what would cause this.. banding on the right. is it from the coils? its happening on all three tubes. can't turn up the contrast above 50

http://bomrat.org/blane/banding.JPG

what would cause this.. banding on the right. is it from the coils? its happening on all three tubes. can't turn up the contrast above 50

The banding is actually on there on both sides. It goes across the screen, but cab be seen more on one side or the other. Or even in some areas more than others.

When doing your initial setup, you'll have to go into "conv" and select "service"

Once you key in the password (0901), you then select random and hit zero (0) on the keypad. That should feather out the grid. And once that's done, you should not see any banding.

Now from there, you're goal is the accomplish geometry without causing the banding. You'll have to do this by constantly scrolling (pressing the button above the power button on the remote) from the setup grid to the white field in the test patterns. the white field helps to reveal the banding.

When you find that you've pushed it too far, go back into the 'conv' 'service' menu and start over.

Banding is not magnetics related, but it's very important to have the magnets centered properly, or you could be creating a setup that would need more geometry setup than it would normally need.

And you'll also want to make sure you "BOW" ('Geom' on remote) stays at 50. If it appears that the bow needs setting, it'll be best to correct that by projector positioning.

I ssem to remember having my screen look like this once and having it go away by switching the retrace rate (a toggle in the menu from long to short or visa versa) or possibly by changing the sync rate (another toggle from slow to fast or the other way). From my experience with my 9500 LC Ultra it can not be set up without banding if the bow controls are set to 50. If you do this and then try to fix everything using converge on green random and/or interpoated) you will get the traditional broad horizontal bands, usualy two banda across the screen darker than the rest of the screen.

As MP says reset the green and then go to geometry. Go easy here and check at every step by toggling between the internal grid and internal white field. It can't hurt to set the internal test pattern frequencies to one close to what you will be watching.The frequency test pattern setter is in source set up under service. As long as you don't introduce banding while using the geomety controls you will be OK. but don't be afraid to use the bow but don't use it so that it introduces banding.

When you finally go back to random access converge on green, the problem will be using it and not introducing horizontal banding. You must continually toggle viewing the green field that will appear very very carefully for the slightest appearance of banding. Every time you make a change TOGLE and then save. If you save and then additional chnges introduce banding you just go back t the saved confiqure and continue. Generally the changes that introduce banding are not edge changes or center changes but the ones along the one thrid lines. Very small changes up will band you. If you get it go back to the no band save and adjust things along the edge and then continue. You will get it.

mike parker, please call ken w...

Set your retrace to long.

Is this just an issue on Marquees, or am I just lucky that I haven't seen it on my 808s? I pushed geometry all over the map while learning the system but didn't ever notice such a thing.

Is this just an issue on Marquees, or am I just lucky that I haven't seen it on my 808s? I pushed geometry all over the map while learning the system but didn't ever notice such a thing.

This isn't a geometry issue it is a retrace issue. Set retrace to long and thr problem will go away.

Chris.

Running retrace at long solves this problem, but introduces another. If you are running Blu ray, (not sure about HD) 1080p @ 60 you loose about 1/4 of the image width. I have the same issue :eek: and posted about it on that other site. I would love to get rid of it.

I guess I will do a full reset again armed with more tricks.

Jim

Is the issue the vertical light and dark bars on the LEFT of the raster? If so, this is caused by "ringing" in the horizontal deflection coils. The dark bands are where the sweep is moving quicker than normal and the brighter bands are moving slower thus allowing more energy to be dumped in a smaller area. No cure for this. The only option is to phase the video to the right so as to avoid the bars.

Scott

Running retrace at long solves this problem, but introduces another. If you are running Blu ray, (not sure about HD) 1080p @ 60 you loose about 1/4 of the image width. I have the same issue :eek: and posted about it on that other site. I would love to get rid of it.

I guess I will do a full reset again armed with more tricks.

Jim


Doing a full reset won't fix this. The only way to fix this is to use long retrace and either a video processor that will allow you to change the porch settings or use a HTPC that will allow porch settings.

I sold my HD-DVD player and Blu-ray player and got the LG HD-DVD/Bluray combo drive. The porches were very short on both my players allowing me to only use the short retrace setting which caused the problem you are seeing.

I can now use long retrace with my HTPC with longer porches and the problem is gone :)

Chris.

Is the issue the vertical light and dark bars on the LEFT of the raster? If so, this is caused by "ringing" in the horizontal deflection coils. The dark bands are where the sweep is moving quicker than normal and the brighter bands are moving slower thus allowing more energy to be dumped in a smaller area. No cure for this. The only option is to phase the video to the right so as to avoid the bars.

Scott

Are there any mods we could due to reduce or eliminate the ringing?

Chris.

Doing a full reset won't fix this. The only way to fix this is to use long retrace and either a video processor that will allow you to change the porch settings or use a HTPC that will allow porch settings.

I sold my HD-DVD player and Blu-ray player and got the LG HD-DVD/Bluray combo drive. The porches were very short on both my players allowing me to only use the short retrace setting which caused the problem you are seeing.

I can now use long retrace with my HTPC with longer porches and the problem is gone :)

Chris.


Chris,

If its not too much trouble, is there any chance you could pm me a list of components for your htpc? I had great success with my first one, but failed miserably with my last one. Also, are you using the software that came with the drive?

Jim

I switched my projector to short trace and bingo, I duplcated the problem. Doing nothing more than switching to long retrace eliminates the problem. Of course it isn't banding. It looks just like ringing though my Marquee guru says it isn't ringing. Regardless, going long retrace should eliminate it. How about some confirmation Bomrat?

My 8500 with a slight amount of banding showing in the first image (short). And with it also in 'SHORT' in the second. Back to long in the last.

Note the ringing to the far left of the image in 'SHORT'


http://img515.imageshack.us/img515/8010/img2358ce1.jpg

http://img263.imageshack.us/img263/3362/img2360ht1.jpg

http://img514.imageshack.us/img514/1151/img2359ts2.jpg

I used to have the exact same problem when i'm running 1080p at 60hz into my projector and when the retrace is set to short. It's gone when retrace is set to long but like someone said, the image is cut significantly. Then I was feeding the HDMI signal into Ophit DDA transcorder (external) then sending the analog signals through the BNC cables.

But I did notice the moment that I start to use my moome card and feed the signal straight into my projector, the banding is reduced significantly, still there but much less obvious.

Is this just an issue on Marquees, or am I just lucky that I haven't seen it on my 808s? I pushed geometry all over the map while learning the system but didn't ever notice such a thing.You will see this also in Barco if you shrink porches too much from source.

MP. What frequency are you running at for these pictures? I have found the higher the resolution, the more severe the banding. Very careful use of geometry and random converge on green will eliminate it on my Ultra. C lins must be used to center the image horizontally and verticallly by marking the screen with center edge markers, This of course requires the use of the other geometry controls to get the lines through the center straight and parallel to the edges. Bow must be used then ne must be very very careful with converge on green but perfect grid geometry can be obtained with a lot of patience. S lins will cause banding if c lins are not set correctly.

To me, this is no fun at all. it takes a lot of time. Its a shame the engineering design of the machine wasn't better. Any banding is simply iunacceptable to me. It is game disqualification for any projector set up that shows it. Same for any blend zone that shows in a blend set up. from what I understand blend zone apearance can be eliminated but the effort to do so is super human with CRTs.

MP. What frequency are you running at for these pictures. I have found the higher the resolution, the more severe the banding. Very careful use of geometry and random converge on green will eliminate it on my Ultra. C lins must be used to center the image horizontally and verticaly by marking the screen with center edge markers, This of course requires the use of the other geometry controls to get the lines through the center straight and parallel to the edges. Bow must be used then be very very careful with converge on green but perfect grid geometry can be obtained with a lot of patience. s lins will cause banding if c lins are not set correctly

It usually starts to show up around 960P, and get worse at 1080P, or higher as you indicated. In my setup I'm running 1080P, and this is an Ultra. As you can see, it was easy to make it raise its ugly head.

You seem to have the process down quite well, because it's not easy to make go away.



To me, this is no fun at all. it takes a lot of time. Its a shame the engineering design of the machine wasn't better. Any banding is simpy iunacceptable to me. It is game disqualification for any projector set up that shows it. Same for any blend zone that shows in a blend set up.

I agree, and I've asked them about this many times. I was told later that it was being looked into. I think i may look into this further myself.

After the power went out at Williams on the day before the meet. I turned on both projectors and one of them had lost all setups. It's a good thing I had a spare control module with me. But as well, I spent that morning redoing the setups and getting rid of that dreaded banding. because you're right, it's not good at all for a blend setup. So It's something that I'm going to look into very soon.

keep in mind, this is only a major problem on the higher scan rates, with it being worse at the higher horizontal rate. And that may explain why It sliped by redesign. I don't think there were that many complaint at the time about it because of the limited horizontal rates they were using then. 1920x1080P is a tough monster all the way around. The banding issue is my last fork for the Marquee.

When I get my Radiance back from William, I try to set up with no banding at 1080p 60. I hope I can do it but I am not at all looking forward to doing it. I am getting much better at it though. Why?

Because I am forced to do it myself. I'd rather pay somebody and do something else with my leisure time but this CRT set up sport has become a do it your selfer. Its only for the hobbyist and I want to change hobbies. I didn't pick this one. It picked me.

But yesterday i watched football all day in my HT. Mostly at ntive 720p. The late game was in 1080i. I had to screw with a new set up for 1080i but at that freq, banding is not an issue.

what causes horizontal banding. What lack in the machine does it. It would seem fixable by making some board or supply better. haven't a clue. Why does it get worse with frequency increases? Tht sounds perhaps like a band width limitatio somewhere. I know how to cause it and balance the machine very delicately to get rid of it but the effort.

If somebody had a clue what exactly was the limitation in design, we could fix it.

The "banding" or "jailbars" is caused by the sweep speeding up and slowing down due to ringing in the deflection current caused by damped resonance of the winding capacitance of the horizontal deflection windings and the series inductance of the yoke leads connecting to the HDM.

It would help to connect the yoke directly to the HDM but is not very feasible.

The fast retrace could be made a little less fast by increasing the value of the quick retrace capacitor on the HDM. The faster the retrace the harder the yoke/lead circuit is "strummed". That is why the long retrace has less of this effect.

Scott

The "banding" or "jailbars" is caused by the sweep speeding up and slowing down due to ringing in the deflection current caused by damped resonance of the winding capacitance of the horizontal deflection windings and the series inductance of the yoke leads connecting to the HDM.

It would help to connect the yoke directly to the HDM but is not very feasible.

Scott

Hmm... I think I'll take a stab at this later tonight..:)

thanks!

Tse. The banding that I have trouble with is horizontal bands, not vertical jail bars. Are you talking about eliminating the vertical bands (which look like ringing on the sides)? Obviously going to fast retrace will reduce the width of the raster. I originally tried slow to make the image wider but got ringing at higher frequencies/.

I tried mikes method for both my PJ my 8000 i was able to get rid of the banding but on my Reference 8(8500 AC) i cant get rid of it. I'm converging the R8's green grid to the 8000's in and up down stack and jst cant seem to get it! tried for a couple hours last night starting over each time. I think its the linerity I'm having problems with and the fact its two different PJ's. Also if i do the long to short i get the jail bars too. I think i'll try next to converge the 8000's to the R8's and see how that goes after i get the R8 not to band. Wish i had two Lumagens in the chain on for each PJ in that it helps to be able to manipulate each image individualy than together as I am now.

You guys are a big help !

Athanasios

Sorry to steal the thread but in AmPro terms, would this be the same as flyback wide and narrow? I have exactly the same thing from 960P and up.

Chip

You have to get the C lins dead on to avoid having the s lins cause banding.

You have to get the C lins dead on to avoid having the s lins cause banding.

I think thats the problem. I throw up the 8000's green grid and try to cover it with the R8's green grid. But i cant get the same C lins for the grid square sizes and have to use the convergence to get close. Should i try as much as possible to get the grid close with convegence then go to the C lin then back to convergencethen back again?
Or do the C lin first before doing the convergence over the 8000's grid then converge the R8 and then touch up the C lin and back to final converg?

confusing i know ! ;)

Athansios

the jailbars effect is the one seen on the left of Arli's picture. this has to do with retrace/porches

the regular banding most people talk about is horizontal, and can be seen on all pictures in this thread (usually in groups of 4 equal sized bands). this is to do with how hard you drive your circuits, geometry and convergence

Some time back I've played around with this "banding" problem, and believe it to be in the vertical circuits (VDM). I've also noticed that the VDM used in the Ultras were a little different than what was being used in the non-Ultra's. However, I also believe that the improvements noticed in the the NON Ultra's may have very little to do with VDM.

Later tonight I'll look at the banding problem again.


The board on the left is out of my Ultra. The one on the right is from a non Ultra.
http://img230.imageshack.us/img230/8236/img2364an3.jpg (http://imageshack.us)

the jailbars effect is the one seen on the left of Arli's picture. this has to do with retrace/porches

the regular banding most people talk about is horizontal, and can be seen on all pictures in this thread (usually in groups of 4 equal sized bands). this is to do with how hard you drive your circuits, geometry and convergence

So this is why I am probably getting the banding ,trying to converge the grind of one pJ onto the other. I guess i'll have to try to do a better mechanical set up. But this is just a temporary set up till i can afford a blending processor.

Athanasios

I've never tried to stack or blend. I have enough problems trying to understand and deal with one projector.

I think its a case of the law of distributed misery.

There is only so much misery in the world and you will get your fair share. However if you buy a British sports car, that will pretty much give you your fair share. If you buy two, you will probably not even get another cold.

I think if I wanted a lot more brightness from my CRT, I'd get a high gain screen and suffer with hot spotting off axis.


BTW. I apologize to all my CRT friends for making meaningful substantive contributions to this thread and for not discussing digitals here. Even I can't be perfect all the time. :)

I've never tried to stack or blend. I have enough problems trying to understand and deal with one projector.

BTW. I apologize to all my CRT friends for making meaningful substantive contributions to this thread and for not discussing digitals here. Even I can't be perfect all the time. :)


No problem Mark, we have problems understanding one aging geezer..:p:D

Aging beats the alternastive.

the way I found to remove this horizontal banding is to slightly ajust the bow and its disapears.

the way I found to remove this horizontal banding is to slightly ajust the bow and its disapears.

Tried that still banding. :-(

Athanasios

I tried mikes method for both my PJ my 8000 i was able to get rid of the banding but on my Reference 8(8500 AC) i cant get rid of it. I'm converging the R8's green grid to the 8000's in and up down stack and jst cant seem to get it! tried for a couple hours last night starting over each time. I think its the linerity I'm having problems with and the fact its two different PJ's. Also if i do the long to short i get the jail bars too. I think i'll try next to converge the 8000's to the R8's and see how that goes after i get the R8 not to band. Wish i had two Lumagens in the chain on for each PJ in that it helps to be able to manipulate each image individualy than together as I am now.

You guys are a big help !

Athanasios

Not sure why you would be having this problem with the Reference 8 Marquee, and not the 8000. The 8000 was an earlier model, and would not have some of the later changes to the DPB and HDM/s daughter board that were done to the Reference 8. btw, who re-badged the Reference 8 and what year was it done?

Evidently the bows in the non ultras cause banding more than in the non ultras. Assuming proper set up of center lins, bow adjustment will be needed and shouldn't and on my machine doesn't cause banding. Misuse of converge on green does. A few clicks at the right spots in converge on green will destroy grid geometry in spots but will eliminate banding. You have to start at the new grid points that don't cause banding and work back from there. I can't emphasize that after reinitializing convege on green, go back to geometry and using the controls there to the best you can checking to see that you don't introduce banding there. Don't be lazy. Mark the center of each screen edge and use the center lins to run the center lines through those points. You will have to use skew and key, pin, size and bow. Then adjust the s lins. keep checking the white field to make sure at each point you don't introduce banding. Then be prepared to spend a long time on the random access converge on green. Finally converge the red and blue to the green. I dislike doing all this. Life is too shot to do this instead of watching what one wants to watch, for me. Our CRT chant should be, look at them bands. Just give me a large chunk of time and I can get rid of them. Miss the game? Naw Tivo. Life on hold.

Evidently the bows in the non ultras cuse banding more than in the non ultras. Assuming proper set up of center lins, bow adjustment will be needed and shouldn't and on my machine doesn't cause bandin.Misuse of converge on green does. A few clicks at the right spots in converge on green will destry grid geometry in spots but will eliminate banding. You have to start at the new grid points that don't caude banding and work back from there. I can't emphasize that after reinitialing convege on green, go back to geometry and using the controls there to the best you can checking to see that ou don't introduce banding there. Don't be lazy. Mark the center of each screen edge and use the center lins to run the center lines through hose points. You will have to use skew and key, pin, size and bow. Then adjust the s lins. keep checking the white field to make sure at each point you don't introduce banding. Then be prepared to spend a long time on the converge on green. Finally converge the red and blue to the green. I dislike doing all this. Life is too shot to do this instead of watching what one wants to watch, for me. Our CRT chant should be, look at them bands. Just give me a large chunk of time and I can get rid of them. Miss the game? Naw Tivo. Life on hold.

Good job there. And as Gino indicated, the goal is to not over stress the circuits, which in the case of the Marquee, is very easy to do.

I did get to look at this last night. And I now have a better understanding as to what is going on...

Athanasios problem is the same problem that we have at Williams. You'll not going to get a blend or stack to do this 'IF' the CLM and a few other boards are not identical. Otherwise you're be really stressing the circuits. And banding could become a bigger problem. Then you'll have to get to a compromise on which is more important - geometry or banding.

The analog circuits with their need for inductors gets stressed. If these functions were performed digitally and then converted back to analog . . . .

Not sure why you would be having this problem with the Reference 8 Marquee, and not the 8000. The 8000 was an earlier model, and would not have some of the later changes to the DPB and HDM/s daughter board that were done to the Reference 8. btw, who re-badged the Reference 8 and what year was it done?

I have a newer HDM the older one was smoking :eek: , and the CLM is original i think i have a few here and have changed them out since i messed up one trying to decouple the power rails, you actually have my fried one mike :D.

I think I'll nail it down and was going to try last night but watched 12 monkeys an HD-DVD instead! Like mark said i got lazy ;) . I think i have newer HDM lying around but i have the raster maxed really nicely on the ceiing mounted 8000 and am afraid i'll loose some width with the other HDM. It seams some hdm can go wider than others and i still cant get my R8 to go as wide on the tube face as my good ole 8000. I'll finally get it all settled once i get back from my december trip to europe, i have lots of thing to do to the projectors any how with new belly fans and HD 144 lenses going in too.

thanks guys,

Athanasios

PS on the hdm raster widening i did add the jumper to pi 7 was it? but still no gain and the high and low bandwidth coil adjusts do nothing, maybe a change of cap C50 might do it? can that value be increased a bit or decreased to achieve this? i still have a feeling its limited by newer clm software for protection...maybe?

Athansios

I forgot the details but there is resistor some place that can be removed to widen the raster. TSE told us about it. Perhaps MP or Tim remembers or TSE can refresh us.

I forgot the details but there is resistor some place that can be removed to widen the raster. TSE told us about it. Perhaps MP or Tim remembers or TSE can refresh us.

I thought it was just the jumper added,TSE told us about that?

Athanasios

Jumping the resistoe is that same as removing it and maintaining circuit continuty.

No it wasnt jumping a resistor it was putting a jumper on Pin 7 under the daughter board.

Athanasios

It was I believe putting a jumper between pis 7 and 10 or something like that. I belive that jumped a resistor but I don't remember for sure.. We are taking about the jump points, that's all. I don't remember which points specifically because MP did it for me.

Ahhhhhhhh, jump this. I mean really, it's only a Marquee:p

Chip

?