[Solved] Trouble with a Sony PGM 2950E

[SuperSpongo]

Hey guys,

I recently bought a Sony PGM 2950E 31kHz monitor which has a defect.
Upon turning the monitor on, it tries to display the picture, but struggles in the vertical direction. After a couple of seconds it switches into a fault mode and the standby LED starts blinking. There's no blink code, it just blinks repeatedly for as long as the set is switched on.
I have uploaded two videos of the behavior:

https://www.youtube.com/watch?v=VKxQdRnIBhI
https://www.youtube.com/watch?v=6QDw3oBDKh0

Apologies for the crackling sound, that comes from my camera.

I assume it is a power issue, maybe a cracked trace or an old capacitor?
Since this is an older, rather obscure model, I cannot find any documentation, let alone a service manual for the set. Can anybody help me out?

I had a look at the manual for the standard PVM 2950 and found an IC which is also on my board: STV9397 which is a vertical deflection booster. Am I on the right path here or should I rather have a look at a voltage coming out of the flyback?

Any input would be appreciated.

[SuperSpongo]

Pity that no one seems to be able to help.

I measured the chip in question and found out that the output matches the waveform described in the pvm-2950's manual. When the signal gets distorted, it can be seen on the scope

https://youtu.be/DsWCZx54q5Q


I went on to measuring the input, the signal is very weak. I went further back and discovered that the input signal is generated on the DX Board, but on Pin 1. This is unfortunate because the DX board is soldered to the A Board with 50 pins and also because the PVM's service manual specified another pin.

So I might have to desolder the board and see if I can find anything.

[gray117]

I don't think you can put this on the vertical deflection. I think it might well just stay on if this was the problem, just with a fcked image... the pvms 15khz used to I think, but then again perhaps the pgms are just different (sorry never really messed with these).

Think the power idea would be the best starting suspect... Just guesses though. And parroting myself that's a shit load of caps to start guessing about. And, again caps are the place to start but don't overlook resistors in case somethings loose/blown.

Nice monitor though, love the big boxes, good luck

[xga]

The PVM-2950 is notorious for having issues with dry solder joints, so I wonder if your PGM-2950 might suffer with the same issue, being from the same time frame. See if giving the monitor housing a few firm taps whilst it powers on makes any difference. It might just be a case of re-flowing solder on some of the components that tend to vibrate a lot like the flyback and transformers.

[SuperSpongo]

Thank you for your responses!

In the meantime I removed the metal shielding of the DX Board. All but two boards can be unplugged from the main board. The DX board of course is soldered in with 50 joints...

I all but ruled out the G board (power supply), as it provides the proper voltages. I also checked voltage regulators for 9V, 12V and 5V, they all seem fine.
So maybe it is something on the flyback, I don't know. I'll try to trace the power supply of that IC a little more.

The solder joints definitely look bone-dry all over. I tried tapping the case while turning it on but that did not do the trick.
I just recently heard about the high-voltage solder-joints being prone to breaking due to vibration so I might reflow the flyback and all other big transformers and give it another whirl.

Problem is that I have to fully reassemble it each time to test as the chassis is so damn crowded.

[SuperSpongo]

So I resoldered all the transformator pins without any luck. I also did a little more research how the vertical deflection is even done. In general, there has to be a sawtooth signal with a frequency of 60Hz somewhere on the board. This signal is generated by an IC which is often used in a multitude of monitors and TVs. Indeed I found an "LA7856" Chip on one of the boards. I reflowed the solder joints on this board as well, unfortunately without any results.

I will try to go further from here. I will measure the sawtooth signal and hope that it looks distorted on the scope. Then I'll try to replace some caps on that particular board.

[MKL]

If it's like the PVM2950, the voltage for the vertical IC (pin 2) is 18V and is supplied by the flyback (pin 9). You need to check if it's present and stable. I suggest to solder a piece of wire to pin 2 of the IC and attach a croc clip to the other end as meter probes can slip and short nearby pins if you hold them with your hands. If there are problems with this line, replace the filter caps (C566 and C582 on the 2950 schematics) and check the diode rectifier D513 and R577.

[SuperSpongo]

Thank you for your input!

Okay, I had some progress over the weekend. Sadly I'm not done yet
I measured the sawtooth signal and it gets distorted, once the picture errors start occurring. I made a video:
https://www.youtube.com/watch?v=6Aki4nJSAr0

So I bought some replacement caps for the DX board.

First, I tried to take out the DX board. 52 pins, soldered in the 90s. I tried desoldering with flux and solderwick and then heating all pins up with a hot air desoldering station while pulling on the card. It did not move one bit.

Then, I started desoldering the capacitors on the DX board, checking each with a capacitance meter and to my joy, the caps were indeed faulty and leaking charge. After tediously replacing 8 caps, I put the monitor back together and fired it up. The result: the picture is stable for longer, but eventually starts crapping out still

Next up is checking the signals again and, as MKL suggested, I might solder some probe wires to the pins to make measurements easier.

[SuperSpongo]

If it's like the PVM2950, the voltage for the vertical IC (pin 2) is 18V and is supplied by the flyback (pin 9). You need to check if it's present and stable. [...] If there are problems with this line, replace the filter caps (C566 and C582 on the 2950 schematics) and check the diode rectifier D513 and R577.

Oh, and which IC did you mean here, MKL? The LA7856 chip does not have supply on Pin 2 and flyback voltage on Pin 9, that's why I was asking. Maybe I missed one?
I'll check out the service manual of the PVM once again.

EDIT: Ah, you meant Pin 2 of the IC and Pin 9 of the Flyback. Found the components on the PVM's schematics. I'll check those out when I'm home, thanks!

EDIT2: This is my DX Board now. It's very different from the PVM's, unfortunately.

Spoiler

[SuperSpongo]

https://www.youtube.com/watch?v=vmF6hKOpC5Q

It's working again! I wasn't able to test it for long, but I'll do that today.

In the end, I think it was a cold solder joint. I'd like to thank MKL once again: I soldered some probing wires to 3 spots of interest on the board and I think that one just happened to be the cold solder joint causing the problems. So the fix was a combination of luck and knowing where to look.

Let's hope that it stays working. It's a little unsatisfying that I don't know what exactly caused the problem. I was right in my initial assumption: the STV9397 vertical deflection booster had something to do with it. The faulty solder joint is located at the output of the booster IC. What I find a little baffling is that when this booster signal gets distorted, it also distorts the sawtooth impulse on the DX board. I get that they all work together and that they operate in a feedback loop, but I thought that the signals were somehow decoupled from one another (via buffer transistors) so that you at least have the chance to somehow narrow down the affected area...

Anyway, I'll do some more extensive testing tonight.

[gray117]

lol congrats