Mac SE repair: what's the best bet, after the analog recapping?

[Addicted]

Hello all. Avid readers may recall I recently repaired an SE FDHD. Along the way, I noticed some infrequent, momentary narrowing of the CRT image, so I recapped the analog board. It seemed to be fixed so I reassembled the case and declared victory. But last night the screen had a period where it was less stable than before.

I've spent part of today combing through the Compact Mac category of posts here, but with over two hundred pages of topics left to go, thought I'd ask here. If there's already a topic all about this - just point me at it, please! TIA

First step will be, of course, to go in and review my work. I did a careful review before I applied the hot-glue and found one mistake, which I fixed. Perhaps I missed something else.

While I'm on that, anyone care to suggest the next most likely repair to try? I have not recapped the logic board, nor the PSU. I replaced no transistors. I will dust off my analog oscilloscope to check the voltages + jitter at the floppy port later today.

One question I have is that this SE FDHD came with two internal 1440K floppies and an internal HDD. I am uncertain whether or not this was a supported configuration, and I wonder if I am sagging the PSU (I also added more RAM), and if that might lead to this type of screen issue? Could I be underdriving the horizontal deflection coil?

More details below. Thanks for reading this, and regards.
-A

Details: The original system pre-repair was in good shape, just a dead PRAM battery (not exploded), noisy fan, and a failed HDD. The screen now behaves very well, sometimes for prolonged periods. When it jitters, it now tends to jitter for many seconds in a row, then let up. I'm not sure, but it may be a sign of a temperature intermittent, or charge build-up somewhere that manages to drain but then builds back up after a while. Possibly I didn't get a good solder contact on one side of a replaced cap. I do not know. I'm not a skilled electrical engineer...

 

[cheesestraws]

Checking voltage and wibble on the floppy port is a good idea, especially during those events. If the voltages aren't stable, you're never going to be able to get things working reliably.

If they are, my next step personally would be to look at the sync signals coming out of the logic board on your scope (this is a bit of a pain but you should be able to get access through the top of the mini fit jr connector). If the sync stays solid while the screen is wibbling about, it's almost certainly an AB problem. If it also goes funny it's an LB problem. What you do with that information, I'm less sure: mine have all turned out to be power supply issues so far and I'm not up on CRT magic!

 

[bibilit]

When recaping the AB board, did you spent some time reflowing the screen connectors ? will explain why you have those issues when the temperature is increasing.

The Plus is known to have those issues, the SE and SE/30 less, but i have seen the same broken joints a couple of times.

Can be a PSU issue, but those are pretty solid.

The Logic Board is not prone to capacitors failure, not having the same kind of caps the SE/30 and Classic had afterwards.

 

[chiptripper]

I'd echo reflowing solder joints on the analog board, check around the top left area, L1, P1, etc.

 

[Addicted]

Thanks all. Planning to do another round of work on it in a few days. Thought I'd get a list of PSU caps ready and on order, just in case. While I'm ordering, might as well get the logic board caps too, I guess.

And I rethought the PSU load theory I had. It's very unlikely I've managed to put too much load on the PSU. One supported, shipping configuration included one HDD and one floppy - and it allowed an optional external floppy. This SE is essentially that configuration, just the external floppy has been moved inside the chassis. So, I'll check and scope the PSU voltages as planned. The PSU may need some caps, but there doesn't seem to be any worry about it being overloaded.

 

[joshc]

If the PSU is marginal/weak, a HD and floppy drive are enough load to cause issues. It's less about overload and more about what state the PSU is in currently. Measure voltages, recap PSU and analog board and go from there. I completely agree with all the advice given in the thread so far, so it's just a case of methodically beginning your repair efforts. Jittery screen (depending on how it acts) can very often be weak solder joints on the yoke connector (on the analog board side).

 

[Addicted]

It helps when the hardware basically throws you the answer. Working while booted from floppy last night when, after a few hours, the HDD spun down. That (dreaded?) descending slide down the scale to silence. I thought, would System 6.0.8 have known to sleep an idle hard drive? I thought not...

I got inside planning to sample the HDD power connector.. wondering if my recent analog board recap had failed (any new caps had popped, come loose, bad joints... and somehow caused this). Nothing looked wrong. Digging deeper, I found the HDD connector on the analog board was faulty. The +5V0 pin had a microscopic crack in the solder all the way around. It could budge if I pressed on the nylon socket. The solder looked fine until I got out a 3.5x lens.

I do not know if this explains the CRT wobble (note that I am a hobbyist, electrically), but seems pretty likely. I would think the HDD was able to keep spinning only because the loose connection happened to be conducting 'enough'. But that tiny air gap caused intermittent arcing -- jerking the load on the +5V 7809 regulator, and percolated transients throughout the rest of the system. Last night, something finally broke the connection long enough for the HDD to spin down and the HDD stayed off. That may have removed any arcing, and left the CRT stable.

I reflowed the HDD connector - and the yoke connector while I was there. The HDD is back 100%. Only time will tell if this was the sole cause of the CRT wobble. I won't have time to test it until the replacement floppy gear arrives (the lower drive sounds like it's missing a couple of teeth during ejects), but I am optimistic.

 

[joshc]

Weak solder joints are common on the SE analog board, so I'm not surprised. CRT wobble is often weak joints on the yoke connector.