Harmful to run SE/30 temporarily without recapping?

[djc6]

I have an SE/30 whose logic board is in decent shape - considering its age. The Original battery was still intact (and now in the trash) so just some electrolyte corrosion.

My question, is it really harmful to run it without replacing capacitors? Capacitor C5 had come off completely.

Why would I want to run it this way? I was to do some troubleshooting to see what chip(s) are causing lack of vertical sync. So I can order the caps and other parts at the same time to save on shipping. I suspect one of the SOIC chips in the video circuitry, their legs are corroded. I wanted to run the machine long enough to probe around with my oscilloscope to hunt down the bad IC. I get a happy mac chime via the headphone jack, so the CPU is executing code and I believe this means the RAM is OK?

I am bench testing the logic board with a CayMac Vintage SE/30 Compact Analog & PSU Replacement Board and a PicoPSU. So I am not using the original power supply or analog board.

 

[Nixontheknight]

I have an SE/30 whose logic board is in decent shape - considering its age. The Original battery was still intact (and now in the trash) so just some electrolyte corrosion.

My question, is it really harmful to run it without replacing capacitors? Capacitor C5 had come off completely.

Why would I want to run it this way? I was to do some troubleshooting to see what chip(s) are causing lack of vertical sync. So I can order the caps and other parts at the same time to save on shipping. I suspect one of the SOIC chips in the video circuitry, their legs are corroded. I wanted to run the machine long enough to probe around with my oscilloscope to hunt down the bad IC. I get a happy mac chime via the headphone jack, so the CPU is executing code and I believe this means the RAM is OK?

I am bench testing the logic board with a CayMac Vintage SE/30 Compact Analog & PSU Replacement Board and a PicoPSU. So I am not using the original power supply or analog board.

if a capacitor came off, DO NOT RUN IT, get it recapped as soon as possible

 

[François]

I don’t think it’s really harmful to run it, C5 is part of the audio circuitry, so you just won’t have sound. The problem is not the missing capacitor, but the leaked electrolyte from failing capacitors that is eating, and thus destroying, the circuit board. So better remove the remaining electrolytic capacitors, and thoroughly clean the board, checking for potential damage, before replacing the caps.

I think we all have run computers with failing capacitors! I sure have run my SE/30 in such a state: when the sound coming from the internal speaker is at a very low volume, even at the maximum setting, is when you know that the capacitors have failed and need to be replaced.

 

[djc6]

I don’t think it’s really harmful to run it, C5 is part of the audio circuitry, so you just won’t have sound.

Thats how I felt, but I see many DO NOT RUN posts when I searched here and other forums, without reasoning why. I thought if C5 was missing the circuit is open and power is not getting to whatever would be powered. I thought the capacitors on the logic board were to smooth out voltages, and I'm not using the original power supply but instead a modern PicoPSU while bench testing.

The problem is not the missing capacitor, but the leaked electrolyte from failing capacitors that is eating, and thus destroying, the circuit board. So better remove the remaining electrolytic capacitors, and thoroughly clean the board, checking for potential damage, before replacing the caps.

I did this on another SE/30 and it worked out well. This is my second SE/30 restoration. I'm still looking for good ideas for cleaning up corrosion around these SOIC package chip legs in the video circuitry. One my other restoration a little vinegar worked well - kept it on the board only for five minutes or so and only in the affected areas. Any other ideas or is vinegar a safe bet?

 

[François]

In case you didn’t know about it already, the schematics for the SE/30 logic board have been redrawn using modern software! Get them on github. They may be useful for troubleshooting…

 

[phipli]

Thats how I felt, but I see many DO NOT RUN posts when I searched here and other forums, without reasoning why. I thought if C5 was missing the circuit is open and power is not getting to whatever would be powered. I thought the capacitors on the logic board were to smooth out voltages, and I'm not using the original power supply but instead a modern PicoPSU while bench testing.

You shouldn't run a circuit that has leaked electrolyte on the board.

The reason is that the electrolyte allows electrolysis to happen - it strips the outer surface off pins and things corrode way faster than when a computer is just sat on a shelf. The damage tends to concentrate on the power pins, so much so that you can tell which pins are the power pins. Running a machine with no visible damage, but that the caps have leaked, then putting it back into storage for a few months gets you this :

 

[phipli]

I dug out the original photo, so here it is in higher resolution :

 

[djc6]

You shouldn't run a circuit that has leaked electrolyte on the board.

I first blew off all the dust with an electronic duster. Then I used a spray bottle of 91% IPA and a toothbrush. On the areas with visibly corroded pins (most SOIC chips in video circuitry) I used vinegar for a few minutes and then rinsed it off with IPA. But I still feel the board could be cleaner - its hard to get in all the nooks and crannies even with a toothbrush. Any tips?

 

[phipli]

I first blew off all the dust with an electronic duster. Then I used a spray bottle of 91% IPA and a toothbrush. On the areas with visibly corroded pins (most SOIC chips in video circuitry) I used vinegar for a few minutes and then rinsed it off with IPA. But I still feel the board could be cleaner - its hard to get in all the nooks and crannies even with a toothbrush. Any tips?

Not hugely helpful, but I got myself an ultrasonic for this sort of thing. Other than that the only thing you can do more is if you can find a solvent proof tray and submerge it in IPA and brush it gently with a paintbrush (make sure you have some ventilation). Make sure you let it dry before powering it on. I'm not a huge fan of scrubbing them because I worry a bristle will catch and bend or break something.

But I guess the obvious question is why not replace the capacitors? What's preventing you from fitting new ones?

 

[djc6]

But I guess the obvious question is why not replace the capacitors? What's preventing you from fitting new ones?

I will - I explained in the first post I just wanted to know if it was safe to run it briefly with my oscilloscope to see whats wrong with the video circuity. I did some troubleshooting last night - appears to be a bad LS393 at UF8. So I'm going to order one of those along with the capacitors.

 

[phipli]

I will - I explained in the first post I just wanted to know if it was safe to run it briefly with my oscilloscope to see whats wrong with the video circuity. I did some troubleshooting last night - appears to be a bad LS393 at UF8. So I'm going to order one of those along with the capacitors.

A fair. I'd forgotten you said that because I have a box full of caps and wasn't thinking.

So that is a slightly different situation, it is still... not great to run a machine without capacitors. Sometimes you get serious transients that can damage a chip, the caps soak them up. I know you mention using a modern solid state PSU, most of the time you'll be fine, and then one day you're not. Plus some of those little pico ATX things aren't as good as others (not saying yours is - just some of them seriously lie about their capacity, and some are budget items with poor quality soldering and questionable parts ratings).

If you mean running with a couple of caps missing, and that most are still there, you'll be fine.

If none were fitted and I hadn't put in an order yet, what I would have done is consider what was missing etc. I'd be happy as long as there was some capacitors on that part of the circuit, so I'd make sure I had a reasonable amount of capacitance spread around the board on the 5v, some on the 12v, -v as well. Sound... the caps on the rails I'd make sure something was there, but I would ignore the ones in series because meh - you'll just not get sound. Kind of something is better than nothing. Old macs are pretty resilient to having the "wrong" capacitance, apple just seem to have taken the attitude of making sure there is plenty enough.

Values wise, it really doesn't matter too much within an order of magnitude, not with the 47uF stuff apple put all over the board. Just tack a few through hole caps, 33uF, 47uF, 100uF, whatever, around the board.

If you're expecting to be working on more macs, I'd just order a hundred 47uF caps of your preferred type (I wouldn't use ceramic like some people do, they have a different capacitance / voltage characteristic, but regular electrolytic, tantalum and polymer would all be fine) and stick them in a drawer. Macs only tend to use a few 1uF 50v, 10uF and 100uFs other than the majority of 47uFs, other than the occasional axial... 220uF??? I forget, whatever they are on the mac II era machines. For the few oddballs I just have a box of misc SMD caps and I fish one out of that. Just remember to uprate any tantalums you use on 12v rails to 25v.

And also, never offer to recap a 7100, it will bankrupt you.