Color Classic VGA Woes

[Marconelly]

I spotted a "for repair" CC analog board on Ebay a few weeks ago. Seller said that the board was VGA modded and worked fine for a while. Seller then attempted to undo the VGA mod, but something got messed up in the process, and it stopped working. I got this board thinking I could try to repair it, and if not, I could at least keep it for parts.

It arrived the other day, and after careful inspection, I spotted a problem that I figured was the culprit. The tiny J78 jumper was not re-soldered properly. In any case, I wanted to try a VGA mod on this board, so I did all the steps very carefully and properly (including the removal of J78). Then I tried it - and it worked! I got a nice 640x480 picture, and after some picture adjustments, I got to try a few games, everything ran perfectly. I was very careful not to make any adjustments on the CRT itself - all the picture adjustments were only done on the analog board pots in the back.

But one thing bothered me - even though I turned the green sub-brightness pot all the way down, the green sub brightness was still too high. I could see a green background all the way around visible picture, and there was never a proper black color displayed on the screen, it was always green-tinged and visibly bright. Watching the @JDW 's picture adjustment video and seeing the CC service manual, I saw that part of the adjustment process is to tweak the larger PL2 pot which controls the "VG2", whatever that is. So I turned that pot a bit - and green sub brightness became perfect! I was really happy with the result, turned the computer off and left it for the day. The next day I tried it again, and that's when the problems started: The picture was really flickery and unstable, kind of scarily so, considering that there were buzzing sounds coming from somewhere on the board each time the picture became unstable. I turned the PL2 back up a bit to where it was originally, and the problem seemingly stopped. Then I turned it back down to make the picture look the way I wanted, powered the computer off - and the next time it would not power on. The front LED just blinked quickly. So I turned the PL2 back up - and then computer powered on.

I figured that maybe the PL2 not being turned in decades was a but rusty and losing contact, so I sprayed it with some contact cleaner (while the computer was off). Let it dry and tried it again. It seemed to make things better. I could adjust the PL2 and the picture remained stable. So here I am a day later, and the computer no longer powers on, no matter what
The light on the front just blinks quickly, turns off, and that's it. The high voltage never seems to spool up when I flick the switch on the back. I tried more contact cleaner on the PL2, some more on the AB main connector, turned PL2 back and forth - but no luck.

Did I mess something up by spraying the PL2? Am I just experiencing the same issue that the seller was? I guess I will remove the PL2 and test it somehow first. Maybe transplant the PL2 temporarily from another board to see if that's really where the problem is. I hope someone can help, if you've experienced the same kind of problem.

*edit* I've checked the resistance of PL2 pot while it's still on the board, and it seems to work OK - I measure the range between 0 ohm and 1.1 Mohm while I rotate it (if I measure the other two pins it's the opposite range).

 

[JDW]

I truly wish I had the answer for you, but I unfortunately do not. I can only guess it must be something more than just a questionable POT. Did you examine the Analog Board for burned parts? There are a couple diodes that often leave black spots. Some parts on the analog board get so frightfully hot it calls into question the sanity of Apple engineers who created it.

You might want to have a look at the following thread where we talk about burned parts and beefier replacements:

https://tinkerdifferent.com/threads/color-classic-analog-board-trouble.2367/

My CC mentor (先輩) is @Kay K.M.Mods . He taught me a lot about the Color Classic. He might not know what's going on either, but maybe he will share his thoughts if he does.

 

[Marconelly]

I truly wish I had the answer for you, but I unfortunately do not. I can only guess it must be something more than just a questionable POT. Did you examine the Analog Board for burned parts? There are a couple diodes that often leave black spots. Some parts on the analog board get so frightfully hot it calls into question the sanity of Apple engineers who created it.

You might want to have a look at the following thread where we talk about burned parts and beefier replacements:

https://tinkerdifferent.com/threads/color-classic-analog-board-trouble.2367/

My CC mentor (先輩) is @Kay K.M.Mods . He taught me a lot about the Color Classic. He might not know what's going on either, but maybe he will share his thoughts if he does.

Thank you, I will look through that thread carefully. I was thinking of replacing the optocoupler and those diodes, even though nothing looks obviously burned.

I’m worried it’s the flyback though. I can hear a brief buzz from it while the LED is on, and it all shuts off a moment later. Maybe that’s what caused my picture instability in the first place.

I wish there was some way to test the flyback without having to transplant a new one onto the board.

 

[Marconelly]

I've made a potentially significant discovery on the nature of the problem. Once I removed the LC575 board, and placed the original CC logic board, the computer boots fine. It of course cannot display the VGA resolution correctly, so the image is distorted in terms of width and barrel distortion, but it shows the picture with the bootup computer and everything - meaning that I will hopefully always at least be able to salvage the flyback and other parts from this board.

But also importantly, having a stable bootup has let me measure the computer voltages - and they're not looking good. I measure 4.6V instead of 5, and 11V instead of 12. So it's really the low voltage problem in the end, it seems. The board is nearly fully recapped from what I can tell though, maybe just those small 1uF / 50V caps are not replaced, which I don't think would cause this. However, I do notice what looks probably like some green corrosion on the low voltage side optoisolator IC legs. Can that be a cause of this? I'm going to be replacing these ICs in any case.

 

[JDW]

I myself have no idea if a corroded optocoupler would be the source of your trouble. The reason I do not know is because I have only ever owned a single Color Classic, and I didn't have much trouble with it. However, ALL THE CAPS (including the 1uF ones) were recapped when I got it, so that no doubt contributed to a largely problem-free experience for me. I also have never found any corroded parts, although I have found some diodes that got so hot the PCB section directly beneath them had become blackened, and yet the machines still works.

I would recommend posting a photo of the corroded opt-coupler you mentioned.

Sometimes, when you are finding individual components that are "suspect," you need to examine the chip markings, then lookup a datasheet, then determine what pins do what, then check if the chip is getting power, then check if the chip's I/O is functioning.

 

[Marconelly]

I've been reading a bit how optoisolator works, and it does seem that a decaying dLED inside of it could cause a voltage regulation issues, and specifically lower voltage issue. Not much point posting a photo of it now as I"ve mostly scraped the green corrosion from the IC's legs, but some of the legs were fairly covered in green. Wonder if I should also replace the TDA4605-2 ICs?

 

[JDW]

I myself have not encountered any "decaying LEDs" in optocouplers I've used in terms of them no longer working. However, because they are light-based, they do require a fair amount of current — almost always at least 5mA. The lower the current fed into them, the lower the brightness of that light, and at some point it would not be able to switch its internal transistor.

You said you're only getting 4.6v on your 5-volt rail, but I don't see that as being so low it would impact an optocoupler that's not degraded too badly, but I don't have a schematic to see how the optocoupler is getting powered. So if the circuitry feeding it cannot supply at least 5mA, then it probably won't switch its transistor.

SE/30 machines will mostly still work at 4.6V, but going lower than that often will trigger a freeze. But on the Color Classic, I really don't know what the lower voltage threshold is, and not having an Analog Board schematic hinders quick analysis of potential problem areas. Too bad no one ever shipped BOMARC a CC!

 

[Marconelly]

On my other CC, with the mystic board I was getting issues when the voltage dropped to around 4.6V, I think. However, the issue looked completely different - the computer would still power on, but I was getting a death chime. This one won't even power on with the mystic board in place.

 

[JDW]

With its 040 processor and design differences, the 575 board will draw more current than the stock motherboard. So if you have power supply issues resulting in low voltage, it would make logical sense that the stock board may work while the 575 wouldn't.

 

[Marconelly]

Good news! After replacing the corroded optoisolator, the board is working again I'm back in the VGA club!
Now, the voltages I measure are 4.8V and 11.5V. Better than before, but still not quite right. Tomorrow I will replace the other optoisolator, although that other one sits next to the high voltage transformer and it doesn't look the least bit corroded. I will also replace two big capacitors on the low voltage transformer side - the ones that are there now come from the newer Console5 recap kit, and are 5600uF / 16V, rather than 5600uF / 10V. They are also kind of smaller than the Panasonic 5600uF / 10V ones I have, and have used before. I know this voltage and size difference shouldn't matter, but since the voltages are still not quite right, I've got nothing to lose trying this.

 

[JDW]

Fascinating. That is very useful info for Color Classic owners experiencing low voltage issues.

Note that modern electrolytic capacitors tend to be physically smaller than those from the 80's and 90's, even though you have the same capacitance and voltage specifications. However, going higher on the voltage spec is a good idea because it generally reduces the capacitor ESR. Also, in my experience choosing electrolytics, I have found the capacitors with a higher "Hour" rating tend to have higher ESR. It's not going to make a huge difference on the CC's analog board, but lower ESR also reduces heat coming from the capacitors, which is a good consideration in light of how hot that board gets. Even when shutdown, as long as the power switch at the back is flipped ON, the machine draws about 8 Watts continually! If you have a thermal camera, you'll be pretty surprised when you point it at the Analog Board (with the back plastics removed).

 

[Marconelly]

Note that modern electrolytic capacitors tend to be physically smaller than those from the 80's and 90's, even though you have the same capacitance and voltage specifications.

I'm aware of this, but the 5600uF / 10V Panasonic capacitors I have are also new. They just happen to be similar in shape to the ones that were originally on the board. The capacitors from the Console5 kit are thinner but longer, so maybe their overall volume is similar. I'll see what happens, no harm in trying.

Yes, some components on CC analog board stay really hot unless the switch on the back is set to Off. For example, the black square component right in front of the mains power plug is too hot to touch. I also think the whole high voltage section, including those two diodes and the 3300uF / 16V cap are permanently powered and heated up unless the switch is set to Off.

 

[jibsaramnim]

Yes, some components on CC analog board stay really hot unless the switch on the back is set to Off.

Ever so slightly off-topic, but do you or anyone else happen to know if keeping parts of a Macintosh hot/live like this was common in those days, and if so why? I had always sort of assumed that this was just one of the several quirks of the odd little duckling that is the CC, but I don't know enough about the inner workings of Macs from this era to know how common or unique this was.

 

[JDW]

@jibsaramnim
In the USA, "Energy Star" came on the scene in 1992, and although the CC was released in '93. In those days, "low power" isn't what we consider "low" today.

Most electronics draw some power when plugged into the wall, AC Adapters being the gizmos we most often forget about. Thankfully, most of those modern adapters only draw about 0.5W continually, but even 0.5W isn't absolute zero. Had the CC been designed better, it would have drawn less than 8W. Leaving the power switch ON (with the CC shutdown) helps to preserve your PRAM, so even if your motherboard PRAM battery is dead, date/time and other PRAM content is preserved.

The following Energy Star PDF which says that monitors and computers needed to draw less than 30W while in sleep mode to be compliant with the FIRST EDITION of Energy Star

https://www.energystar.gov/ia/products/downloads/CompMonitors_Highlights.pdf

It goes on to say that was lowered to 2W over time. That explains why the 8W power draw of the CC (switched ON but shutdown) wasn't considered bad at all in 1993.

 

[jibsaramnim]

It goes on to say that was lowered to 2W over time. That explains why the 8W power draw of the CC (switched ON but shutdown) wasn't considered bad at all in 1993.

Thanks for the thoughtful response JDW! I get that from a power consumption point of view this wouldn't have considered problematic, especially when it was released. For me I was considering this more from the point of view of seeing several components remain live that get so very hot. I am not sure if all these particular components are necessary to produce the lower voltage(s) necessary to keep PRAM charged, and whatever else they felt like keeping powered. I guess if that's the case it would explain it.

Either way; I keep mine fully turned off unless I use it. Keeping those components constantly cooking just really doesn't feel quite right .

 

[JDW]

I really hate fiddling with batteries and PRAM, to be honest. Back in the day, prior to the internet, having the machine auto-connect and set the date and time wasn't possible. I get that. But having the mouse default to a super slow setting? Having the CRT of the CC default to a super dim setting? And don't get me started on cursor blink rate and menu blinking, all of which drive me completely insane! So when a silly PRAM battery dies, I have to spend an enormous amount of my time getting everything setup right again. And if the battery remains dead and if I flip off that back side power switch, then next time I boot, it must all be done over again. It's something about vintage Macs that has always irked me. But back in the day, I didn't mind it because I always had a good PRAM battery in any Macs I used. It's only when they die that these things become a major issue.

Now I suppose these things are considered trivial by others. But hey, some people don't mind leaving their dirty clothes on their bedroom floor either, but that would drive me insane. So to each his own, I guess.

 

[jibsaramnim]

Oh yes, totally get that. Even with modern replacements like those CR2032 adapters, with us not really being able to use these macs as frequently as we could back in the day, you're bound to run into empty batteries again sooner or later.

I wonder, as your CC is running some version of Mac OS 7.x, wouldn't using something like PRAM Reader or PRAM Auto-Restore serve you well, even if you do usually keep your CC plugged in with the power switch on? I've not yet tried this myself as my CR2032 has so far been doing a good job of keeping up, but from their descriptions they sound like they might make your life a bit easier even if you do end up in a situation where you lost (or had to zap) your PRAM. You're probably already very aware of these tools of course, I just wonder if you perhaps had a reason why these tools aren't useful.

 

[JDW]

Oh yes, totally get that. Even with modern replacements like those CR2032 adapters, with us not really being able to use these macs as frequently as we could back in the day, you're bound to run into empty batteries again sooner or later.

I currently use a triple LR44 solution in my Color Classic Mystic (using a cheap battery holder sold on Amazon). That lasts 75 days when unplugged from the wall or with the power switch at back turned off. The stock battery would have lasted much longer, possible 2-3 years, similar to how long my compact Mac 4.5V batteries typically last.

...wouldn't using something like PRAM Reader or PRAM Auto-Restore serve you well.. I just wonder if you perhaps had a reason why these tools aren't useful.

I have used TechTool Pro's PRAM backup and restore before, so I do know it is useful. I was merely saying that a good PRAM battery saves you the time and trouble of even needing to restore the PRAM from a backup. And even then, you'll need to set the date/time regardless. But yes, you are right to preach the merits of doing a PRAM backup. Thanks for mentioning that!

 

[jibsaramnim]

I currently use a triple LR44 solution in my Color Classic Mystic (using a cheap battery holder sold on Amazon). That lasts 75 days when unplugged from the wall or with the power switch at back turned off.

Interesting. The "MacBatt" CR2032 solution I'm using seems to be lasting much longer than that. Since my CC has been a work-in-progress due to one particular issue it's been having that I've not yet been able to solve, but for the ~1.5 years I've had it so far and with it off most of the time, the MacBatt still seems to be going strong. I wonder why that is, you'd think your solution with three batteries would be able to keep up much longer compared to my singular CR2032?

was merely saying that a good PRAM battery saves you the time and trouble of even needing to restore the PRAM from a backup.

Ah yes, very true, you're absolutely right.

Edit: I realize I'm making this thread go rather off-topic, my apologies! I didn't mean to do that.

 

[JDW]

@jibsaramnim
I measured the Current Consumption of the LC575 motherboard to the PRAM battery to be 63µA @4.5V. Just FYI.

Also, I would say it's a "related deviation" rather "off-topic." @Marconelly mentions having an LC575 board in his CC, and because those boards all have PRAM batteries, our short conversation about PRAM batteries will probably be useful info to whomever finds this thread in the future.