SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

[nick3092]

Hello all. I have a 21" Sony Trinitron CRT monitor I'm hoping to repair. It's technically an SGI monitor, model GDM-5011P. I showed the attached pictures to a couple people who are slightly knowledgeable in monitors and/or electronics in general. And the general consensus is that it's an issue with the deflection board, or less likely the power supply. Most likely the deflection board though, as it looks like an issue with the horizontal sync. I'm hoping that it just needs a recap to fix it. The service manual complete with schematics, block diagrams and parts lists is available to view or download at https://www.manualslib.com/manual/99...Gdm-5011p.html

I've removed the deflection board (after ensuring the CRT was discharged and removing the anode cap from the tube), but don't see any visible signs of bad caps on the deflection board. Nothing bulging or leaking. But I know that caps can still go bad without exhibiting those symptoms as well. It looks like Sony primarily used good caps. The majority are Chemi-Con and Rubycon, along with a couple of ELNA's.

I'm mainly looking for advice on if this seems like something that could be caused by bad caps (or something else easily fixed). And if it is the caps, advice on selecting suitable replacements. Along with if I should pursue the deflection board (more likely) or power supply (less likely).

Here it is at the desktop:


And to rule out the PC itself, you can see it distorting the OSD with no signal:


Looking at Digi Key/Mouser, most of the caps are available in exact replacements (ie. Chemi-Con KMG series), or there is a designated replacement for older ones (ie. Rubycon KY replaced by PK series). Personally, for other projects I usually tend to stick to Nichicon or Panasonic. I also generally try to stick to only 105c caps for longevity. About half the caps on the board are 105c, and the other half 85c. So It would be nice to replace the 85c ones with a better cap, if possible. So I don't know if I should stick with these, or there is a better suited replacement.

Chemi-Con KHA x2
Chemi-Con KMF x1
Chemi-Con KMG x16
Chemi-Con LLA x1
Chemi-Con SMG x11
ELNA No Marking x1
ELNA RE3 x3
ELNA RJ4 x1
Rubycon YK x23
Rubycon YXA x1

And here are some general notes of mine:

YK replaced by PK
KHA replaced by KXG
KMF replaced by KXG
LLA = low leakage (doesn't seem available, and no replacement listed)
The one ELNA has no series marking, but has the same numbers written on the back as the RE3s do: 9849 85c

Thanks in advance for any info. I really want to get this monitor working again, as I more or less built my MAME cabinet around it. So throwing in any other display wouldn't fit well. And its been dead in the water for about 6 months while i was trying to find someone local to look at it. Which I couldn't find anyone.

-Nick

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

It looks like it could be poor filtering on one of the vidio drive supply voltages, I would check C485, C459,

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

So you are thinking the issue is on the A/neck board instead? I was able to check those two caps in circuit with my ESR meter (Anatek Blue, in circuit).

C459 22uf 100v Rubycon YXA .34 ESR
C485 10uf 250v Chemi-Con SMG 4.3 ESR

Depending on which ESR chart you look at, C485 could be too high. But I'm not sure that it is. The SMG data sheet is readily available. But I've honestly never been able to correlate my ESR readings back to a data sheet. I just know if a value looks abnormally high, it's a problem. I'm also not sure if these can be read in circuit.

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

It seems like the general consensus of people I've talked to say this is a H sync issue. So I've decided to give recapping the deflection board a try to fix this. As this board is a bit of a pain to take in and out, I'm just going to shot gun replace everything and hope for the best.

Not much feedback here, but I figured I would document what I did Incase it works, and anyone else comes across the same issue. This is the approach I took for selecting new caps.

Also, this is my first attempt at trying to match up new caps to the old. Most of the stuff I have worked on in the past wasn't picky about the types of caps used. So I welcome any feedback in my choices. In most cases where the cap was 85c, I tried to find a 105c version instead.

• Chemi-con KHA (2): Couldn't find a data sheet, but these seem to be high ripple caps. Replaced one with Panasonic EE (33uF/250v), the other with Nichicon UPJ (2.2uF/250v).
  
• Chemi-con KMF (1): couldn't find a data sheet, but they replaced this series with KXG (33uF/400v). So I went with that.
  
• Chemi-con KMG (16): This is still a current series and a 105c cap, so replaced it with the same.
  
• Chemi-con LLA (1): Went with a Nichicon KL for low leakage (2200uF/10v).
  
• Chemi-con SMG (11): This is a current series, but the 1uf caps are out of stock till July. Looks like the SMQ is basically the same specs. Used that for the 1uF/50v and kept SMG for the 4.7uF/50v caps. I forgot to try and find a 105c replacement.
  
• ELNA unmarked (1 - 100uf/16v) & RE3 (3 - 220uf/16v): Replaced these all with Panasonic FR.
  
• ELNA RJ4 (1): Replaced with Panasonic FC (2.2uF/50v)
  
• Rubycon YK: They list the PK as the replacement for YK. But I decided to upgrade these to 105c. I used Panasonic FC for 21 of them (multiple values), and FR for the other two (1000uF/25v).
  
• Rubycon YXA (1): Went with Panasonic NHG (22uF/100v).

The caps should be here tomorrow, so hopefully I can get the job done this weekend and have an update on the effectiveness of the deflection board recap soon.

[stj]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

nice arcade backdrop,
the woble is caused by unstable voltage to the deflection coil,
first suspects are caps smoothing the B+ rail on the main board.

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

nice arcade backdrop,
the woble is caused by unstable voltage to the deflection coil,
first suspects are caps smoothing the B+ rail on the main board.

I work on old tube radios, and in that world B+ means the high voltage supply. I'm assuming that means the 200v rail in this monitor. So you are suggesting C671 and C672 on the power supply board (G board in the manual, page 28 in the PDF?

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

After recapping the deflection board, I had no improvement in the picture. I don't think it looks any worse, so, that's a plus. I guess next I move on to recapping the power supply board.

The only cap I may have trouble replacing there is C611 (330uF 450v, can). It uses an odd "L" shaped configuration for the terminals. Looks kind of like this: ( | _ ). As where most can caps I've found at DK/Mouser seem to use a two prongs either 180 degrees apart, or the run parallel to each other.

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

I would check the caps in the power supply like the +80 volt supply and +200v supply, You don't need to replace the caps, you can temporarly solder one across it to see if it makes a change. IE; C672, could be bridged with a another 2.2~3.3µf 250v if that does'nt help, unsolder it, disscharge it and solder it across C674
It could also be caused by one of the other supplies, a scope would find it easily.

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

I have a scope. But poking around in a live CRT chassis with it makes me a bit nervous. I did try bench testing the supply, but since it's a power saving monitor, it goes into low power mode on the bench.

There aren't that many caps on the power supply. I have no problem with replacing them all. If one is going/gone, others could be also on their way out. Or not, maybe they could run another 20 years. Who knows? But if they all get replaced now, the odds go down quite a bit of another failure in the near future. Plus it's easier to work on them all at once, instead of doing one, putting it back on the metal bracket, reinstalling the metal bracket to the cage only to find out it needs to come out again.

Replacing the small ones should be no major task. It's just the big can cap I might not be able to find a suitable replacement for due to the terminals.

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

The only part you need to stay away from with the scope would be the primary side of the power supply, and any of the high voltage around the highvoltage transformer.

What is your scope model?

You should be safe checking for any high frequency ripple on CN671 and CN672 on the power supply plugs. If the voltages look clean, then there is no need to replace the caps. It is likely the power supply is ok and the problem is on the A board
I would then scope the outputs of the +5 & +12 volt regulator ic's on the A (video) board

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

I have a Rigol 1054z with a 100x probe. I mostly bought it to help troubleshoot tube radios. I'm not very good with scopes in general though. So I usually need some pretty explicit guidance on how to set it up.

I gave a shot at taking some snapshots off the power board. Not sure how good or useful these are. If something isn't right, let me know what to change on the scope and I can try again. I clipped the ground from the probe to the cage, as I confirmed with my DMM it has continuity to the GND pins on the power board connector. I pulled the spring loaded clip attachment off the tip of the probe, and used it with nothing on it. The needle point on the bare probe was able to reach inside the plastic connector to take a reading. All readings done from CN671.

-15v rail (measured -14.5v)


+15 rail (measured 14.5v)


+80 rail (measured 78.2v)


+200 rail (measured 199.3v)

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

that amount of ripple is not that bad, You are not looking at the dC value, you are looking at the AC ripple. I would now check the A board, First check the regulator outputs, If there output cap goes bad they can oscillate, I would also just check all the electrolytic caps with the scope, If the cap is used as a filter, and there is lots of noise/ripple on it, it is likely bad.
I would check around IC402 & ic403 since both the video and osd show the problem.

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

Thanks for the help so far. I took a couple measurements off the A board on IC406 and IC408, the 12 and 5v regulators. I don't think I see any major ripple here, but then again I'm not expert at this stuff. I did adjust the scales here to try and get a better look. All other settings should be the same.

IC406 (12.08v)


IC408 (4.9v)


Are there any particular pins or caps that I should look at around IC402 and 403 I should focus on trying to get a reading?

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

what you can do is check the video signal, If you are using the vga input, check the R,G,B input right at the jack, CN402, to see what the signal should look like, (if you change the sweep rate, you will see a full field, or a single scan line) then check the output of ic401, pins 27,30,35, if it looks ok (likely different amplitude), the check the output of ic402, 37,42,51

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

Hmm. I'm not sure I'm doing this right. I can't seem to get a stable wave form on the screen. Not sure if I'm supposed to though. For now I only did the red channel, before I go further with a bunch of useless/messed up screen prints.

IC401 can't really be scoped directly, as it is a surface mount IC, and it's facing the cage. There is no way I can get in there to probe it. So I figure next best thing would be the in pins on IC402.

Here is what I scoped for the red line:

CN401 - Pin 1


IC402 - Pin 17


IC402 - Pin 37


I apologize if these aren't right. Like I said earlier, I'm not all that great at using a scope. It's not helping I'm doing this on the floor as its the only place I have enough room to work on the chassis with a scope and laptop attached. Plus I'm one handed for making scope adjustments. Due to the fact I can't actually clip the scope probe, I have to manually hold it on the pin.

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

which input is connected to the pc? I guess it was not the 15pin vga connection and was there video on the screen when you scoped the pins?

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

I had the 15 pin VGA hooked up to my laptop. The laptop was at the desktop screen. After work tonight I'm going to try resetting the input on the scope, make sure I get a nice clean waveform on the test terminal, and recheck the monitor. I'll post back once I've had a chance to do that.

[R_J]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

well that would explain the first waveform, you checked the wrong connector, CN402 is the one for the 15pin vga. I am starting to think the video portion may be ok. and that it is the horizontal sync that is being effected more than the actual video.
so it might be an idea to check CN511 on the D board (horz sync should be on pin 3 or 5, which ever pin it is on, it then goes to ic013 and then into ic007. there are a lot of small electrolytics in that circuit, I would scope each cap. if the caps negative is connected to ground, there should be no real noise on the positive.
Have you tried changing the screen resolution to see how that effects the picture, a different resolution uses a different horz frequency.

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

Looks like that was a typo on my part. I was actually probing CN402 for the vga connector. I'll still check again after work tonight to try and get some kind of stable waveform.

I'll have to see later if CN511 is accessible on the D board. It looks like it's down at the bottom and would not really accessible when the monitor is assembled and running. Scoping the caps on this board will be virtually impossible. The board is mounted to the metal cage with no access to the solder side of the board. Also, I recapped every electrolytic on D board with no change in the symptoms.

Not to say there couldn't still be an issue on the D board. Just much less likely it's an electrolytic on the D board causing issues.

[nick3092]

Re: SGI GDM-5011P CRT Monitor (Sony Trinitron) Help

OK, this time I used the auto function on the scope when taking measurements. I believe that does what it can to give you the best picture of the waveform instead of relying on me to try and set it right.

I retook the red pins from the A board. And I was able to open the cage and swing out the deflection board. I was just barely able to get to CN511 this way. I took pictures of pin 3 and 5. Hopefully these make more sense than my last ones.

CN402 - pin 1 - red


IC402 - pin 17 - red in


IC 402 - pin 37 - red out


CN511 - pin 3 - c_sync_1


CN511 - pin 5 - c_sync_2