but for welding equipment
(from Mig-Mag to Tig all the way up µPlasma and stud welding) ;o))

And since I also dabble with Tesla Coils (250kV upwards), I think I
can handle the 34.9kV ;o))

Should I keep the (or rather move) the stigmator coils and the two control
boards when exchanging the tubes? Running the 8110 on a windows
desktop (yes, low brightness and contrast ;o)), I noticed the 8110 is a
little fuzzy on some parts of the desktop (esp fonts). I think the stig
stuff would help reduce this to a minimum. If I deconverge the
three colors, I also noticed that red (using the dot test pattern) isn't
100% round.

Of course I won't be using a windows desktop on the CRT ;o)) However
it shows me points where the projector needs more adjustment. (Yes, I
didn't take *that* much time to adjust the 8110 ;o)) since it isn't in the
final installment, yet.

Next question: I looked at the belly fans (for the tubes). Why did E-home
only make one hole for the air intake down below? (will be running the CRT
right side up & floor mounted) instead of one set of hole for each fan?
Wouldn't this increase air flow and reduce noise. If I remove the top cover,
I feel a strong flow of air from the fan (more or less) directly above the
hole, but from the two fans left and right from the hole (i.e. red and blue),
there is *much* less air flow (at least from gut feel). Seems to me the way
E-home did the air plenum design below the fans, the fans fight each other
for air (i.e each fan tries to suck as much air as possible, but has to fight
the other two fans, which are trying to do the same).

Also would it be a good idea (if I add the two holes) to add three (window)
screens to prevent the fans from sucking in too much gunk? I wouldn't
mind vacuuming the screens every few months to prevent clogging.

LVPS fans: Would it be a (good) idea to build a duct between the LVPS
and the HVPS (so that air is really pulled through the HVPS)? I was
thinking about cutting part of the LVPS case facing the HVPS. That should
also reduce noise a bit (will be foam padding the LVPS case and fans).
The duct between the LVPS and the HVPS would be fashioned from
nylon (so flash over or voltage creeping shouldn't be a problem)

Last question (for now): What is the color correction board for? I have
this board in the Ultra, but not in the 8110. Should I remove or keep
it for HT use?

Oh, BTW: I paid 2200 Euros for the 8110 and 1500 for the 8500 Ultra.
Are these prices okay? (I think so ;o) Just wondering what these CRTs
run for in the U.S.)

Greets,
Reinhard

Bump ;o))

C'mon guys, you have gotta have some kind of comments ;o))
on my ideas (even if you think they aren't any good)

Greets,
Reinhard

Reinhard,
You're asking too many questions at one time, you'll need to do a little at a time.

So can you start over, but ask one or two questions at a time?

Hi Mike!,

Okay, I guess I got a little carried away with the number of questions ;o))
It's just that I have been looking over the projector and this forum, so I
posted a few (too many ;o)) questions. So let me start with the last two
questions again:

What is the color correction board for? I have this board in the Ultra, but
not in the 8110. Should I remove or keep it for HT use? I plan on running
the Marquee via RGB using DeScaler and a Radeon (probably the 9800).
If I can find a Multistandard Decoder board I might also run it via S-Video.
I was thinking about using an Image Shifter (unluckily I don't have this
baord either ;o() when watching TV so that the channel "icons" don't burn
themselves into the CRT tubes.

I paid 2200 Euros for the 8110 and 1500 for the 8500 Ultra. Are these
prices okay? (I think so ;o) Just wondering what these CRTs run for in
the U.S.)

Greets,
Reinhard

Mike I have a 8111 great unit got it for a song run time are 356hrs on 432 standby. it was a demo unit. Well my quest is as so, like everybody around here Im trying to get the best pic possible I think my best start was to get the Marquee. I purchased the stig amp board still needing the rest to accommodate the entire system. Is the stig system going to give me a sharper playback via my HTPC ofcorse, or will just a simple good video card like the radeon 9500 do the trick? Im using ffdshow on the htpc for video processing but must admit it isnt the friendliest in use. Will a modded video card such as yours do the trick, or like I mentioned the stig system is the way to go?

Well on the fans for the CRT;s
I think the FCC EMI had something to do with it as well as X-ray leakage.

also think what will happen if one of the three fans fail..
air plenum will help some but the air jet affect is in play also some of that air is redirected in a hole to cool the cards
CPU and video input cards in the rear card cage.

Hi sly,

>Well on the fans for the CRT;s
>I think the FCC EMI had something to do with it as well as X-ray leakage.


Hmm, maybe that was E-Home figure, but if I use metal window screen,
this shouldn't be a problem in either case (EMI or X-ray). Look at the cage
of a monitor. It's full of holes. (Okay their voltage is slightly lower @ 27kV)

>also think what will happen if one of the three fans fail..
>air plenum will help some but the air jet affect is in play also some of that
>air is redirected in a hole to cool the cards.
>CPU and video input cards in the rear card cage

I'm afraid I don't get that ;o)). The plenum itself won't change. In order
for that to happen, what (I think) you are talking about, the air plenum
would need to be much deeper. The way E-Home did it, air convection
won't *really* happen. If one of the fans fail, there will be absolut minimal
air flow to that particular tube (simply because the other two fans will
suck the convection action away). If I apply my mod, however, convection
will be possible as each tube has a "chimmney" pretty similar to the old
days of RF transmitter tubes. Since the whole chassis of the E-Home acts
as a heatsink for the electronics, I would imagine that it's most important
to keep the tubes cool.

It most certainly would be interesting to see the actual air flow through
the projector. Maybe using smoke to make the air visible.

I remember in a post you wrote (at least I think it was you. Or was it
Mooneyass?) that you where thinking of building a thermal watchdog
system that would switch the projector off in case of temperature rise.
I'm not sure if this would be easy to realize as we don't know where
thermal hotspots can occure (first) in case of fan failure. I think it would
be easiest to monitor the fan rpm (of course this means exchanging all
5 fans with ones having an internal tach-generator.) Since we aren't
interested in actual rotational speed, it should suffice to check if all 5
fans are delivering pulse trains. As soon as one fails we could shut off
the mains supply.

Greets,
Reinhard

I was up too late ;)

what I was trying to say was. That the air plenum is there for fan fail I think.
But due to air jet is is mostly ineffective. But is does work some if you stop a fan you still will get some air flow from the other fans. Air convection has nothing to do with it.
Also most of the chassis is NOT a heat sink. It is air cooled for all parts except the one big heat sink in the back. But everything else is air cooled by forced air.

I removed the three fans in mine and use a remote blower with a air filter on to keep every thing clean.
so having a shutdown system is good for clogged air filter and fan fail. I think there is some very easy hot spots to sens for bad air. Mine is still a work in progress I have been very busy doing my day job so I have not had time to finish it yet.


Perfed metal should work OK for EMI but window screen will not. It kinda works but not too well for many reasons.

The stuff used for EMI can be very complex built up of layers of different materials electroplated one layer after another. At lest the stuff we use is..

Also I was trying to say do not forget about that hole that sucks air over the rear card cage. The boards in there need cooling too.

Hi Sly,

> I was up too late

No problem :D That's what discussions are for.

>what I was trying to say was. That the air plenum is there for fan fail I
>think. But due to air jet is is mostly ineffective. But is does work some if
>you stop a fan you still will get some air flow from the other fans.

Okay, I see, but this only helps *IF* a fan fails. In normal running
conditions, the E-Home plenum is more hindersome than helpful, as it
isn't high enough (okay I realize that E-Home's designers didn't want
to make the projector too tall).

>Air convection has nothing to do with it.

But the air convection would work if there was an opening below
each fan. You have the hot running tubes above the holes and cool
dense air below. This air would be pulled into the projector by pure
convection and should help, if nothing else, cool down the tubes, should
a fan fail.

>Also most of the chassis is NOT a heat sink. It is air cooled for all parts
>except the one big heat sink in the back. But everything else is air cooled
>by forced air.

Ah, misunderstanding here. What I wanted to say is, since the case is
made almost completely out of aluminum, it helps dissipate heat. Almost
all parts (be it the horizontal stage, vertical stage, LVPS, HVPS) are screwed
into the chassis. While this isn't the same as directly mounting the driver
transistors to the case, it still helps take the heat away (the heat from the
heatsinks is transfered to the specific housing by air and from the housing
the heat is dissipated into the case (okay it's indirect cooling ;) )

>I removed the three fans in mine and use a remote blower with a air filter
>on to keep every thing clean.

Hmm interesting idea. However, this is one step too much (for me ;) )
Do you >blow< air into the case or do you suck the warm air out of the
case? Either way, I'm not quite sure if this has the same effect as
having the fans actually running inside the case. I assume you are cooling
the projector by overpressure (i.e. blowing cool air into the case).
This would change the air flow characteristic (and the way parts are
cooled) in the chassis parts >below< the fan mounts. It might even
distort the air flow coming from the LVPS (just taking a wild guess here).

>I think there is some very easy hot spots to sens for bad air.

Let me clarify what I wanted to say: Of course there are certain places
that will heat up, should a fan fail, but how can we be sure that we
cover each and every part that might be thermally stressed? This might be
as easy to spot as a horizontal deflection transistor, yet it might also be
a simple resistor, which when properly cooled does it's job, but if air flow
fails, can die perhaps taking more expensive parts with it.

My point was simply if you really want to measure the temperature at
each and every possible failure point, this would amount to a lot of
sensors, some of which might even be difficult to affix to the part in
question (be it a voltage problem or simply to physically glue the sensor
to the part). If we simply check that the fans are rotating, it simplifies
things quite a bit. Setting the point (i.e. in case the fans still rotate,
but not fast enough) where we say the fan has failed is easy enough.

I still hope to catch Tim's eye as he seems to have really investigated
the air flow within the projector

> Mine is still a work in progress I have been very busy doing my day job
>so I have not had time to finish it yet.

Keep us posted as your work progresses. I'm most definitely interested
in your results. Since the projectors last as long as they do, E-Home
obviously didn't do that bad a job in designing the cooling system. But
then, it's just like tweaking your Chevy V-8 for more power. If everything
was perfect no one would do it --and we wouldn't have the fun we are
having-- :D :D

>Perfed metal should work OK for EMI but window screen will not. It kinda
>works but not too well for many reasons.

Okay, I could also use perforated (sp?) metal from an old monitor shield.

>Also I was trying to say do not forget about that hole that sucks air over
>the rear card cage. The boards in there need cooling too.:D

Yes, I did catch that. I'll have to look into this problem more closely.

Greets,
Reinhard

It seems that it would be a lot easier to go ahead and replace the fans with ones that have the speed sensor in them. If you don't want to do that, what about a flow meter of some sort. I know they make flow meters that just have to have a small tip in the air flow to get a reading and don't restrict air flow; but I don't know about cost. They might get pretty expensive.

Just a thought,

Andy

Hi Andy,

>It seems that it would be a lot easier to go ahead and replace the fans
>with ones that have the speed sensor in them.

Exactly my point :D :D

However, the Marquee posses no electronics to check the rotational
speed of the fans (as the original ones are simple two wire & fire units)
So we would need to add this (easy). Since we aren't interested if the
fan is turning 2000 rpm or 2100 rpm it makes the survailance electronics
a lot easier. The mains supply should be shut off, if any (AND gate)
of the fans drops below point "X" (count to N and reset).

There *is* a small problem, however.

I always remove power completely from the projector when not in use.
(after a cool down period). Not all people do this, so we need a method
of overriding the safety feature when in "shut off" mode (as three of the
five fans aren't running then). The saftey circuit still needs to monitor
(IMHO) the two fans in the LVPS when the projector is powered up,
however.

> If you don't want to do that, what about a flow meter of some sort. I

Here again, the problem *where* do you place the flowmeter? Same prob
as when using thermal probes.

> flow to get a reading and don't restrict air flow; but I don't know about
>cost. They might get pretty expensive.

I don't think so. We have these vane switches (we aren't interested if the
airflow if 10 cfm or 11.5 cfm. we just want to know if there is air flowing)
in some of our welding equipment. It's just a microswitch with an air vane
attached to them. It would be pretty tricky to place them correctly (there
isn't that much space around the fans on the blow side)


I think an rpm monitor, which is relatively insensitive to rpm is the way to
go (for the safety circuit). However I would like to improve the air flow
through the projector, too :p


Greets,
Reinhard

A quick question...

I am working on the neck boards and I can't find any 22uF >=100V caps locally so I am wondering if I can use leftover pany 33uF 250V from my ECP days in place of these.

thanks,

Andy

Originally posted by mp20748
Now I'll share a real nugget.


First we'll start on the vertical... There are 8 resistors I'll recommend for change on this board... the later version boards have one half ohm (1/2 ohm) on each supply rail... They're supposed to be fusable metal film resistrors at 1/2 watt :rolleyes:

You'll also want to replace R5 and R6 with the same resistors also.

Now for the convergence board... There are 12 of them on the convergence board... There are also 48 1.2 ohm (same mickey-mouse resistors) right near the power transistors.



So I am going to need a total of 70 fusible metal film 1/2 ohm resistors at 1/2 Watt right?

Sorry to bring this up again so late in the game.
Forget the previous post I'm putting together the order.

Andy

Ya know Andy, if you were nice, and I know you are, you'd post the Digikey part number when you get it so I can order me some too. :) :)

Cary

Are the caps low ESR and 105 degree?
if so they probably would work. But I would think that they are too big to fit.
But digikey has the right parts .


On the fans the only problem with RPM is if the thing get plugged with a dust the rpm will go up a bit but no air so it can over heat and the tach will tell you nothing.

On mine the blower reversed the air in the rear card-cage.
And runs the same direction for the main part.
But it has no affect on the LVPS/HVPS colling system .
They are two separate systems that do not interact .( in the system due to the walls sealing each system off)
But what was happing with mine (ceiling mount) is that the hot air leaving one vent was just sucked in the next vent so the air was just running around in a circle getting hotter and hotter. So by blowing in fresh /filtered air in the system, the exhaust is cooler then the air would get going in the circle so the LVPS/HVPS should run cooler now due to just having cooler air there to suck in.

Cary, it looks like Digikey does not have the resistors but I will list the rest when I finish. I think I am going to have to order 1000 (the mim. order)from Micro-Ohm and sell off the extras for those resistors.

And do I really need 22 1/2ohm 1/2 watt and 48 1.2ohm 1/2 watt resistors or is it 70 of the 1/2 ohm 1/2 watt.

For some reason that 1.2 ohm looked like a typo to me....

Andy

Andy, the schematics specifically show 2 1 ohm 1/2 watt fusible resistors in parallel. They didn't just change the value to 1/2 ohm.

Note that the paralleling is not in all places where there are 1 ohm resistors, there are some circuits where only one leg has the dual piggy-backs. If you don't see this, let me know and I'll tell you which ones.

If you look carefully at the PC board, you'll find that the 1.2 ohm resistors are also in parallel pairs side by side on the board. These are not fusible, just standard metal film.

33mf 160v 105 degree caps on the neck boards are fine, and in fact recommended on upgraded boards.

Last time I called Micro-Ohm (about two weeks ago) they were quoting 4+ weeks lead time on those resistors. Florida Circuit, Inc., also carries them, but I have not called for stock information. There seem to be about 4 European manufacturers of fusible resistors, but I've only found 2 local US distributors. So far.

Hope this helps.

--Bill

Andy,
you'll need either 1/2 ohm or the 1 ohm from Micro-ohm.

The manufacturers upgrade was to add a second resistor to the 1 ohm resistors that are on some of the earlier boards. The later versions has the added resistors.

Also from the same tech bulletin (reliability Upgrade). The following parts should be replaced as follows:

- Remove R704, R804 and R904 (36Kohm) and replace each with a 30Kohm ohm resistor.

- R715, R815 and R915 (1.5 ohm) and replace with a 1 ohm resistor.

- Remove C703, C803 and C903 (470nf) and replace each with a 1 mfd 50dc electrolytic.

- Solder a 1 ohm resistor in parallel with R721 (i.e. across R721). repeat for R821 and R921. These parts are to remain elevated since they serve as fusable safety resistors for the module.

-------

Concerning the "color correction module" - As I stated before it should be removed. The exception is if you're using a curved screen, then it'll allow you to tweak the contrast on the outer edes for a more even image. Or if your're doing a multiple (side-by-side) projector setup, and you'll want to match the edges.

I'll try to get back to this thread later and finish up with upgrades and such.

I'll try to get back to this thread later and finish up with upgrades and such.Mike, that would be wonderful!Concerning the "color correction module" - As I stated before it should be removed. The exception is if you're using a curved screen, then it'll allow you to tweak the contrast on the outer edges for a more even image. Or if your're doing a multiple (side-by-side) projector setup, and you'll want to match the edges.You know, I've never been able to get behind this particular suggestion and would like to know what your objections to the use of the contrast modulation card are.

IMO, the bright spot in the center of a flat screen is very obnoxious (to me)and obvious on many scenes. I have oscillated back and forth a few times, but what I keep going back to for the most realistic left to right image would require that card.

What I do is after the full gray scale calibration (using Colorfacts), I then measure the color temperature at the left and right sides of the screen, and utilize lens striping of red and blue to bring that temperature as close to that of the center as possible. Then remeasure temperature from each edge to about 1/4 of the distance in from each edge of the screen, striping on the green lens to smooth temperature deviations. The result is +- 100 degrees color accuracy from left to right. Without doing this, the right side of the screen is red dominant (several hundred degrees) and the left side blue dominant on a ceiling mounted pj.

But lens striping does take some light away so there is a little more lightness falloff to the sides. I correct that partially by increasing the light output on the three left zones and the three right zones equally on all colors. I say partially increase it because if you do it too much it doesn't look as smooth. Zone numbers of 65 to 70 are usually sufficient.

The net result is a lot less of a hot spot in the middle, smooth color rendition from left to right, and you don't need to run contrast as high to present a solid image. The picture is just more believable and 'there'. It's a real PITA to do, but so are many things about CRT projection.

Is anyone else doing anything like this or have comments pro or con?

--Bill

Originally posted by mp20748


- Remove C703, C803 and C903 (470nf) and replace each with a 1 mfd 50dc electrolytic.



The values on my caps are different, just wanted to make sure I still replace the same ones.

Cap 803 is a 330 (nf?)
http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=803.jpg&.src=ph&.view=t&.hires=t (http://www.avsforum.com/avs-vb/images/avsforum-logo.jpg)

Cap 801 is a 470 (nf?)
http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=801.jpg&.src=ph&.view=t&.hires=t (http://www.avsforum.com/avs-vb/images/avsforum-logo.jpg)


Do I still replace cap 803 with the 1uf cap even though it does not start with the same value?

thanks

Andy

Reinhard

Sorry. Have had limited time on the board the last month and just caught up on this thread.

I am surprised that no one has said to take the tubes out of the 8110 and put them in the Ultra. The 8500 has a lot more features than the 8110.

The Ultra should already have the CLC449AJE chips and the blue gamma circuit.

Walt

Also on those 48 resistors, I have two types of "pairs" of resistors along the power transistors. There are 12 pairs like the two on the left of the picture and 6 like the two on the right. There are no "piggybacked" resistors at all so I am assuming I want to replace the 24 similar to the ones on the left with the 48 1.2ohm metal film resistors right?

http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=48+resistors.jpg&.view=t

thanks,

Andy

PS. The board I have is 50-2023-03P Iss. 2

PPS.
According to Micro-ohm's resistor color codes these are 10.0 and 10.2ohm resistors.
http://www.micro-ohm.com/colorcode/rescolor.html

Andy, no that's not correct. 1.2 ohm resistors are brn-red-gld-gld, and from left to right (lettering right side up) at the bottom of the board, are the following position numbers:

R169,R171 R168,R170
R269,R271 R268,R270
R369,R371 R368,R370
R469,R471 R468,R470
R569,R571 R568,R570

The 1 ohm fusibles which should be replaced with pairs of 1 ohm fusibles are brn-blk-gld-gld in the middle of the board. From left to right (and up/down):

R118,R117
R218,R217
R318,R317
R418,R417
R518,R517
R618,R617

On the vertical board, there are no 1.2's, only 1's to deal with. Pairs of 1 ohm should replace R721, R821, R921.
Single 1 ohms should replace R720, R820, R920.

--Bill

Hi guys,

@sly:

>On the fans the only problem with RPM is if the thing get plugged with a
>dust the rpm will go up a bit but no air so it can over heat and the tach
>will tell you nothing

Ahk!, Darn it. Your most certainly right. I didn't think of that situation. Hmm,
seems like I'll have to do some more thinking. Thanks for pointing this out !

>But it has no affect on the LVPS/HVPS colling system .
>They are two separate systems that do not interact .( in the system due
> to the walls sealing each system off

Yes, true. I was @work and I tried to recall from memory how the Marquee
is built. Seems like I mixed up the rear PSU panel and the divider wall
underneath the fans. :D

>the LVPS/HVPS should run cooler now due to just having cooler air there
>to suck in.

True. My idea was to make an air channel between the LVPS & the HVPS.
Seems like some of the air from the LVPS won't reach the HVPS. because
the front sheet metal covering both PSUs has vent holes (the LVPS
however sucks air in from the right side (and not the front).


@mp20748 & bblue:

Hmm, now what :D ?? One of you says ho, the other hum to the color
correction board. @Mike: Does the signal quality degrade (visibily or
measurably ) when running the projector with the color correction board?
To me it seems like a good idea, if I could adjust the contrast differently
across the screen.

@aspec2:

>Sorry. Have had limited time on the board the last month and just caught
>up on this thread.

No problem :D I was just asking some questions that popped into my mind.
Glad I am getting lots of resonance and setting me straight on some of my
thoughts.

>I am surprised that no one has said to take the tubes out of the 8110
>and put them in the Ultra. The 8500 has a lot more features than the
> 8110.

That was my idea. The 8110 is my projector to play with at the moment.
The 8500 Ultra will be the true McCoy (i.e. also the one to mod). However,
I was exactly planning on doing what you said, since the tubes on my 8110
are in almost mint condition. I'll keep the 8500 tubes as spares and will be
looking for another green tube (to keep as a spare), since the green tube
out of the 8500 isn't in that great a shape. However, since the 8110 tubes
are okay, I'll have enough time to look for the (spare) green tube.

However, what I do need to know is if I should keep the stigmator stuff
on the modded 8500 (when moving the tubes). It seems to be a good idea
to be able to tweak the whole screen with the stigmator adjustments.

>The Ultra should already have the CLC449AJE chips and the blue gamma
>circuit.

Yes, it does. Donald mentioned this and I have since had a look at the
board. Walt, do you know where are the tube hours stored? Does the
EEPROM on the backplane board hold this info? If not, where is it stored
and what info does the EEPROM on the backplane PCB contain? :D

Greets,
Reinhard

Originally posted by aspec2


I am surprised that no one has said to take the tubes out of the 8110 and put them in the Ultra. The 8500 has a lot more features than the 8110.

The Ultra should already have the CLC449AJE chips and the blue gamma circuit.

Walt

The tubes from the 8110 (non-LC) 'WILL NOT' fit in an "ULTRA" The Ultra has LC (liquid coupled) tube assemblies.

-------

Bill,
the 1.5 ohm (R715, R815 and R915) where in some of the early version vertical boards. Periode: June 1995 to September 1995.

Hi Mike,

>The tubes from the 8110 (non-LC) 'WILL NOT' fit in an "ULTRA" The Ultra
>has LC (liquid coupled) tube assemblies.

Not to quibble with an expert, but are you sure? The Ultra Series comes in
four flavors: The 8500LC 9500LC, the 8500 (non liquid coupled) and the
8520 (two piece projector) At least this is what my manual, VDC Display
Systems and the label on my projector say ;o))

Greets,
Reinhard

Reinhard,
you've got me on that one. I've never seen a non LC Ultra, nor does any of my literature show one. It must have come along later.

That's one over me!

Hi Mike,

>Reinhard,
>you've got me on that one. I've never seen a non LC Ultra, nor does any
>of my literature show one.

Here is a link to a sales PDF from VDC:

http://www.vdcdisplaysystems.com/Products/Simulations/VDC_Marquee_Preliminary_Brochure_2002.pdf

>It must have come along later.

Hmm, It's dated December 2000. It's also no longer the old style case (i.e
beige & grey with perforated (round holes) sheet metal along the bottom
of the projector case. It's black and has plastic sides covering the bottom
of the projector. This plastic contains slits instead of holes for ventilation.

Getting hold of the contol board:
U35 is version 4.2C
U29 is Rev 2.0 (no longer socketed)
U28 is Rev 3.0 (also no longer socketed)
U16 is Rev 3.1
U106 is Rev 3.1
and U100 is Rev 3.4.

The Electrohome part number for the complete control board is 50-2036-03
ISS.


Greets,
Reinhard

Originally posted by bblue
Andy, no that's not correct. 1.2 ohm resistors are brn-red-gld-gld, and from left to right (lettering right side up) at the bottom of the board, are the following position numbers:

R169,R171 R168,R170
R269,R271 R268,R270
R369,R371 R368,R370
R469,R471 R468,R470
R569,R571 R568,R570


--Bill

Ok I'm lost... I have had 3 other guys look at this thing and none of us can find any of these resistors.

They are supposed to be on the Convergence Section of the Convergence Vertical Module (50-2023-03P ISS. 2) right? I can't find them on the Vertical Section either. Can someone point out where the should be in the picture?


thanks,

Andy

http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=Convergance-+Vertical+Board.jpg&.src=ph&.view=t&.hires=t

Newark has 1ohm .5watt fusible metal film resistors for $.145 each

http://www.newark.com/NewarkWebCommerce/newark/en_US/support/search/parametricSearch.jsp?category=710000000&action=2&First=0&parameterLabel=Resistance&parameterValue=1+Ohm&searchString=2032%7E2044%7E&level=1

Andy

Andy, that's an older revision of the board than I have schematics on. But from the picture it looks like the piggybacked 1 ohm fusibles should go at positions R117,R118 R217,R218 R317,R318 R417,R418 R517,R518 on your board. They should be maintained approximately at that height as well.

Your board doesn't appear to have the 1.2s at all. In their place is what looks like standard 1 ohm 1/2 watt metal films. Check and see if the Rx01 and Rx02 (x=1 through 5) positions are in parallel with each other. I think they are, and if so they should both be replaced with the same value. The other leg appears to be Rx10 but it is not arranged in parallel pairs, so just the one should be replaced with the same value.

There might be a technical bulletin with value upgrades, but I've not seen one.

On the vertical section there shouldn't be any 1.2's. Perhaps you can include a picture of it.

--Bill

Thanks Bill, There do not appear to be any 1.2's on the vertical, there are a few 1.0's in the upper right hand corner, R4, R5, R6, R7 though.

Here's the pic.

http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=Convergance-+Vertical+Section.jpg&.src=ph&.view=t&.hires=t

Mike, do you know anything about this board?

thanks,

Andy

Andy,
what is the model of that projector, and what are the numbers on that board (50-XXXX-XX)?

Marquee 8000, the board number is 50-2023-03P ISS. 2

Andy, it looks like R712,R713,R714 R812,R813,R814 and R912,R913,R914. You'll probably find in each triplet, that two of the three are in parallel with each other and one is by itself. If this is so, then replace them 1 for 1, no piggybacks..

Mike, it's a combo board part number 50-2023-03P ISS. 2.

--Bill

Bill, for R712,R713,R714; R812,R813,R814 and R912,R913,R914, each group has all three in parallel

"Check and see if the Rx01 and Rx02 (x=1 through 5) positions are in parallel with each other."

Rx01 and Rx02, on the convergence section, are not in parallel

Hi Andy

Excellent macro photography. What camera/lens are you using?

DH

Thanks Donald, I am using 2 different ones, one is an Olympus, I'll get back to you on which one, that I use for the whole board shots and for the closeups, I am using a digital hooked up to a microscope and a computer (back-lit ones are just with a light underneath).

Here are a couple pictures of the traces, front and back of the two groups.

R912,913,914
http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=Back-R912-102.jpg&.src=ph&.view=t&.hires=t


R101, R102
http://f1.pg.photos.yahoo.com/ph/andyebr@sbcglobal.net/vwp?.dir=/Marquee+8000&.dnm=Back-R101-102.jpg&.src=ph&.view=t&.hires=t

Andy

Andy,
I'm not sure, but I think you have the very first version of the 'separate' vertical board that went into the some of the 8000's.

The very first 8000's have a single (combined) vertical/convergence board that was not detachable, and it was used on the rear heatsink that did not swing down once the four screws were removed.

They later went with the separate versions (detachable convergence/vertical) that are so common today. So that's why Bill has been getting confused with the resistor locations, and it appears that both the vert. and convergence are slightly different from what we've been discussing. The good news is, it's not the single combined board. the not so good news is, it has us a little confused :confused:

Anyway, have you checked your neck boards for the CLC409 chips?

mike

Yes, they have the H1100 chips, I have 409s to replace them with but have not gotten to that yet.

Andy

Andy

Um.. I hope you meant to say "I have 449s" in your previous post.

DH

yep, I did...

I have CLC449's to put in place of the H1100's

Thanks Donald

Andy, I hope you meant 449s to replace them with, not 409s.

Well since those triplets are all in parallel, I'm not sure what to suggest. The circuits are usually fed with two rails, a + and a -, and each will have some low resistance value as a feed through. On your board, though, it looks like it is not a split rail configuration and one side is ref to ground and the other through the triplets.

Without a diagram about the only way to figure out what might be feeding the other rail is to trace the circuit out as best you can to determine how its supplied.

If it really is single ended and the supply is through the 1 ohm triples, I'm not so sure I'd even bother replacing them. They'd have to be pretty poor quality to have much of an impact as a group. Perhaps someone else reading will know this answer. I'm at an impass for now. Sorry.

--Bill

Ok guys, I just got back from Don Howard's, neither of our have the 48 resistors RX68, RX69, RX70, RX71 (X=1-5) but they do have the 12 resistors-RX17 and RX18 (X=1-6) so I will just replace those with two 1ohm resistors and forget about the others.

Thanks for all the help and sorry about all the confusion.

Andy

Originally posted by mp20748
[B]Andy,
you'll need either 1/2 ohm or the 1 ohm from Micro-ohm.

Also from the same tech bulletin (reliability Upgrade). The following parts should be replaced as follows:

- Remove R704, R804 and R904 (36Kohm) and replace each with a 30Kohm ohm resistor.



Mike, these are the same resistors Tim says to replace the 56K resistors for the anamorphic mod. Does this only apply if you are not doing the anamorphic mod?

Andy

Originally posted by andyebr
Mike, these are the same resistors Tim says to replace the 56K resistors for the anamorphic mod. Does this only apply if you are not doing the anamorphic mod?

Andy

My bad. I quoted from the tech bulletin, and had forgot that those are the same resistors we replace with 56K for the anamorphic mod.

But yes, only if your're NOT DOING the anamorhic mod, and only if your board has the 36K resistors in it. But 98% of the vertical boards has the 30K already in it, and these should be replaced with the 56K, for HT applications.

I've moved these topics over to this thread since my next question is more of a maintenance type.

Thanks to everyone who replied in the previous threads about my tube phosphor concerns and the lens issues. I have located a set of GT17 lenses which should be here in a few days. Meanwhile I have the tubes removed and on the way to VDC for rebuilding again. Though they were better than what was in there a year ago, I never was happy with how they set up and performed, especially blue and to a lesser degree green.

What really surprised me was today when I removed them, all three had mild burn-in after just 1157 hours of operation. Nothing severe but certainly visible. I wasn't expecting any at all and have never considered the picture particularly bright. On setups my 100 IRE reference is about 7.5 ft.lamberts with a 1.3 gain screen. Anyone have any ideas on this? Is it normal?

The real maintenance question is this. Now that I have the chassis completely open and exposed again (it's still mounted on the ceiling), is there anything at all I should consider changing or upgrading? It looks like I can get to the 3-fan plate easily now, would it be worth changing those out to the Papst fans? Anything else?

--Bill

Say Bill

The Papst fan upgrade could buy you seriously better reliability, and 1-2 db lower noise, if done in conjunction with our Papst LVPS upgrade; e.mail me if interested. I also advise replacing the CLM lithium backup battery cell, and Mike Parker's VIM mod also deserves major consideration. HD-10L lenses are excellent out to 120"+ image width, be aware that GT-17s do not focus inside of 97-98" throw. Otherwise, if your 9" Marquee has v3.3 main software or higher then you are pretty well set. Be aware that replacing the lithium cell, or main software, will dump all established settings, unless downloaded to a PC using Marquee Librarian software, then uploaded after. Librarian software is a free download at http://www.etechvideo.com/

Bill,
as you can also see from Tim's post, the HD-10L lenses should have worked for you. Like I have said before, most of the 9500LC's out there have HD-10L lenses. So if what you've been told is true, all of these happy 9500LC owners are operating their projectors on 78" wide screens or less.

-------

Next week, I'll post an announcment on where and how to get the VIM/Neck board mod. I chose 9 people as testers. The jury is still out on three of the testers, because two of them have not got their modded boards to evaluate (they're waiting to ship), and one has his, but is waiting for a tech to install it. I was going to wait for the full jury to return, but I have decided to go ahead based on the results from the six who has already reported. Everyone who has evaluated or has seen the modded boards gave a very good report as to the results. My testers/evaluators included some average HT folk, as well as a few engineers and some other technical folk. It was a well balanced jury.

Now, this mod is considered to be a High End upgrade to the Marquee projector. It will not fix a broken projector, or make a projector with weak or bad tubes look better (though such a projector may show some improvement). It will seriously upgrade performance. The improvements, if used on a well aligned, properly funtioning projector that has a good source and cables should be very noticeable.

Full details will be posted on my site later.

Tim, thanks for the suggestions. I may do that on the LVPS though I do already have new bearings on the existing two fans and they are isolated from the case. The LVPS already had a little voltage regulator circuit for the heaters (much to my surprise) that holds the filaments at 6.31 volts.

I also have Mike's VIM and Neckboard upgrades installed (as a tester), all of the maintenance thread mods plus a few, and have 3.3 firmware. I'll check on the battery age.

On the lenses, I'm not sure what to say. Your assertion that the HD10L's are good for up to 120" wide screens flies in the face of every chart I've seen so far. Even 3M's web page (took over US Precision) calls the HD10L a low magnification version of the HD10E. The L range according to their chart is 60 to 90" diagonal, and for the E is 90 to 300" diagonal, similar to a plain HD10.

I've got 120" throw to work with so the GT17 should do fine. Plus, at 12 lines/mm vs 10 lines/mm it is a higher resolution lens to start with. I just don't see the downside. In any event I'll report what I find either way. They're in pretty high demand, so it wouldn't be hard to sell if necessary.

--Bill

Originally posted by mp20748
And if your board has "U22" you have the gamma circuit.


Well I have one question about this... The picture earlier of U22 was an 8 pin chip I have a U22 but it is a 14 pin chip. Does this make any difference? Here's the pic.

Here's a zoomed out picture

I have two 9500LC. The one up now has the G17 lenses, the other the HD10L. For the 1.78 110 inch diagonal screen (54 x 96), the throw with the G17 is about 130 inches, we have it closer than the book 147 inches in order to fully utilize the tubes' rasters. Both sets of lenses will cover this size screen no problem. The throw changes with the 10s by about 4 inches longer. BBLUE what size screen do you have?

Andy,
that VIm is an early version 8000 VIM. It does not have the gamma circuit.

Mark Haflich had the HD-10L lenses on his 9500LC, until he purchased a second 9500LC that had the GT-17 lenses. We swapped the lenses from the two projectors so that he would have the GT-17's on his present Marquee.

We had good focus with the HD-10L lenses on the 96" wide screen before the lens swap, and we noticed no difference at all with the GT-17 lenses installed.

Here's a zoomed out pictureYou do not have the gamma chip. It would be a CLC4x9 (or at least the same as the R G and B output chips if it was there.

--Bill

BBLUE what size screen do you have?Mark, mine looks like the same as yours, 96x54 or 110" diagonal.

According to the Madrigal MP-9 documentation, they used only the GT17 lens, and for a 96" wide 1.78 screen, the recommended throw is 120.2".

They write "These throw distances are minimums, assuming that you wish to use the maximum available phosphor area. Used correctly, they will deliver the best possible picture. They will, however, require careful attention to proper setup, as there is little room for error at these distances. We recommend leaving yourself +- 1/2" adjustment leeway in the mounting system for final fine tuning of placement."

But their definition of throw distance is "measured from the projector's center lense to the center of the screen". Using that definition I have a little over 124" throw, so it should be fine. I hope.

--Bill

Bill,
the following is copied from the manufacturers tech sheet (54-017088-01P):

"5) Calculated values are for reference only. It is good practice to simulate the setup to determine the necessary throw ditance."

"6) Display size is affected by input signal characteristics. Once the sprojector is set, use the size function to match the size of the projected image with the size of the screen."

This is the proper way to determine throw distance on a Marquee. Every piece of literature that you'll see will have a different number to represent the proper throw distance. Also every rebadger claims to know the proper throw. But from our experience with the projector, and the direct results from the manufacturer itself – there is no “throw distance” figure for the Marquee.

That’s why the literature reads “simulate the setup to determine the necessary setup”. And that is the best way to determine where the projector should go to the screen. From my experience, there are two reasons to do this. One is the differences in the horizontal deflection modules and other circuitry (to include input sources timings), the other is that there may have been slight differences in the HD-10L lenses manufactured. This has been a long debate, but has yet to be confirmed.

One thing for sure, to eliminate that focal problem on the other edge as much as possible, one must follow the recommendation in the manufactures tech sheet. And that even applies to the GT-17 lenses. Even that lens can create an outer focal problem if not setup properly. Especially if you’re mating them to an earlier Marquee, because the Ultra has a different horizontal deflection module than the non-Ultra Marquees. So ignore all reference for throw distance, especially if you put GT-17’s on your projector, the MP-9 is different, it has the Ultra deflection module, and that's a different deflection module than what you should have in your 9500LC.

Thanks Mike, that makes a lot of sense.That’s why the literature reads “simulate the setup to determine the necessary setup”. And that is the best way to determine where the projector should go to the screen.Originally that was how I arrived at the current throw distance. I set the tube image width to as wide as I could (within 1/4" of the tube face edge) and then moved the projector to fill the screen. That doesn't take into account where in the range of the lens you are, though mine seem to be close to the middle.Especially if you’re mating them to an earlier Marquee, because the Ultra has a different horizontal deflection module than the non-Ultra Marquees. So ignore all reference for throw distance, especially if you put GT-17’s on your projector, the MP-9 is different, it has the Ultra deflection module, and that's a different deflection module than what you should have in your 9500LC.In particular, what is different? If the HDM can scan full width with reasonable linearity (and presumably tube focus) what else is required? Or, what else matters?

--Bill

Originally posted by bblue


If the HDM can scan full width with reasonable linearity (and presumably tube focus) what else is required? Or, what else matters?

--Bill

Ok, I'll answer this way:

I have a customer in Virginia Beach, VA. He has a 9500LC projecting on a large curved screen. If I remember correctly, his screen size is 109" wide. He also has HD-10 lenses, but there's a slight focal problem in the corners, but should be expected at that width.

His horizontal deflection board went out. The seller of the projector sent him a replacement board. The replacement board did not have enough amplitude, so he sent his original board to me to be repaired, so that he can maintain his present setup.

So his projector had two of the same part number deflection boards in it, but one was different... and that was also true even with new replacement boards. This is something that both Tim and I are hoping to one day identify with the boards.

I do not know if this matters but by moving out the slug you can gain some width. I wounder if that is all you are seeing between the different boards with the same #

Hi Mike,

Are you *sure* it was the exact same part number? From my service
manual the Ultra part number is 00-260303-09P and the non Ultra
part number is 00-260303-07P. The board numbers are 50-002003-08P
and 50-002003-07P respectivly. It's only a one digit change from Ultra
to non-ultra.

In December 1998 the ultra board was revised. Here are the changes:

R1000 was changed from 49.9 Kohm to 47.5Kohm
R63 was changed form 133K to 100K
Jumper at P7 is no longer in use.

P7 is part of the width sensing circuit. In the original design the LM6361
Op-Amp inverting input has a 7.5K resistor to ground. The later revison
removes this 7.5K resistor from ground (i.e. left floating). The other two
changes are part of the Key & Pin Tracking circuit.


Greets,
Reinhard

Originally posted by Lightningman
Hi Mike,

Are you *sure* it was the exact same part number? From my service
manual the Ultra part number is 00-260303-09P and the non Ultra
part number is 00-260303-07P. The board numbers are 50-002003-08P
and 50-002003-07P respectivly. It's only a one digit change from Ultra
to non-ultra.

In December 1998 the ultra board was revised. Here are the changes:

R1000 was changed from 49.9 Kohm to 47.5Kohm
R63 was changed form 133K to 100K
Jumper at P7 is no longer in use.

P7 is part of the width sensing circuit. In the original design the LM6361
Op-Amp inverting input has a 7.5K resistor to ground. The later revison
removes this 7.5K resistor from ground (i.e. left floating). The other two
changes are part of the Key & Pin Tracking circuit.


Greets,
Reinhard

Reinhard,
the board number is the same (00-260303-07P) on the non-Ultras, but there were changes made to that board number that may show up in the part number to reflect that boards changes. Not sure what the changes were, or when they were made, but just like what you've indicated with the Ultra board, revisions were being made to both, but they still have the same *board* numbers.

----

Sly,
not sure if that'll solve the problem, because the tech sheet was refering to new Marquees, so we assumed that the coils were properly adjusted when they left the factory, so no adjustments were made to them unless it was a special display setup, and the width adjustments needed to be more precise. But to properly size the raster using the coils, the +85 volt rail should be checked first, to make sure it has a stable +85 on both high and low bands, this should be done before adjusting the coils. If that voltage is low on either band, the power supply is either weak or there's a problem with the deflection board. In either case sizing out the image could damage the deflection module.

The huricane has robbed us of power, so I'll be scarce for awhile, until they get our power back on.

Hi,
first excuse me for my disturb,
please help me,
I need the specification of 2 resistor and 1 capacitor behind the r223 of the 8500 power supply LVPS because I have 2 of this LVPS and if I misure the voltage vhen the CRT are on I can read 6.60v and when I look for adjust the trimmer this is missing and in the same place there is a little daughter board and there is not the 2 resistor and 1 capacitor.
I post the picture of my LVPS

The first resistor is a '121' ohm 1/4 watt 1%

The cap (C127) is a '104' (or 0.1 uf). I believe it's 50vdc.

The second resistor (R226) is a '536' ohm 1/4 watt 1%

The resistors must remain at 1% tolerance.

Starting this week we’ll be working with Tim Martin on Marquee board repairs. This co-laboring venture will also include my Marquee "Performance Mods."

Tim (E-Tech) will be the source of both board repairs and Performance Mods. We’ve combined our talents, tech literature and notes to make possible a very comprehensive board repair and exchange service for the Marquee projector. This new service was in the workings for the past year, and was the main reason for my visits to the desert (Arizona). So after a year of hard work, we’re now ready to offer this service to all.

This service will cover about 90% of all the Marquee boards. And before the end of next month we plan to also offer rebuilt HVPS’s. And after two years of laboring with the HVPS, we feel the rebuild process of the HVPS is complete. I now have two technicians in place to help with this great work.

We wanted to not only offer repaired boards, but we also wanted to make sure that the boards we offered would have the latest known upgrades to include the replacement of any known or commonly failed components.

Starting today, All Marquee board repairs to include the Marquee Performance Mods that I’ve mentioned in the past will be handled through Tim Martin. Since I’m not the source of contact for these items, I can’t discuss pricing. Please email Tim directly at this email address:

ehometech@earthlink.net

More details on this and my MP-5 Transcoder will be posted on my website soon.

Also this month, The release of the MP-1.3 -- will also be discussed on my website soon.

The Repairs, Mods (Tim) and Transcoder (to be announced) will all be handled by someone other than myself. I need to get my life back :(

I have my 9501LC tubes back from VDC and am in the process of doing the full magnetics alignment and setup. They really look quite excellent so far, though red behaves quite a bit differently than the other two.

After disassembling and cleaning the GT17 lenses I installed them, replacing my HD10L's, to do a rough positioning adjustment along with a preliminary electrical and optical focus. (If you missed the earlier thread, the issue had to do with how large a screen the HD10L's could project to. Mine is 10" wider than the widest supported, yet others reported no problems at all with even larger screens than mine at 96" wide.)

I cannot believe the difference between the GT17's and the HD10L's I have been using on my system. The focus is very precise, and the flapping and corner adjustments are a breeze to do. With the HD10L's, flapping changes were hard to see clearly and difficult to balance uniformly from side to side, corner to corner.

No more. The sides, top-bottom, and corners are very easy to adjust because you can see what's going on quite clearly, 10ft back from the screen! I haven't tweaked them down to the last nad yet, but after just a few minutes spent on each one they are very sharp and detailed all the way to the edge of the screen and corners. Quite impressive, and something I've never seen before here.

Even better, at my existing throw distance (just over 120") there was no significant change in picture width and height. The overall image size is the same.

The only gotcha in the installation was that these particular GT17's (from a Barco projector, not that that should matter) have a slightly thicker mounting lip. The four cap screws used to mount each lens were just long enough to get a couple of turns thread grip. I'll need to replace all of those before I button things up.


Now a little tip. Maybe this is common knowledge but I've never heard anyone mention it before. On a ceiling mounted pj, when you're doing the flapping adjustments, you'll note that the lower left corner of the front of each tube has adjustment screws for up/down tuning, and the lower right corner of each tube has adjustment screws for left/right tuning. For any given directional adjustment, you're supposed to loosen one screw slightly, tighten the other one against it, note the differences and keep repeating this until things look right.

What they don't tell you is that the longer of the two screws in each set is actually the lockdown screw, and the short one actually does all the moving. Assuming that your tube assemblies are free within the case (no wires or other things holding them from moving freely), and your pj is ceiling mounted, you can simply unscrew the lockdown screw a few turns and make the adjustments with the short screw only. By having this freedom, you can see the immediate and continuous results from a single adjustment. Once you're in the right range, you can actually see the lens move in the image on either side of balanced. It kind of widens slightly on one side or the other at +-1/4 turn on the adjustment. Once you're centered in that range, tighten down the lockdown, make sure it didn't shift anything, and you're done.

Of course, you still need to work up/down or left/right against master focus if flapping is off by very much. With the corner adjustment ring, which also interacts some with flapping, if one or more of the corners does not sharply focus at the same time the others do, your flapping adjustment is still off a little.

--Bill

Throughout this and other threads, whenever there has been a discussion on how to set G2 on a Marquee it's described as being done only by reference numbers of control settings. For example, setting brightness at 50 and then adjusting G2's so the lowest step on the internal stair-step pattern is not seen, and the 2nd step is 'just' visible. Or setting all G2's at 60, brightness at 50 and then do gray scale tweaking from there. 60 because that's the way a lot of them come set from the factory.

Doesn't that leave some latitude for a possibly significant error in terms of tube life depending on age and origin of the tube? Contrast and Drive settings do the same thing -- vary the amplitude of the video signal on the first grid of the CRT. But Brightness and G2 do quite different things. G2 is a fixed voltage adjusted only by the G2 grid pot on each tube, and brightness controls the DC signal offset between cathode and grid of the CRT. So I think you could say that Brightness controls traditional black level (relative to the peak video signal itself), but G2 controls tube cutoff.

Isn't there a more precise way to adjust G2 to get it correct for each tube as well as relative to brightness?

It seems like you would want G2 as low as possible and still give you enough upper range of brightness, as that would minimize the electron output of the cathode. Right? Not?

Does anyone know more about this and/or a precise procedure for setting G2?

--Bill

Bill, as you say, drive/contrast/brightness effect the cathode (and G1 grid for the Marquee) and G2 is Grid 2. I hear what you're saying, but even if one is set to the extreme, the other will compensate for it. If brightness is too high, if you adjust G2 as per the instructions, it will end up being slightly lower than it would have otherwise. The yardstick we're using to measure their "proper" position is the beam current (indirectly, as we're really looking at how much the phosphor is being excited). Since we adjust so that the "lowest step is not seen", I don't think it matters how we got there, brightness or G2; in terms of tube life, it will be the same because the beam current is the same.

Now what about the effect of having one of the controls set to an extreme? Let's say brightness is too high so G2 gets lowered to compensate. I don't really think it will effect the G2 circuit at all. Moving the brightness to full range probably requires only a small voltage change on G2 to compensate.

Cary

hi cary,

don´t you loose some contrast/ punch when setting g2 too low and compenate with increased brightness?
this is what i examined on my barco. with proper g2 adjustment the overall picture has very much more punch. with lowered g2 it´s a little washed out.
don´t know yet how this works with marquees.

regards,

gernot

Gernot,

As I typed my reply above, I wondered about that. Increased brightness *could* effect the dynamic range available for video since you're raising the floor on the signal. I decided that I'd really have to look at the schematics to tell and I was too lazy. :)

Cary

For the last few days I've been running G2's high (70-78) and brightness low (23). The picture is pretty good, though the tubes are still settling in. I'm about to do another tracking adjustment and this time will shift G2's down and brightness up and see how it compares.

Looking over a variety of reference literature on vacuum tubes and crt's, it seems to be indicating that in a projection type crt the 2nd grid is used more as an accelerator than a control grid, just before the focus and other magnetics come in to play. If this is true and it is more-or-less amplifying further whatever beam it gets as a result of cathode and grid modulation, then it would seem that you'd want the cathode to emit less (lower brightness) and G2 to pep things up. But where is the happy medium? There must be a preferred current through the cathode and a preferred voltage at G2 for optimum performance. Hard to imagine that these two different areas of beam control can be completely linear through both their operating ranges.

--Bill

Hmmm, well, at least we only have to worry about the cathode. The G2 voltage is constant (except for when you adjust it) so it's only the cathode that varies the beam current. If I remember right, brightness does change the bias point for the cathode amplifier (it's really a differential amplifier with outputs on the cathode and G1) so I guess it's possible that whatever bias point you end up with because of how you set your G2 could effect the linearity or dynamic range of the cathiode/G1 amplifier. I'm not sure that the bias point would actually shift enough to worry about though.

I guess you're in a good position right now to see if you can tell a difference Bill when you go to higher brightness and lower G2.

Cary

Well. I spent most of the night this morning trying different things and was getting inconsistent results. So I noted the G2/Drive/Brightness/Contrast settings I used when Brightness was being set really low ( as above) and started from scratch using the recommended starting point numbers (all G2's at 60 and drives at particular values for each tube) like in earlier threads here and also on Tim's web site.

I've started with those numbers before and they were reasonably close but this time (new rebuilt tubes, different blue phosphor, add red c-element) those numbers weren't anywhere close. When I was happy with the tracking according to Colorfacts I was surprised to see that I was +-500 degrees of 6500 from 20 to 100 IRE, and most of that variance was just in the blue tube curve. Not bad, actually.

But I've now watched the pj in this config for several hours and don't like it a bit. G2's are each 4 units below what they were on my previous setup but drives varied quite a bit both directions. For what approaches a balanced picture now, I have contrast at 55 and brightness at 60. Before I was around contrast 58 and brightness at 23. That brightness shift is from 4 G2 units on each tube! Anyway, at any combination of brightness/contrast the picture looks too dark, except for bright peaks above 60 IRE or so, and they just jump out at you. Turning down contrast and raising brightness reduces the range but doesn't undarken the picture detail-wise.

Before doing any of this I watched my various sources (several HD and the HTPC) on the scope making sure that video waveforms stayed within the 0-7v p-p range, and that the HTPC output also conformed to that. All the sources now have the same general appearance, dark, though HDNET and Discover HD when airing excellent source look very good except for the darkness at lower levels.

Tonight I'm going to try to recal with higher G2, forcing lower brightness and see how that changes things.

--Bill

Bill

I haven't read this thread for a while. You, however, have piqued my interest.

Thanks

Walt

As some of you know, I am very new to the 9500LC. Just got my first Wednesday.

This thread has been great, I commend all of you for your contributions.:)

I got to one of my neck boards and it definitely shows signs of the high hour burn that MP describes. (I will attempt to attach a picture.)

What is the recommend part to replace L8 and L9?

I have been planning to buy a digital Multi-meter, do some of you have recommendations? I have been looking at Fluke, but there are so many choices..... (I could just get a 189, but would never use all of its features.)

Thanks

I will try the picture attachment again. :confused:

Wow, and I thought mine was bad...

Glen,
Probably not too much conductivity left there! Probably should remove the resistor and wire brush the ends, reinstall and resolder making sure there is good flow between the trace and wire. You should resolder all those resistors regardless of how they look (even good ones). I've got one board in a spare set of neck boards that is a lot worse than that one, to the point of having to replace sections of trace with wire. It was really fried.

On my spare set of neck boards I used the Miller 5300-25 100mh chokes (Mouser) and they seem to do ok. On my main boards with Mike's modifications, he uses something different, though it looks similar. I don't know what it is by number. Perhaps he will comment on this.

Definitely do all the neck board mods you find in this thread. I'm not sure which thread it's in but later on Mike mentioned using 33mf 160v 105 degree caps instead of the 22mf's referenced in the thread.

--Bill

Following up on my last G2 post, there does seem to be something to the relationship of G2 to brightness.

In my last post I had started calibration to what appears to be more or less reference G2 levels, and ended up with remarkably good gray scale tracking from 20 IRE up but the resulting picture had a too high on whites too low on blacks look that was really hard to watch.

This time, I started calibration from that point (where it was), but first lowered brightness to 30. The net result was, basically, moving G2 up 3-5 numbers on all colors. The look of the resulting picture is completely different! The high whites look more like part of the picture, and the low blacks show more shades of gray. The whole presentation is more realistic and life-like and subtle details are clearer, as if it was higher resolution, but the resolution didn't really change. It's sort of analogous to an audio system with its treble and bass controls set too high, then lowering them to flat or a point at which they become part of the whole presentation rather than jumping out at you.

There probably is a point somewhere between these two extremes that is the ideal place to be, but how do you determine it other than visual trial and error as I have been doing? I can't help but believe there is a specific 'right' way to set (or start) G2 (or brightness) that just hasn't been disclosed as of yet.

I can't quite visualize what is happening in the tube with these changes. At first I thought it was a range issue -- the amount of visual range there is from black to white, but that is what white level (contrast) would control. Black level (brightness) itself seems to be a DC biasing issue of the tube which would control where in the operational curve of the tube the black to white range actually exist. G2 would then act like a booster amp, accelerating the tube output results on to the surface, with further acceleration and beam guidance handled by magnetics and anode voltage. This currently makes the most sense but I'm just not sure yet. Anyone with another explanation?

--Bill

Bill, That is the plan, and I will make all three the same. Mike told me yesterday about the 33mf I just need the specs for L8/L9 the CLC449s have been ordered. I think that is all i need to replace.

Now for my "dumb" question: Is all of this neck board work done with the heat-sink in place?

Yep, sure is. The heatsink only is partially in the way on a couple of components. With a slim shank soldering iron and standard pc board tip you can get to everything with little effort, just a bit of cleverness!

--Bill

anyone? No thoughts or opinions?

--Bill

Bill,

I recently did some G2 adjustments on my Barco 1208. As a result, I have noticed that I am having to back down the brightness on some signals.

Where can I find some of the CLC449's??? I need to order a dozen or so...

Thanks,

Nick

Hi Bill,

Originally posted by bblue
anyone? No thoughts or opinions?

--Bill

Well not directly, but what would be of interest to me is what
beam current values are safe for long lasting tubes? I think
beam current is the only real value which can be considered
in terms of talking about tube life. A lot of people say a contrast
& brightness setting of so and so is good for long tube life.

Personally I don't think you can give any "potentiometer" values
like that, since it also depends on how "hot" your input is. Meaning
what is the Vpp of the video signal, etc.

Of course, I realize any beam current above zero is going to lead
to phosphor wear, but where is the happy midpoint between
brightness and tube life? And how would one go about measuring
beam current? One might be able to measure it via a cathode
resistor (being the sum of all current flowing in the tube i.e. not
explicitaly the pure beam current), but I doubt that G2 or other
control grids take that much current, so it should be accurate
enough. And how does beam current comapre to tube life? Is
it a linear function? exponential? If it's exponential (which I
suspect --> at least above a certain limit), then tweaking a little
more brightness is going to reduce tube life radically.

Bill wrote:
> but the resulting picture had a too high on whites too low on
> blacks look that was really hard to watch

Yes, I too have noticed this when setting up the tubes the way Tim
describes on his website. Using the other (described elsewhere)
method (getting all tubes to light up from dark black equally), also
doesn't give me results which I really like. Having a colorimeter (sp?)
is nice, but I don't have one *and* I think you can't use a purely
scientific method to adjust color. Everyone likes their "tone" a bit
differently. Some a little colder, some a little warmer.

I also find the method, which Tim (no offense ;o)) describes a little
vague. What *exactly* does it mean, when he writes "so that the
second darkest square just lights up". Should you be able to actually
"see" the bar or just the divider line between the bars (for example)?

I mean the smallest amount of ambient light or light being reflected
from the walls, furniture and what not else are going to affect this
adjustment. It also depends on how good one can identify the
difference between really black and "just lit". Other questions which
pop into mind: Is this to be checked with your nose directly on the
screen or for which screen gain values is the true? The higher the
screen gain, the easier it would be to see, but then the easier the
screen will hot spot.

However, I am not sure if you can work with voltage values for G2, as
this also depends on the degree of wear on each tube. As a matter
of fact I'm not even sure each tube (say green for example) has the
same exact same color range as another tube even of the same kind.
In other words does each and every tube (of the same color) have
a true spectral range from "x" to "y" (low end to high end range).
Does this change with age (w/o taking wear into consideration)?
How much does this vary from tube to tube? How much does this
vary in terms of tube wear? All of these things are going to affect
the color temperature output of each tube and therefore also the
color temperature of the whole picture.

A lot of questions, but I haven't been able to find any really
satisfactory answers ;0))

Greets,
Reinhard

Originally posted by SunTECH
Where can I find some of the CLC449's??? I need to order a dozen or so...

Thanks,

Nick



I could use a a dozen or so also. I can't seem to find them unless you order 1000 or so.

I am getting ready to replace the tubes on my 9500LC. Any recommendations on cleaning the black electrostatic dust from the hardware and Thomson yokes?

I also find the method, which Tim (no offense ;o)) describes a little
vague. What *exactly* does it mean, when he writes "so that the
second darkest square just lights up". Should you be able to actually
"see" the bar or just the divider line between the bars (for example)?Lightningman, you should be able to just see the 2nd bar and not see the first bar at all. I'm not sure what you mean about a line in between them -- I don't see that. Maybe your contrast is too high?

On G2, I'm sure there must be an ideal or 'target' cathode current range for the tubes but so far no one has reported any knowledge of it.

I wonder if one of our resident techs who has a full service manual set could look in the waveforms/voltages or calibration areas and see what it says about the subject?

I've been watching HD video on the scope at the same time as on the screen and noticed that the majority of it is referenced to 0 IRE not 7.5 IRE as 'black'. But many will only drop to 7.5 IRE on a fade to black and then resume 0 IRE on presence of signal. This means that if you have brightness set for the right black level as seen on program content, fadeouts will fade to gray, not black. Upconverted HD (from regular NTSC) mostly uses 7.5 IRE as 'black' but that doesn't mean much since you don't usually adjust to (or even watch) it.

This doesn't have anything to do (directly) with our G2/brightness tube calibration dilemma but does explain the need to push brightness for general HD viewing.

--Bill

I am getting ready to replace the tubes on my 9500LC. Any recommendations on cleaning the black electrostatic dust from the hardware and Thomson yokes?If inside the house a vacuum cleaner with wand+brush, and perhaps a small 1" dry paint brush. If outside, an air compressor and dry paint brush. If using an air compressor (at any time) be sure to drain the tank of any accumulated moisture beforehand.

--Bill

Hi Bill,

Okay, after the stupid Windoz explorer screwed me :mad:,
here's my second try.

Originally posted by bblue
Lightningman, you should be able to just see the 2nd bar and not see
the first bar at all. I'm not sure what you mean about a line in between
them -- I don't see that. Maybe your contrast is too high?


Maybe I just described it wrong. Each graytone of the stairstep has
a distinct edge, making it easy to see each rectangle. That is what I
called "line". This distinct edge is very easy to make out way before you
see the actual graytone of the specific bar. So my question was do you
just barely want to see this edge or do you actually want to barely see
the rectangle (second to last or second to first). While calibrating I have
contrast set to 45 and brightness to 50. While viewing a movie actual
contrast is set between 41-46 and brightness between 31 and 38. Sort
of depending on the movie. My screen is almost 2,50m (roughly 98 inches)
wide. I haven't really tweaked my 8500 as I want to rebuild the 8500
Ultra (non LC :D) using the tubes out of the 8500 (almost mint condition)
and the pj isn't in it's final location, yet. People like Mike would probably
call my picture "broken" :D


On G2, I'm sure there must be an ideal or 'target' cathode current range
for the tubes but so far no one has reported any knowledge of it.

I wonder if one of our resident techs who has a full service manual set
could look in the waveforms/voltages or calibration areas and see what it
says about the subject?
--Bill

Yes, I will agree there is probably an optimal current range for the beam
or cathode current. However, the true adjustment is going to be largely
dependant on the condition of the phosphor (which is why I asked if
the coating ages, even w/o electron bombardment) and will most likely
also be different for each R G or B tube. To me it seems that beam current
is the only way to adjust the pj (unless you have enough tubes so that
tube life isn't critical :D)

As to the service manual: I have an original Christie service manual
including the schematics. However, there is no real calibration process
described in the manual (i.e. in that respect it's similar to the owner's
manual) and there are no waveforms shown. The only manual with a
few waveforms, that I have, is an older Marquee Training manual. IIRC
these waveforms were for trouble shooting only (no cal process).

However, it may well be that there is another "flow blown" manual out
there, which I haven't found yet. :D I do have the secret manufacturer's
code to enter the Service 0 menu. This isn't in the (my) service manual
either. As far as I know there is also a third (even more secret) code,
which delves even deeper into the Marquee internals. I haven't gotten
hold of this one yet. I believe only certain Electrohome (Christie, VDC)
employees have knowledge of this code, so most likely it will remain a
secret (Sorry, I can't disclose the Service 0 code, as I promised to
keep it under lock and key, and no I didn't find it on the internet and
I didn't receive it from a forum member either :D)

Greets,
Reinhard

Can any of the knowledgeable technicians and engineers tell me if the Resistors R13/R28 (400 ohm, 15 watt, 5%) should be closely matched?

On one board, one is 388 ohms (-3%) and the other is 405 ohms (+1.25%).

Since re-soldering is necessary/advisable, would it be advantageous to remove all 6 resistors and match up the closest pairs?

Originally posted by mp20748
Now I'll share a real nugget.


First we'll start on the vertical. ..........On the early version boards, you'll see that resistors R720, R721, R820, R821, R920 and R921 are mini one (1) ohm resistors. On the later version boards, there are two of these one ohm resistors in parallel, they are the plus and minus 15 volt supply rails for the vertical output IC's. So that would mean that the early version boards had a one ohm resistor on each of the IC's supply, and the later version boards have one half ohm (1/2 ohm) on each supply rail. In other words the upgrade was a doubling up of these resistors. I only had three problems with the design of these projectors, and the use of these mickey-mouse resistors is one of them. They're supposed to be fusable metal film resistrors at 1/2 watt :rolleyes:

........................There are twelve of them on the convergence board (24/stacked on the later version). R118, R117, R218, R217, R318, R317, R418, R417, R518, R517, R618 and R617. There are also 48 1.2 ohm (same mickey-mouse resistors) right near the power transistors.

:( Mike, On my vertical board, there are both configurations, R720, R820 & R920 are 1 ohm and R721, R821 & R921 are two 1 ohm. Am I correct in replacing all six positions with the double 1 ohm 1/4w 5% ?

As for the 48 1.2 ohm, I only spotted 24 on the convergence board (R168, R169, R170, R171, R269, R270, R271, R368, R369, R370, R371, R468, R469, R470, R471, R568, R569, R571, R668, R669, R670, R671) however, there are 24 of the same, in 2 ohm (R164, R165, R166, R167, R265, R266, R267, R364, R365, R366, R367, R464, R465, R466, R467, R564, R565, R567, R664, R665, R666 & R667).

Look back, Look way back
Earlier in the thread (page5) there was a lot odf discussion re the heater filament fix. The posters at the time were suggesting a combo of resistors plus a pot to reach the desired voltage. When reading that it struck me as being a little over complicated and I think I have a easier way.
Why not just remove the leaky pot (R2) and replace with a higher rated (larger pot) of the same value.
When I modded my PS supply all I did was remove said leaky pot.(R2) Since the new pot would never fit on the board I ran jumper wires from the board to the new pot which I installed on the side of the PS by drilling a hole in the case and mounting the new pot accordingly. This also has the advantage of easy adj access since the pot has a large adj arm.
I was able to adj the voltage in this manner with no trouble and since the new pot is rated so much higher than the original I feel the chances of a re failure arevery slim. Sorry i can not remember what the rating of the new pot was except it was 0-1K and it was a large bracket mount type as opposed to the tiny board mount original.
What do you think.

Mike, On my vertical board, there are both configurations, R720, R820 & R920 are 1 ohm and R721, R821 & R921 are two 1 ohm. Am I correct in replacing all six positions with the double 1 ohm 1/4w 5% ? They should be replaced in like values. The first three should be replaced with single 1 ohm's, and the last three should be replaced with double 1 ohm's.

There are only 24 1.2's that are actually 12 parallel pairs.

--Bill

When I modded my PS supply all I did was remove said leaky pot.(R2) Since the new pot would never fit on the board I ran jumper wires from the board to the new pot which I installed on the side of the PS by drilling a hole in the case and mounting the new pot accordingly. This also has the advantage of easy adj access since the pot has a large adj arm
Can you post some pics, what is the value of the new pot? Can anyone else chime in with opinions on whether this is actually the best choice for fixing this problem? Seems like we have a lot of ideas on what's the best fix but no consensus!

Originally posted by draganm
Can you post some pics, what is the value of the new pot? Can anyone else chime in with opinions on whether this is actually the best choice for fixing this problem? Seems like we have a lot of ideas on what's the best fix but no consensus!
Seems to me that if the fix is a 1W 1K Ohm pot, then the Bourns 3290H would be the best fit fix. It's a little expensive, available from Mouser, here (http://www.mouser.com/index.cfm?handler=displayproduct&lstdispproductid=212373&e_categoryid=31&e_pcodeid=65213) or review the Spec Sheet (http://www.bourns.com/2/pdfs/3290.pdf)
I have not tried this yet, my projector is very stable at 6.35V with the original pot.

draganm
I can't say if this is the best fix but I will say I have done 3 8500s this way and never had a problem.
I'll see if I can get some time for a pic tomorrow.
Steve

Bblue said:

"This time, I started calibration from that point (where it was), but first lowered brightness to 30. The net result was, basically, moving G2 up 3-5 numbers on all colors. The look of the resulting picture is completely different! The high whites look more like part of the picture, and the low blacks show more shades of gray. The whole presentation is more realistic and life-like and subtle details are clearer, as if it was higher resolution, but the resolution didn't really change."

I found the same thing prior to your post, and after first following Tims procedure, and was hoping some of these mods might fix this situation. If g2 is set on mine with brightness in the <35 range and contrast in the low 40's then blacks show best, whites look best and the whole pic just has more "pop" to it. What happens though is brightness adjustment over 40 or so is useless and makes raster light up like a christmas tree and totally wash the pic out if brightness is accidentally turned up.

I'm also wondering if this has any adverse affects as it really helps the pictures 3d effect. If it's not detrimental then it would also be nice to drop the available brightness range so it no longer can go so far too high.
(Whether that is already an option in a hidden setup menu I don't know, but both brightness and contrast could use a reduction of max range for me to feel safer with the pj adjusted this way.)

Troy

Thank you Glen and Steve, $25.00 is kinda steep, but a heck of a lot cheaper than shipping my LVPS to another state. I have downloaded the sepcs to the Bourne and I'll ask my elect. engineer about matching up something locally. I have one more stupid question, I will need to bring a multi-meter home after this mod to check and set P14. What do they mean by "checking across P14" . The motherboard says HTR1 and HTR2, is this the circuit you check across?:o
BTW, the G2 versus Brite/Contr posts are very interesting. I don't think Tim meant for his instructions to be anything more than a basic/beginner type guidline. I would like to see a more advanced procedure posted as a FAQ. Of course it would only be applicable to units with fresh tubes(less than 3K hour ).
How do you all feel about the recomended drive levels? R35, G90, B50

I have a quick and cheap way of checking the LVPS. Just hook up a pair of car jumper cables to your truck. If the projector can start it then the voltage to is high! *grin*

Im with Dragan on this one. Can someone post a pic what the pot looks like to adjust the levels?

I've learned on there that the best way to set G2 level is to turn contrast down then stick your face in the lense. Turn up the G2 level for the tube your looking into up untill you just "barely" see the raster glow.

Ok Here are some Pics.
This pic is of the front of the PS installed in the PJ. You can see the silver adj arm protruding. When the PJ is buttoned up the arm is just accessable through the plug in hole.
If you are going to do this you need to make sure the arm will not hit the PJ front cover. I had to install a couple of washers on the inside to make clearence.
This is my first attempt at posting pics on any forum so be kind.

OK
Here is the guts. This is a 1 Watt resistor so to my thinking it should be more than robust enough to last forever.
Please if i am wrong let me know before I have a failure.

Thanks
Steve

Here is one more all closed up to show the importance of checking your clearence
Steve

wow great pics, thanks energeezer. Your right, that looks like a pretty slick set-up. I will see if my electronics guy is comfortable with attempting something like that.

If you look back in this thread there is the correct fix for this..
The problem is not with the wattage of the pot you can put in a 1Watt 5Watt 1000Watt that does not fix the problem of the adjustment being way way to sensitive. The fix is to make the circuit have a limited range so if the pot moves a bit is does not matter. After the fix you should be able to move the pot all the way both ways and still be in the safety zone..

Could I get a pic of the stock pot? My TV repair main friend wants to check his new 8500 we just got in last night. We are going to check for the 6.5v but if its off a bit I want to know where the pot is to adjust this.
THanks )

Sly, why don't you resurrect fix that you designed. It sounds like a better way to adjust the voltage, mainly because it'll keep the voltage in the safe range.

This is a repost from page 9
click the link and look at the 2 images then the page should look OK

============================================================ ========
I keep getting asked about the LVPS mod I did so here it is.

This is the trace that needs to be cut.
http://www.avsforum.com/avs-vb/attachment.php?attachmentid=156330&d=1256592674

And this is the 475 Ohm resistor installed. I used a 805 size part rated at 1/8 watts you might want to use a 1206 for more wattage just to play it safe.
http://www.avsforum.com/avs-vb/attachment.php?attachmentid=156331&d=1256592674

Now just install a 20 ohm pot.

@mike parker,

have a problem when cleaning the boards with denatured alcohol.
after drying the board seems to be some kind of tacky or sticky.
when inspecting the surface there´s seems a thin layer of laquer which the alcohol is destroying.
the only way is to do the cleaning several times until the tackiness is gone.
afterwards the pcb have a matte look. before it has a shiny surface.
i found out that when using a chemistry for cleaning pcbs the result is the same.
did i miss something?

regards,

gernot

Wondering if someone who knows calibration can post their 8000/8500 settings. I ask this because I always followed the G2 guide, but had the drive set at around 60 for all tubes. I talked to a person in town who did Ehome setups and he said that was FAR too high and they should be around 10. I was very surprised when I lowered all 3, that it still looked good and the stairstep was fine.

So I was wondering if some people who do calibration could post what the settings range 'should' be, as well as the overall bright/contrast. I know it will vary for people, but I think if I had been told to tame the drive a while ago, my tubes would have been all the better for it.

Mine are:

bright- 50
cont - 50
red drive-10 g2-71
green drive-10 g2-65
blue drive-41 g2-76

hi,

need a little help, please.
on a friends 8000er, i did the chip upgrade to cl449s and resoldered the 2 resistors and 2 coils. also two new e-caps were fitted.
sadly the streaking is less but not gone.
did i miss something?

regards,

gernot

Did you do the VIM or just the neck boards?

hi,

just the neckboards.
after reading this thread again, maybe the 2 coils need to be replaced too?

regards,

gernot

Originally posted by gerni g.
@mike parker,

have a problem when cleaning the boards with denatured alcohol.
after drying the board seems to be some kind of tacky or sticky.
when inspecting the surface there?s seems a thin layer of laquer which the alcohol is destroying.
the only way is to do the cleaning several times until the tackiness is gone.
afterwards the pcb have a matte look. before it has a shiny surface.
i found out that when using a chemistry for cleaning pcbs the result is the same.
did i miss something?

regards,

gernot


Gernot,
I don't know specifically about the Marquee boards as I have not examined them too much yet but many circuit boards have a Conformal coating that helps prevent corrosion of the traces on the board. I know from first hand experience (was actually trying to take the coating off) that with enough alcohol and friction from say a cotton swab the coating will come off. It should not really hurt the function of the board at all.

Steve

Gernot,
You have to get rid of all 409's in the video chain, including any in the VIM. Until you are 409-free you'll have various degrees of streaking and reduced resolution.

--Bill

hi bill,

the vim has h1100 in it.
also i swapped in a newer vim board with 449ers.
the streaking is still very visible.

two neckboards are 50-2000-02p issue1, the third is issue2 and this has absolutly no streaking.
when comparing the layout, they look the same.
also i replaced the coils L8 and L9 but this had no effect.

i´m wondering what i missed or is it the boards.
if there´s no remedy i need 2 replacement boards.

best regards,

gernot

Originally posted by Tester007
Wondering if someone who knows calibration can post their 8000/8500 settings. I ask this because I always followed the G2 guide, but had the drive set at around 60 for all tubes. I talked to a person in town who did Ehome setups and he said that was FAR too high and they should be around 10. I was very surprised when I lowered all 3, that it still looked good and the stairstep was fine.

So I was wondering if some people who do calibration could post what the settings range 'should' be, as well as the overall bright/contrast. I know it will vary for people, but I think if I had been told to tame the drive a while ago, my tubes would have been all the better for it.

Mine are:

bright- 50
cont - 50
red drive-10 g2-71
green drive-10 g2-65
blue drive-41 g2-76

The numbers mean nothing here, they are only relevent to the many variances of the Marquee.

There are three different board numbers for the VIM, with different part numbers for each. Each of these boards would display a different number based on the same or similar display image. Two of the board numbers have differenent analog multipliers (electronic gain controls). One of the version of the VIM has resistors added to the output of the first OP Amp's output to ground. And don't forget the condition of the tubes...

The numbers should be used to document the settings for that particular projector only. Even when we installed brand new Marquees on a multi-diplay, or stack setup, the numbers were not the same.

:D

Time to bump this thread at the top


Brido

hi,

almost forgot, streaking problem fixed.
thanks to aksor, billblue and curt palme.

also i´m very delighted how stable the convergence is with the micro-ohms.
this rocks.

regards,

gernot

hi,

Q13 on the neckboard was the bad guy.

regards,

gernot

Hi,

to bump thread and :

horizontal deflection module 220 mfd @ 160vdc cap (C50)

what type of cap do I have to buy ?

Thanks!

Brido

Hello fellow Marquee gents!

I have a 9500LC that I'm getting ready to be ceiling-worthy so I've been reviewing this thread again to make a list of things to do and buy.

Thanks to everyone, especially Mike, for contributing to this.

I have a question/suggestion regarding the 1ohm (piggybacked) resistors on the HDM and Convergence boards. These resistors are used for power supply decoupling to keep noise from going into or out of the power supply rails on the board. Typically a designer will choose a capacitor for bypass and then a small value resistor (1/2 ohm in this case) because resistors are cheap. It seems like a ferrite bead has the potential to be better in this application since the temperature coefficient of the bead is going to be smaller (better) than the resistor. Anyone tried this?

I'll probably give this a shot unless someone has had a bad experience with using a bead in this spot. I'll need to find some way of measuring any improvement or degredation though.

Cary

Originally posted by mp20748

First we'll start on the vertical. <snip>
On the early version boards, you'll see that resistors R720, R721, R820, R821, R920 and R921 are mini one (1) ohm resistors.

My 1993 (re-mfg in 1996) 9000 with non-hinged heatsink does not have any resistors with a number higher than R1xx. There are no R7xx, R8xx or R9xx anywhere on the vertical board.

Any ideas on how to mod it in conjunction with the convergence board to ensure tight, non-drifting convergence?

ALSO, is the max4224 still the 449 replacement or is there something better now?

Thanks :)

Can anyone help with getting some CLC449AJP

I am looking to replace the 409's.

Any help would be great,
everyone I have checked with has NO stock at all.

I am looking for 24 of them, enough to do 2 projectors.
12 for each projector is correct isn't it???

Apprecaite the help

Cheers
Steve

Originally posted by CZ Eddie
My 1993 (re-mfg in 1996) 9000 with non-hinged heatsink does not have any resistors with a number higher than R1xx. There are no R7xx, R8xx or R9xx anywhere on the vertical board.

Any ideas on how to mod it in conjunction with the convergence board to ensure tight, non-drifting convergence?
:)

CZ,
I have this info in one of my folders. However, I don't have the folder at my present location. But from memory, if you're able to identify the 1 ohm resistors on both boards, then replace them with 0.47 ohm @1/2 watt, or double stack the ones that are there with another 1 ohm resistor. These resistors should be fusable, and should be placed away from the PC boards surface.

2nd. Resolder all of the power transistors on the convergence board, as well as the three output IC's on the vertical board.

3rd. apply a thin coat of patroleum jelly on the heatsinks insulator, where the transistors (convergence), and IC's (vertical) will make contact with the heatsinks.

last, but not as important, replace all 1.2 ohm resistors on either board to a 1 ohm @ 1/2 watt fusable.

-------

Steve,
CLC449's may be difficult to locate (obsolete). There's a replacement from the National Semiconductor site. I've tried the replacement, but did not like the results, but others have had good results.

Thanks Mike, I have had a hard time finding the CLC449's.

I have had an offer from someone here to supply me some,
This should happen next week.

Mike, Any Ideas for my MP1 hookup onto the ATI X800?

Cheers
Steve

Originally posted by HK-Steve


Mike, Any Ideas for my MP1 hookup onto the ATI X800?

Cheers
Steve

I've yet to attach a mod to the new ATI X800. I'm expecting one in to transfer a mod to. I'm looking forward to getting my hands on one, because I'll need know how to do this for the soon to be released MP-1v3. So far, you and two testers are the only ones with the new MP-1.

Do you have pictures?

Yes, I have already sent them through to you, a week or so ago.
Did you not get them??

I can resend if you need me to.
Let me know.

Thanks
Steve

Mike, I just saw your Thread about your email problems.

I have just sent them to your AOL address,

Cheers
Steve

Mike,

Do you have any details on how to hookup the MP1 to the ATI X800?

Any Impressions on the X800?

Cheers
Steve

Originally posted by HK-Steve
Mike,

Do you have any details on how to hookup the MP1 to the ATI X800?

Any Impressions on the X800?

Cheers
Steve

Not yet, I'm still waiting to get my hands on one. Someone is supposed to be sending me an X800 and I'm really looking forward to its arrival. So far I've recieved a Gforce 6800 that I've attched a MP-1 to, but I've only had time to play with it for a few minutes.

With the newer mod that you have, it's very important to use either RG59 or RG6 cables.

I believe you'd sent me a picture of that card.. I'm not sure if I've seen it, as i've had email problems, so could you send it again to the listed email (below). I'll do what I can to help you make this happen.

I've been busy with several things, but with much attention to the MP-5. I was able to test the demo power supply this past weekend, and I must say I'm very pleased with it :) , it's far far better than expected.