Troubleshooting Apple Multiple Scan 17 Display

[nicktasios]

I recently picked up a Quadra with a Apple Multiple Scan 17 Display. The Quadra was not working, it clearly had corrosion from leaking caps so I replaced them and luckily it chimed and worked properly.

The next problem to tackle was the screen. Sometimes it turns on and off multiple times before finally turning on, and then it sometimes turns purple. After waiting some time, without doing anything, it regains all colors. Another symptom is that it sometimes seems to momentarily blink.

Also, not sure if this is related, but sometimes, after experiencing these problems, the Quadra will not chime. I have to wait a bit before it starts working properly again. I think that without the monitor this doesn't happen, but I'm waiting for a VGA adapter to confirm this only happens with this monitor.

By the way, I confirmed the monitor was the problem by testing the monitor on a different Macintosh, which it again showed this purple coloring, and also by testing the Quadra with a different monitor, which seemed to work fine.

I checked the boards for any leaking caps and can't find anything. Everything looks pristine, with the exception of some dust, and the caps are not bulging. What are some things I can troubleshoot? Could it be the flyback transformer, or do you suspect a bad cap anyway? If I am to replace caps, which board should I start with?

Btw, not sure if this is normal, but the monitor makes a high-pitched sounds when switching resolutions as you can see in this video https://youtu.be/nkRL_RYf5hQ.

[momaka]

The monitor going purple looks like a loss of Green signal.

Does the wacky color (green loss) fix itself instantly... as in instantly switch from the faulty state to the normal state after some time? Or does it take a long time where the color slowly corrects itself?

The former would imply a bad connection somewhere, whereas the latter is more likely to be caused by small caps in the signal path / on the neck board.

That said, the monitor looks like a Sony build (both the PSU and the apperture grille screen), so the caps are all going to be good Japanese brands and likely without any issues. In particula,r the big caps on the PSU should still be OK. Just check to make sure none of them have signs of corrosion (electrolyte leaking from the cap's bottom.) Some early low-ESR series eventually do leak. Nichicon PR are typically the worst offenders... though IIRC, the PL series would also do it sometimes. Panasonic FC is another one from what I've been reading on this forum, though a lot more rare. I don't remember if UCC (United Chemicon) had any series with this particular problem... but again, just check the board(s) carefully.

My first suggestion is to completely take apart the monitor - i.e. pull both the PSU/deflection and neck boards out, then follow up with a thorough cleaning. On the power board, inspect all large components for bad solder joints. And on the neck board, inspect for bad joints on the video connectors. Also, if there are any small electrolytic caps on the neck board that were near heat sources (including very close to the monitor's neck / gun assembly), those caps might be drying out and due for a change. But before pulling out any caps, just re-seat all connectors and the neck board onto the CRT tube a few times, then test the CRT again. Sometimes, the issue could simply be bad contact in a connector somewhere.

As for the noise while switching resolutions - that's pretty normal for some CRTs. I've never used an Apple Multiple Scan, but I do have several Sony CRTs, and some models do squeal / screech while changing the resolution, others don't and just click (relay noise.) Never had issues with any of these, so I assume they run normal. Many of them have served me for many years. I think yours looks/sound pretty normal with the resolution switching... unless you have another Apple Multiple Scan 17 CRT and have confirmed that one doesn't make the same noise when switching between resolutions.

[nicktasios]

Funny that you ask, but indeed, the monitor gets back its color suddenly. I disassembled the monitor completely and inspected both boards thoroughly and I found that they looked superb, shiny pads, no leakage close to caps nor any bulging. Close to the flyback transformer it's a bit darker, I guess from dust and heat. I did clean both boards from the accumulated dust.

I didn't find any cracked joints, but I guess these are often tricky to detect with the naked eye.

If I post some pictures could you point to where the hot areas are, or which caps are likely close to heat so that I can order some caps to try replacing?

PS: This is indeed a Sony Trinitron. Looks like a fine monitor, so I'd like to save it if possible.

[nicktasios]

I posted images of the boards here https://imgur.com/a/VbrZR9i . Below I attached an overview of the two main boards.

The capacitors are mostly Matsushita, with some ELNA, Rubycon, and I think some Nippon Chemi-Con.

[momaka]

Originally Posted by nicktasios

Funny that you ask, but indeed, the monitor gets back its color suddenly.

Then that is almost certainly going to be caused by a bad/marginal connection somewhere. It could be on a connector, or it could be a solder joint.

First, take all of the multi-pin signal wire connectors that run between the deflection board and the neck board, and inspect carefully. Sometimes, it could be just slight surface corrosion on some of the pins to cause this. in which case, reseating the connectors a few times will usually fix the issue. But also inspect where all of the wires are crimped into the female pins for each connector, as that is another weak spot, where the wire strands might have started to corrode and are making poor connections with the metal crimps.

Next, the neck board: you can see the PCB underneath IC001 (LM2205N) has turned slightly darker, meaning this IC probably runs a little hot. My suggestion would be to retouch all of the solder joints on this IC with a bit of fresh solder. See if standard 60/40 solder wicks well and make the joints all shiny. Early lead-free solders can be a bit strange sometimes and not play too well with leaded 60/40 solder... in which case, you might have to retouch the solder joints with some lead-free solder.
While at it, also retouch the pins on IC002, as well as connectors CN307, CN305, and CN304. And finally, retouch the solder joints for the CRT socket. They do look OK from afar in your pictures, but earlier lead-free solder joints from Sony can be a bit tricky and shouldn't always be trusted. If you look at the solder joints of electrolytic capacitor C007 on the neck board (right by IC001), you can see the solder on the positive terminal doesn't look very optimal and could use some resoldering.

If you have more patience/time, add flux on the rest of the small SMD components under IC001 and retouch their solder joints as well. Also retouch C204 and L202 - these are components directly connected to the KG line.

Originally Posted by nicktasios

I disassembled the monitor completely and inspected both boards thoroughly and I found that they looked superb, shiny pads, no leakage close to caps nor any bulging. Close to the flyback transformer it's a bit darker, I guess from dust and heat. I did clean both boards from the accumulated dust.

Yeah, there's always thick dust by the flyback

Originally Posted by nicktasios

I didn't find any cracked joints, but I guess these are often tricky to detect with the naked eye.

Yes, sometimes.

Originally Posted by nicktasios

If I post some pictures could you point to where the hot areas are, or which caps are likely close to heat so that I can order some caps to try replacing?

Like I mentioned, most of the caps should still be OK, despite their age. In particular, the larger electrolytics on the PSU / deflection board should still be fine. Only the small caps on the neck board might be getting closer to EOL, as the neck board usually tends to run hot from the RGB amps and the tube heater itself. But before you try replacing anything, see if fixing / reflowing the solder joints mentioned above fix the issue.

If not, then you can try changing all of the small caps on the neck board. In particular, you can start by checking C101, C201, and C301 as these are your RGB signal (in that order too) coupling caps. Since you seem to have a loss in green, check C201 (and R203, since that is the next component downstream for the Green input.) Apart from these caps, also worth checking would be C062, C105, C205, C305, and C007. But again, before doing anything with these, first try the monitor with the reflowed solder joints. Good chance that should fix the problem. Bad caps in the signal part would usually cause only a deterioration in the signal line and not snap back instantly to normal when the monitor heats up.

Originally Posted by nicktasios

PS: This is indeed a Sony Trinitron. Looks like a fine monitor, so I'd like to save it if possible.

I'm with you on that one - if you can / have the time, it's certainly worth saving.

BTW, thank you for providing such clear & well-focused pictures. Makes troubleshooting a lot easier.

[nicktasios]

I reflowed the solder joints as you suggested, and then some, but unfortunately I still get a purple screen. Not sure if it's related to the reflowing or not, but now it doesn't seem to get its color back and convergence somehow became worse (look attachment).

I then went ahead and replaced C062, C201, and C007, since I had replacements at hand, but unfortunately that didn't have any effect.

Maybe worth mentioning, the monitor has adjustments for horizontal and vertical convergence. Horizontal adjustment seems to work, but vertical produces little to no effect.

Any other suggestions? Or should I just try to replace all caps? Something tells me it's not a cap, but what else could it be?

[nicktasios]

I made some progress! The nice thing with these neck boards is that in principle the red green and blue parts of the board should actually be clones of each other. So I put my multimeter in diode mode and checked the pins R.In, G.In, and B.In, and found that R and B were reading the same (0.66V), but G was different (0.86V). This leads me to believe that there is a faulty component somewhere along the green part. Following the path with my multimeter, I end up at IC001 (R and B reading 0.56V and G reading 1.28V). If I'm understanding the PCB correctly, all RGB inputs to the IC read the same value in diode mode (1.28V), which means that IC001 might be faulty!

I attached a picture of the IC. Could you help me identify it and perhaps find out if I can buy a replacement?

[nicktasios]

I went ahead and removed IC001 from the board, and (un)fortunately it was not the component causing the problems, so I went ahead and removed IC002 too, but again, the difference in the diode readings remained.

The remaining circuit for each of the R, G, and B is not that big, so I mapped them. In the attachment I show two separate schematics for the R and B parts. You see that some parts are the same, and in reality, the RGB parts are connected at these points (note that there are some other capacitors connected between the double diode and ground but I left them out of the schematic). The part of the circuit that is different for each RGB component is small, and I don't see any visible faults. I have to add that after working on the board for a bit, the G.IN reading in diode mode sometimes gives the expected value, but if I wait a bit it suddenly jumps from 0.683V to about 1V, and after a bit it snaps back to 0.683V. The R and B.IN give 0.673V. Weirdly enough, the B.IN sometimes doesn't immediately read 0.673V, while the R.IN does. So perhaps the same component in the blue part is also starting to fail/age? This intermittent behaviour matches the observation of the color suddenly coming back.

As there are basically no electrolytic caps in the faulty area, what other components could show such behaviour? Could it be the transistors Q104,Q204,Q304? I guess it's either that or the double diodes D109, D209, D309.

[nicktasios]

So, this becoming more of a repair log, but hopefully someone might benefit from it in the future.

After the last post I spent like 12 hours trying to figure out what the problem was. I removed the transistor Q204 on the green line and then got the expected readings on D209, so I thought that the transistor was the problematic part. In diode mode it was reading fine so I decided to swap it with the one from the red line, Q104. Alas, the problem was exactly the same. I then swapped D209 with D109, but again the same. Just to be sure, I also swapped the electrolytics C105 and C205 which are the only other set that is RGB specific. No effect here either.

By sheer luck, at some point, as I was heating the board to reflow one of the transistors, I noticed that the diode mode reading that was different on the green line was slowly moving to the expected value as the board was heating. I though it had to be something on the other side of the board because I checked virtually everything on that side, so I turn it around found out that pushing one of the resistors R220, I think it was, made the diode mode reading be the correct value. This resistor is really close to the heatsink of IC002 (VPS07S) so it could be that the solder cracked due to the temperature? Anyway, I reflowed it and also the one on the blue line, R320, which was also near the heatsink, just for good measure. To my surprise, the diode mode reading now stayed at the correct value. I was convinced I solve the problem, so I was pretty disappointed when I turned on the CRT again. The problem got even worse!

In the end, I had the feeling that the diode mode readings would give me something if some components was faulty, and everything was reading the same, so I suspected that IC002 was failing. To test the hypothesis, I removed IC002 once more and hooked it up to the board using wires like you see in the picture. This allowed to perform a couple of tests:

1) Remove G.OUT and B.OUT and connect them to R.OUT. This gave me a grayscale image, and it means that the output part of the board should be fine.
2) Without undoing (1) Remove R.IN and B.IN and connect G.IN to R.IN of the IC. This again game a grayscale image which confirmed that G.IN comes fine into the IC.
3) Same as (2) but instead of G.IN, connect B.IN to R.IN of the IC. Same result.

I think that the tests are pretty conclusive that IC002 is faulty! So I ordered a spare of the IC, LS07S. Unfortunately it's not easy to find those nowadays, so I ordered LS07T, which hopefully is the same. When I receive it, I'll update with the result!

[momaka]

Originally Posted by nicktasios

So, this becoming more of a repair log, but hopefully someone might benefit from it in the future.

Sorry, for the delay - I haven't neglected the thread yet! Just been busy doing some "interesting" changes to the plumbing of a 90 Y.O. multi-family home, so wasn't available the last few days.

Originally Posted by nicktasios

Any other suggestions? Or should I just try to replace all caps? Something tells me it's not a cap, but what else could it be?

Agreed.
At least going by the way the symptoms change, it appears the issue is not capacitor-related... but we still have to leave our minds open. I think a recap might be more of a last resort idea to try, though I also don't expect this to fix the issue.

Originally Posted by nicktasios

I made some progress! The nice thing with these neck boards is that in principle the red green and blue parts of the board should actually be clones of each other. So I put my multimeter in diode mode and checked the pins R.In, G.In, and B.In, and found that R and B were reading the same (0.66V), but G was different (0.86V). This leads me to believe that there is a faulty component somewhere along the green part. Following the path with my multimeter, I end up at IC001 (R and B reading 0.56V and G reading 1.28V). If I'm understanding the PCB correctly, all RGB inputs to the IC read the same value in diode mode (1.28V), which means that IC001 might be faulty!

OK, looks like you're on the right track here with checking components on the green line. However, I'm not yet convinced that IC001 is faulty. The issue is, IC001 is an "active" type of circuit with transistors in it, so you can't really test it in diode mode. You can check components attached to the R, G, and B pins (both on the input and the output), but once those are verified, you really can't test much else without an oscilloscope. In particular, you need to check the RGB signal going into and out of IC001. If you don't have an oscilloscope, the next best thing we can do is just continue checking components after IC001 on its RGB output or back from the KR, KG, and KB pins on the CRT tube.

Originally Posted by nicktasios

I attached a picture of the IC. Could you help me identify it and perhaps find out if I can buy a replacement?

So I looked for a datasheet the other day and really couldn't find anything (though I will admit that looking up obscure datasheets isn't my strongest side.) That said, I did dig through the pictures and service manuals for several monitors that I own and the best I can come up with is that IC001 (LM2205N) must be somewhat close to a LM1203. From what I could identify from your neck board pictures and the LM1203 datasheet, at least the R, G, and B input pins are the same number (pin 4, 6, and 9 for R, G, and B, respectively) and ground (pin 7). So there's a chance these might be cross-compatible... or at the very least, perhaps verify with the LM1203 datasheet if your LM2205N IC appears to be connected similarly.

Here is a link to the LM1203 datasheet:
http://cdn.goldmine-elec.com/datasheet/A10692.pdf

Originally Posted by nicktasios

After the last post I spent like 12 hours trying to figure out what the problem was. I removed the transistor Q204 on the green line and then got the expected readings on D209, so I thought that the transistor was the problematic part. In diode mode it was reading fine so I decided to swap it with the one from the red line, Q104. Alas, the problem was exactly the same. I then swapped D209 with D109, but again the same. Just to be sure, I also swapped the electrolytics C105 and C205 which are the only other set that is RGB specific. No effect here either.

Excellent troubleshooting work there! Indeed the RBG lines should have similar component layout for each, so I think we are on the right track here. I did some further looking into this, which I will share a little further below.

Originally Posted by nicktasios

In the end, I had the feeling that the diode mode readings would give me something if some components was faulty, and everything was reading the same, so I suspected that IC002 was failing. To test the hypothesis, I removed IC002 once more and hooked it up to the board using wires like you see in the picture. This allowed to perform a couple of tests:

1) Remove G.OUT and B.OUT and connect them to R.OUT. This gave me a grayscale image, and it means that the output part of the board should be fine.
2) Without undoing (1) Remove R.IN and B.IN and connect G.IN to R.IN of the IC. This again game a grayscale image which confirmed that G.IN comes fine into the IC.
3) Same as (2) but instead of G.IN, connect B.IN to R.IN of the IC. Same result.

I think that the tests are pretty conclusive that IC002 is faulty! So I ordered a spare of the IC, LS07S. Unfortunately it's not easy to find those nowadays, so I ordered LS07T, which hopefully is the same. When I receive it, I'll update with the result!

Sounds good, but in the mean time, I looked at the service manuals of several different Sony CRT monitors to see how they designed the circuit after the RBG output amp (IC002 in your case), just to see if anything pops out. Going by the Dell D1025TM and Sony CPD-E230 service manuals, I came up with the following:

The R, G, and B lines from the RGB amps of these monitors go through several components before terminating at the KR, KG, and KB pins at the back of the CRT tube. Namely, the KR, KG, and KB outputs are pulled up / biased towards the +B power line (around 180V for older/smaller? Sony CRT monitors, it seems) with one or two high-resistance resistors, typically close to 1 MOhm total resistance. In your case, those resistors appear to be R107, R207, and R307. Meanwhile, the RGB Cut-Off amp pulls down these same KR, KG, and KB lines down to around 70V DC (66-68V for the CPD-E230, at least) through three diodes (one for each line.) In the case of your monitor, those diodes seem to be D105, D205, and D305. Unfortunately, that's about as far as I can trace back towards the cut-off amplifier. Actually, I don't see a dedicate Cut-Off Amp IC, so I suspect it might be done with discreet transistors or perhaps through an IC somewhere else?
Whatever the case above, my suggestion is as follows:

1) First check the values of resistors R107, R207, and R307. Might have to take them out of circuit, as their resistance is quite high and likely won't be possible to measure in circuit. I suspect these should still be OK, but still worth a check.

2) Next, check diodes D105, D205, and D305. If you feel experimental, perhaps swap these around with each other to see if the faulty/dim signal follows one of these (the one on the green line, D205.)

3) Check the DC voltage, (CAREFULLY!) at the KR, KG, and KB lines at the back of the CRT neck board with the CRT powered On. While at it, also check the output voltages from the PSU, just in case.
--- 3a) Namely, check the value of the +B line, 75V line, and +12V line at connector CN305 (to avoid shorting any pins together, test at the pad of a component somewhere that connects to one of these lines - e.g. resistor R316 on the side connected to the +B line for the +B value.)

4) De-energize / turn off power to the monitor, and try tracing back to where the cathodes of D105, D205, and D305 go to. I suspect they might also be going to the main board / D board though connector CN303 on the neck board (I see some labels "R CutOff", "G CutOff", and "B CutOff".) If they do, we might have to trace where these go to on the mainboard.

5) Check and/or try swapping capacitors C104, C204, and C304 capacitors between each other. These couple IC002 (VPS07) RGB output lines to the KR, KG, and KB lines. See if the problem in the image changes when you swap these around - that is, the lack of green changes to a lack of blue or red instead.

6) IC002 RGB output should also have a pair of clamping protections diodes for each line that limit its DC value to be between ground (0V) and 75V rail voltage (75V DC.) On your monitor, these diodes appear to be D103 and D104, D203 and D204, and D303 and D304. Check the resistance of each of these diodes out of circuit (forward and reverse) and see if all of their readings match or are fairly similar. We are looking for a leaky diode here. I suspect these would also be fine. But if you don't mind doing a bit more soldering, you can swap these around with each other to again see if this changes the lack of Green to a lack of Red or Blue.

And that's about all of the tests I can come up with for now. Maybe try them before the new replacement for IC002 comes. I haven't had a chance to check yet if the replacement you got is compatible or not. So just in case, feel to do any of the above tests prior to swapping the IC.