Strange Mac SE startup anomaly

[spaceinvader12]

So, I was pulling out my favorite SD, my FDHD model. I haven't used it at all this month because I've been so busy with other projects. And now, an old problem I thought was gone is back. Essentially, I flip the switch and it doesn't start up right away. Rather than describe the symptoms I'll attach a dropbox link to the video of me showing it

https://www.dropbox.com/s/vqydlzlqikiyyei/V_20170412_191041.mp4?dl=0

So, here are the details that are hard to see in the video. The flyblack sounds like it's clicking (not audible over speaker popping) and the hard drive spins up and down, but gradually gets louder and faster as time goes on (pulses with the drive LED). Similarly, the checkerboard pattern gets more and more grey pixels on it as time goes on, but eventually snaps to the right screen and it finally has the correct startup bong. The time can be anywhere from a few seconds to several minutes before it boots correctly, but so far, it always has done so. It happened to be relatively short the time I videoed the issue.

I've checked the voltages, which read around 5.1V and 12.2V on the floppy port. My only guess is that it could be flyback or heat related. The problem seems like it disappears entirely in the cold as I only had it last summer, less in the fall, none in the winter, and now it's showing up again. The flyback clicking suggest it may be voltage related, is it likely that it's bad and needing replacement? I've done a lot of googling but found nothing like this anywhere and I'm stumped. What should I do?

 

[Johnnya101]

Caps.

Assuming those are original...

Just watched that video.

Get those caps replaced NOW. If they are replaced (You didn't say they were) then traces and PSU are probably the culprit?

 

[hunter44102]

strange..  Maybe just need to open it up and reconnect everything really good and look for cold solder joints or obvious things.  But who knows, could be caps on the analog board starting to go bad.   

 

[Themk]

Johnnya101, SE doesn't normally have capacitor problems with its logic board. Could be one or a couple capacitors, but not all of them. SE/30 on the otherhand, to quote techknight "YIKES!"

You might check your analog board.

 

[Johnnya101]

But its always good to at the very least check them.

They USUALLY don't have cap problems. A few might!

Now, if it was flyback related, you would have heard the bong, and it wouldn't of lasted this long. Power related, you probably would not be getting anything at all (PROBABLY). The reason why I think its caps is because when the caps heat up their problems disappear...

 

[Themk]

Which is why I was suggestion analog board. Logic board rarely has problems, but analog board has them more often.

Also, the style/type of capacitors between the SE and SE/30 is completely different, if you didn't already know that.

Not trying to dismiss SE logic board cap problems, they are just more rare, then common.

 

[spaceinvader12]

Alright, wow, thanks everyone! Consensus seems to be on the analog board, at least so far. I think the caps are original on it, but there's not odd smell or anything. When I have more time (likely this weekend) I can open my other SE and switch the analog boards/PSU to see if the problem persists. As far as the caps heating up, that may have something to do with it; the times it takes for it to start up normally get shorter and shorter after each time in succession if there's a short interval on leaving it off before powering it on. Leave it sitting (to cool off, perhaps?) and I'm back to square one.

Besides caps, which solder joints should I check, if it could be cold solder? I should be able to retouch those pretty easily. Are there any other things that I should be on the lookout for? I have a second working SE that I can use for parts but I think the power supply in that one needs a recap.

 

[Compgeke]

You said you checked the voltages, but are they cutting in and out with the pulses? Based on the hard drive cutting in and out along with the system problems I'd look towards the PSU myself. 

A somewhat sketchy method for solder joint checking that may or may not work is to give the system a couple whacks on the side and see if that changes anything. As for which to check, check them all for obvious bad joints.

 

[mattislind]

Can it be the crowbar in the PSU which false triggers?

This repair guide page 10 mentions the crowbar.

 

[Compgeke]

That's what I'm thinking. The PSU is entering protection and cutting out of it over and over again.

 

[spaceinvader12]

Had some free time and decided to take a look while armed with diagrams of the PSU, which was slightly different from the diagrams in the PDF you linked. I pulled it apart and opened the power supply. Tweaked the voltage pot from 5.2V (it was reading high) to 5V. 12V was high too, around 13ish. Currently, with it set as is, it's booting regularly again and both are back to normal. Hopefully it stays that way. But now, since I know where the pot is in the PSU, I can adjust it without taking the whole thing to pieces as my screwdriver can tweak it while it's still installed on the analog board.

Mattisland and Compgeke, you were right about the crowbar. Voltage was just a bit high enough to trip it until the trickle of volts spun the hard drive up (gradually) and that brings it just low enough to boot (still 5.15ish). The symptom on page 10 of that manual mentioned the "characteristic whupping sound" when the power supply resets which was what I was getting, and even mentioned misadjustment of the voltage pot. Thankfully no parts to repace here, at least for now.

Thanks for all your help, I'm hoping it doesn't resurface!

 

[Johnnya101]

Wow, definitely way off what I thought! Lol

 

[Themk]

So Analog board related (More specifically PSU)... Lol, glad it is figured out!

 

[spaceinvader12]

Yup, that's what it turned out to be. I guess the true test will be letting it sit unplugged overnight and testing it in the morning, since cold starts used to be the problem. We'll see how it goes then, but it seems like it's far better now.

 

[techknight]

Capacitors in the power supply. The fact that it starts up rapid and slows down means a capacitor is trying to reform with an internal heating process because of its ESR being way too high, while power is running through it. 

Your chasing a red herring with the voltage pot. 

5.2V would never trip any crowbar under any normal sense of operation. 6V on the other hand.... 

 

[spaceinvader12]

Techknight, I think you're right on this. Thanks for your post. All the tweaking and time I spent last night involved a lot of power cycling and the fact that it was under constant use meant the problem basically disappeared: a false positive. I thought just tweaking a pot was too good to be true.

So, PSU recap time. I did some googling and checked maccaps, but couldn't find a list of values for the caps in it. Does anyone have a list of such? I'd hate to just go off sight and what I see and then miss one because reordering caps would be a bit of a hassle and extra cost. Still, preferable to the issue at hand, which did actually resurface this morning. Once I locate the cap values I need I'll get on recapping the PSU, and hopefully that'll put the actual issue to rest.

Also. Johnnya101, guess you were actually on the mark there. Got a bit misled in hoping it'd just be a simple adjustment, but it likely is indeed caps.

 

[Themk]

I think johnnya101 was implying logic boad caps...

Anyway, do you have a Sony PSU or an Astec PSU?

Here is one schematic, dunno if its for Sony or Astec though: http://web.archive.org/web/20141003215555/http://museum.dyne.org/gallery/apple/stuff/mac/andreas.kann/SE_Netzteil.GIF

My documents on my computer say this is supposed to be sony:



My computer says this is supposed to be Astec:

 

[Johnnya101]

No I was implying all caps. They all will go bad...

 

[Themk]

And the SE power supply is known for needing recapping. Se logic board, not so much, but yeah these SE PSUs need help.