AVS Forum banner
Status
Not open for further replies.
1 - 20 of 25 Posts

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #1 ·
Could someone explain to me how to target a specific gamma in a Marquee series projector? Mine is a 9501LC which I've retubed with the newer Panasonic-gun rebuilts from VDC. I've done the full setup from physical to electrical setup, and am doing the final calibrations with the Colorfacts 6000 system and Eye One.


I've tried a variety of techniques for G2 and drive settings and general color tracking and though the resulting display is pretty good, the overall 'look' of the picture is dark, even when black level is raised above normal. I concluded that it's a gamma issue, which Colorfacts is reporting around 2.5 -- it should be closer to 2.2.


I found by experimentation that the gamma seems to be influenced not just by the ratio of G2 and Drive, but also the relationship of those to brightness (black) and contrast (white) levels. With some juggling I got the gamma down to about 2.35 -- which is an improvement -- but doesn't quite go far enough.


So can someone explain the best sequence, order or relationship of these adjustments? No documentation that I have found seems to discuss this directly or indirectly. If start with the usual setting of g2 to *almost* diminishing levels and work up from there I seem to get the highest gamma numbers. If I go the other way and balance the high luminance levels first, and work G2 down to a suitable measurement from Colorfacts I seem to get an excellent black level but still a slightly elevated gamma.


The second, and possibly related question is about the blue tube. I'm getting an overall light output with guns calibrated for reasonable gray scale and working at 100 IRE, of about 25 nits which calcs out to about 7.3 ft.lamberts at the screen. I can push it beyond that (30 nits) at a cost of some detail. But looking at the gun output curves, red and green are pretty flat from about 20 IRE to 100 IRE, even pushed. But blue starts rolling off between 50-60 IRE. The effect is that at 50 IRE I can be at a color temp of 6500, but at 80 could be down to 6000 or so, directly related to the blue rolloff which is consistent and predictable regardless of G2/drive settings.


Is this a 'normal' characteristic of blue output? And if not does anyone recognize what might be the cause? I have attached a couple of charts as a reference. uh, well just one chart. Not sure how you add a second one.


--Bill
 

·
Registered
Joined
·
4,916 Posts
Bill,


I can see that your B is low at the higher IRE's. You need to setup the projector so that the 80-90 IRE is at 6500 using the Gain or Drive controls. Which in this case should bring up B and force you to raise the R as well. Then go to 30 IRE and set it for 6500. This will initially yield a hump in the B through the middle of the range which is normal. The trick is to get the hump (or as much as you can) out of the middle. You can use a slight defocusing of the B to help with this.


Terry
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #3 ·
Thanks, Terry, but I'm well past that. This rolloff is not something that can be corrected by balancing G2 against screen like you suggest. Regardless of how it is set -- even with a large hump between 20-60 IRE or so, as output increases all the tubes increase output except blue. It seems to hit an output ceiling and then just stays there regardless of any adjustments.


I had orginally thought there might be a problem with the HV voltage divider, but measuring at each tube it's right on at 34.9kv.


Focus and astigmatism look good on it and quite comparable to the others, and the dots stay uniform through the usable lightness range. So far I'm NOT underfocussing the blue, though. It doesn't seem like that would help to the degree I'm seeing.


Also, the rolloff on the blue is worse the more screen you use. For example, the graph I attached was the RGB peformance with the test area about 4' square centered in the screen (54" x 96" 16:9). If I do the same test full screen, the rolloff on blue will occur much earlier, while red and green stay uniform through the entire range.


--Bill
 

·
Registered
Joined
·
264 Posts
What is your source??

if it is a HTPC you can just load gamma values into the video card and make it perfect.. I think most if not all the video cards now have a gamma table in them . I have not looked much on the pc I know you can set a static gamma for each color. But I bet there is a way to load a table.

for a 8bit per channel card there should be a table of 256

for each color the value is most likely a 10 bit value but that depends on the hardware.



find the table and fix the blue if the table is right it should be gamma of 1 and perfect. after everything is set up

that is what it is there for is to fix this kind of thing.
 

·
Registered
Joined
·
4,916 Posts
Bill you are correct in that under focusing the blue isn't going to get you where you need to be.

But can you set up 100 IRE so that you get 6500K and say 30IRE so that you also can get 6500K? Also what are the coordinates of your Blue Primary?

If you are able to get 100 IRE and 30IRE set, what are your light output readings at 90, 80, 70....down to 30 using the gray scale?

Terry
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #6 ·
Sly, no HTPC here. There are several sources, a Samsung HD receiver, GI C/Ku Satellite receiver with HD outboard, and a Dish6000 receiver. Those are all RGB/RGBHV out, fed to the ehome switcher. There's also a Meridian DVD player through a Faroudja scaler, also with RGB out.


So I have no way to adjust gamma with or through the sources, only at the projector. I guess one underlying question is what can be expected from projector, i.e., can it actually perform at perfect gamma (achieved somehow) or is gamma typically high on that series of devices?


--Bill
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #7 ·
Terry, within reason I can drive blue to match RG at 100IRE and also maintain it at 30IRE, but if I do that the hump in between 30 and 100 is enormous. Whatever I set blue to at 100IRE, increasing light output (contrast adjustment) will increase RG but not B (it just hits a certain point and can't go any higher). Forget a setup for D9300, you end up with the enormous hump.


I've attached the primaries CIE chart from Colorfacts for the coordinates.


I don't think I have saved a luminance chart with the blue set that way, but will check and retest if necessary.


I'd be happy to retry suggestions, but am pretty certain that there is a non-adjustment something wrong -- the tube itself, something that is current limiting when it shouldn't, incorrect filament voltage, or something. (I just ran across the maintenance thread and will be checking that out.) Was hoping this was an ehome recognizable type problem. Machine was manufactured Jan 1997.


--Bill
 

·
Registered
Joined
·
4,916 Posts
Hopefully Tim will step in.


Terry
 

·
Registered
Joined
·
264 Posts
A few things


if you at the max output of the blue you will need to bring the red and green down to that level.

I think you might be trying to run it too bright??


the 2nd thing is check that you do have a gamma part on the VIM

look at the maintenance thread for how to check.

it should but you never know what might have been swapped. this is the first I have heard of the blue bump

I think that it can be fixed easy with some values in the gamma circuit. as long at you do not try to over drive the whole PJ. I do not have a color meter to test this my self.

in fact I have not even had the time to get my PJ running again I bought a newer one but it is just setting in the garage.
 

·
Registered
Joined
·
6,833 Posts
Swap the Blue neck board with the Red neck board. It's more likely the neck board, or even the VIM. Inability of any of the three colors to properly track is the best indicator of a neck board that's not funtioning properly.
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #11 ·
Sly, I'm running it such that 100IRE is roughly 7 ft.lamberts, which to me seems marginally low. The rating of 270 Ansi lumens for a 9500 should be able to produce 9.75 ft.lamberts. Even at 7, I'm in blue rolloff.


Mike, here's what I know so far in checking the various boards. First, heater is 6.31v in run and 4.0v in standby on all three tubes. The VIM has the newer component configuration and blue gamma mod as discussed in the maintenance thread. While I was there I removed R152. I've pulled the red and blue neckboards, also newer with CL449A front end chips and SOIC's.


The neckboard on the blue tube showed mild heating around R13 and R28, but no evidence of heat on L8 and L9. I cleaned and resoldered both ends of the two resistors but have not replaced the caps yet.


However, the neckboard on the red tube is a mess (this one has been working just fine, but probably not for long). The R13 side of the board showed some pretty severe heating, and both ends of it had very little solder left. Additionally, R47 was burned open (circular burn line around the body center) and R48 had turned white with heat. When I probed it lightly both ends it just fell off the solder pads, but it measures correctly at 22 ohms. These two are in series between 85v and the collector of the crt grid driver. How the red was working is a mystery, but it showed no outward signs of a problem! (my schematic calls the +/- 85v lines 85v and +85, no reference to a - anywhere). Does this sound like anything else is wrong other than C26 as you state in the maintenance thread? L8 and L9 look fine.


C26 is a brown 100v 22mf, whereas C1 is a blue one of the same rating. Both are Rubycon's and neither look to have been replaced since manufacture. My parts list for this board says these two should be 160v minimum.


Do you know why R47 and R48 are spec'd as metal film? I've only got some carbon films handy, and will have to go capacitor hunting tomorrow ... Saturday, good luck. Maybe Fry's but I doubt it.


I'll continue with the red/blue neckboard swap after I've located the replacement parts.


--Bill
 

·
Registered
Joined
·
6,833 Posts
Bill,

I'm surprised that you got the gamma curve that you did. And blue will always be stubborn if there's a problem with the neck boards.


I recommend that you replace R47, R48, C26 and C1 on all three boards, and repace R46 and R31 as well. It appears that your neck boards had experienced some heat damage, and that if any of the blue caps (C26,C1) have changed color, you'll have to change them all. The original caps are low temp and should be replaced with high temp caps (105c). But make sure that they are 160Vdc.


R46, R47, R48 and R31 should be replaced with 22 ohm 1/4 watt metal film resistors. Carbon resistors should not be used in this circuit. And make sure to clean the surface of the board with denatured alcohol or some other cleaner after you replace the components. Try to replace the parts without removing the heatsink. If you have to remove the heatsink be very careful, because when turning the six bolts to the power transistors the 4 leads on the power transistors could break loose (damage). If you cannot perform the task of replacing the 6 parts with the heatsink in place, and should have to remove the heatsink to get to the bottom of the board, unsolder each of the four leads on the six power transistors first.


When done, you should experience a much better image.
 

·
Registered
Joined
·
264 Posts
Well it looks like you have some Work to do.

The cap that is brown is that way because it is burned out ( it got so hot it changed color).

where are you? I am in Milpitas.

if the necks have that much wear and tear you might want to check the other parts that Mike has outlined in the maintenance thread.


you may need to change your proverbial timing belt and oil.
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #14 ·
Thanks, Mike. Will pull the green neck board as well and replace all as you suggest. Now my Saturday quest for parts begins.


I don't see any reason why I should need to remove the heatsink as both sides of the board are accessible with a longer soldering tip. I guess I should be happy that the boards are all the newer variety and searching for chips as well won't be necessary!


Sly, I'm in San Diego (actually El Cajon, on the east side). The timing light and grease gun are standing by!


I've attached a photo of the damaged red board. R47 (the one that split) is directly under R13 in the photo, and R48 should be on the pads to the left of R47. Had to drop the photo down to 640x480 so a lot of detail is lost.


--Bill
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #15 ·
Well, no joy on any of the components today. Hopefully, Mouser will have quick delivery stock on all the upgrade items, though right off the bat I don't see any listings for fusible metal film resistors.


--Bill
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #16 ·
Finally, the new parts are in and the neck boards swapped. The pj even works!


I took the downtime opportunity to perform most of the mods in the Marquee Maintenance thread (except 1 ohm replacements, that's for later) plus the 22mf's in all neck boards and 22 ohm resistors in the red neck board. Originally, the red neck board was really fried, the blue had just a hint of heating at the solder joints of R13 only, and the green board was perfectly clean. I put the green board on the blue tube, blue board on the red, and the burned/repaired red board on the green tube.


After all this when I powered up, the LVPS wouldn't start up. I traced that down to the green neck board (the old damaged red one) and found that C58, a shunt cap to ground had a 300 ohm resistance. Apparently it had been fried when the 22 ohm resistors burned, or perhaps it went first and helped them along. C58 is a 10nfd SMD which I temporarily replaced with a .015mfd tubular. After that, everything powered up and worked fine.


The behavior and 'feel' of the tubes calibration is quite different than it was, but the blue still rolls off starting at about 80IRE. I've been able to get pretty good tracking from 10 to 80IRE, and red/green are quite good well beyond the 100IRE level, but not blue. I've attached the RGB chart to show its behavior.


So what are we down to, the tube or VIM?


--Bill
 

·
Registered
Joined
·
6,833 Posts
Bill,

it appears that you still have a blue "gamma" problem. As you may already know, the blue gamma circuit (U22 on the VIM) should have improved on the 80 IRE rolloff.


At this point you'll need to swap out the VIM to determine if the problem is the tube or the VIM. And it would also be good if you were able to scope out at the BNC's (BNC 'T' connectors on RGB input of VIM) to rule out the input source.
 

·
Registered
Joined
·
264 Posts
If you do not have a extra VIM try swapping the blue with the green in the VIM just move the mini-RCA plugs

and redo the plot . If it looks the same there is something wrong with the gamma circuit on the VIM. If it a lot worse then you are at the what the blue CRT can do..

It would be nice to see this plot taken on a different Electrohome. Just for control..If the gamma kicks in at the right spot but it is just not giving it the boost needed and the CRT can go brighter .The circuit could be changed to add more boost.


if you have a chance look at VIM R200 R201 R157 R159


and R177 R179 and see what the values are.


I show 475, 330,435,100


and R177 and R179 depends on the Vim

one is 24.3, 39.2

and the other is 56.2,18.2 this sets the gain so it might be different.
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #19 ·
Interesting. Well, I don't have a spare VIM so will have to work by process of elimination. Will try switching blue/green and do basic measurements of the gamma circuit components. Could be a defective zener diode I suppose.


It hasn't been exactly clear to me what specifically the blue gamma circuit does. The documentation I have says that it adjusts the blue tube gamma to more closely track with the red and green tubes, but it doesn't show just what that difference is. The circuit looks to me like it has at least two 'poles' that would affect blue gain in different ways. One looks like it increases gain at a certain point, and the other decreases gain at a certain point. Where those points are relative to the start of my rolloff isn't clear. Does anyone know?


I don't think my blue problem could have anything to do with with blue input to the pj, as this rolloff effect is clearly seen as a response to contrast and drive, regardless of level of input. Also, there don't appear to be any current limiting circuits except through b-limit, which affects all colors in unison, not just blue. So the problem would seem to have to be in the gamma circuit or the drivers right in front of it, unless it's an issue with the tube or something directly connected to it that isn't under control of the neck board, like the G2 voltage. G2 is pretty high impedance, is this area prone to failure of any kind?


I do have a friend with a Vidikron 2 (non-LC) but we haven't been able to get together yet to plot his pj with my Colorfacts . It's planned at some point, but I know not when.


I received the Commscope 73501S video wire today and may get into replacing that while the VIM is out. Any caveats I should be aware of other than matching length?


--Bill
 

·
Registered
Joined
·
1,056 Posts
Discussion Starter · #20 ·
No joy on the VIM components. I checked everything in the blue gamma circuit except the 449 IC itself, and for the driver stages ahead of that compared component measurements to the corresponding components in the R & G sections -- all were extremely close.


I reversed the blue & green connections from VIM to neckboards, and also my inbound B & G feeds so I could still setup and calibrate. The results were, uh, interesting and showed that the gamma circuit seems to be working (unknown if it's working 100%) and that there's no way at all to get any sort of gray scale tracking in this mode. The blue rolloff was even more severe without the gamma circuit. I've attached an RGB graph of the results in this reversed mode with the blue gamma driving green, and the stock green driving blue.


So now what? Unless there's a degree of correction problem on the VIM, everything that drives a tube has been swapped. All that would appear to be left is the blue tube itself, and the G2 supply.


Is there anything mechanical on the tube that if misadjusted enough could cause this behavior while at the same time allowing reasonable positioning range, focus and astigmatism adjustments?


Does anyone have a colorimeter that could show me an RGB plot of a LC ehome pj to serve as a comparison to what I'm getting?


--Bill
 
1 - 20 of 25 Posts
Status
Not open for further replies.
Top