IIx died, possible bad PSU?

[Kipio]

I purchased a recapped IIx on ebay and everything seemed to work okay the first couple of times I booted it. The third time, something seemed wrong after it was on for a few seconds, and suddenly I saw and smelled smoke. I pulled the power right away. I assumed this was a blown Rifa, but no, in fact this was a dramatically blown trace on the motherboard! Tracking this back, it appears to be the trace connected to the -12V line from the PSU.

I've been trying to diagnose whether this is an issue with the PSU going too negative on that line, or something shorting on the mobo, or something like that. I've poked around the motherboard and so far I'm not seeing any obvious shorts or bad traces (except those which were burnt up) or anything like that.

I've pulled the PSU out and tried putting a load on the 5V and 12V rails. It starts for a second or two, and then shuts down (when I feed 5V from my bench PSU to its soft start pin.) When I do that with a volt meter on the -12V pin, it peaks out at -16.5V consistently before the PSU shuts down.

I suppose I have a multitude of questions:

1. Is it likely that this extreme voltage on the -12V rail could have caused my trace to go up in smoke?
2. Is the PSU refusing to stay on because it is detecting the -16.5V on that rail and going into protect?
3. If this is a PSU problem, is there a way to fix it? And if not, is there a way to replace the guts of this PSU with a pico-ATX or ATX PSU or something like that?
4. Are there other things I should be checking on my motherboard to determine if the problem was on its end?

Any help would be greatly appreciated!

 

[Fred1212]

Power supply fix or replacement, lots of options but replacing the guts with a small ATX supply is a relatively easy start.
Once supply is done, don't worry about the soft start just leave the main switch on and plug in power to supply.
Double check all your supply voltages and output connector pin positions before attempting to power up the board.
Check board with damaged track for other problems, shorts, components obviously blown etc
Double check the replaced capacitors if Tantalum for correct polarity placement, the marked end is positive, and I have seen them put in the wrong way around resulting in similar smoke from the main board.

 

[Unknown_K]

Get a picture of the trace and the area around it.

I don't think a shorted tantalum capacitor would melt a trace since the current surge is so short term (tants explode).

Take the motherboard out and see if there is a screw or something metal loose under the board.

The only boards I have seen with melted traces tend to have a component short out and the thinnest trace acts like a fuse and melts. Capacitors leaking and eating a trace so that it is barely there could also create a hot spot but probably not release a lot of smoke.

 

[Kipio]

Thanks, I will get a picture of the trace and post here. I've already cleaned it up, but the damage is still quite obvious. As I said it was a larger power (-12V) trace and it killed a couple of signal(?) traces around it.

I did check the case when I removed the motherboard and I don't see any signs of a screw or any other metal item at the bottom of the case which could cause a short. The thin plastic shield at the bottom of the case is now scorched, but I think it was intact, so I don't believe anything was arcing to the bottom of the board.

I thought of the tantalums too, since it had been recapped recently and it is so easy to reverse them because their markings are reverse of electrolytics. But they all looked like they had the proper polarity to me.

As far as leaking capacitors go, as I said the board was recapped before I got it so I don't really know the condition of them beforehand. But everything looks really really clean to me. I don't see any evidence of capacitor-related corrosion anywhere on the board. So while that's certainly possible as a cause, I'm not seeing the evidence. The board looks really clean, which is why I have sort of focused on the power supply.

As I said, I will post a picture or two! Thanks for your help!

 

[Kipio]

Here are 4 photos of the burnt traces.

As you can see, the larger trace comes from the PSU. It is connected to the -12V pin, as I mentioned. The soft power trace runs right along side it, and if you look closely you can see what appears to be a via sort of "sticking up". My theory is that maybe the heat impacted the solder in that via and made it rise to the surface!? Either way, that soft power trace seems fine, there is continuity between the PSU side and the via.

It appears that 2 or 3 other presumably logic traces also got burnt. The outermost one still appears to have continuity.

All in all I think the damage should be easy to fix with 3-4 bodge wires. But what I'm mostly worried about is understanding the root cause of the issue, because I obviously don't want to get it put back together only for it to happen again the next time I power it back on!

Any further help is appreciated.

 

[Unknown_K]

What does the top side of the board look like, is there a lot of capacitor leakage?

 

[Kipio]

I can get pictures of it as well, if that would be useful. But as far as I can tell it looks immaculate.

As I said, the previous owner recapped the board. But I haven't seen any evidence on the top side of damage from previously capacitor leakage, although I'll definitely take another look.

 

[Kipio]

Here are some photos of the top of the board. The first photo is of the entire thing and the subsequent photos are concentrated to the part of the board where the magic smoke came out (on the underside.)

Looking again, I'm still not seeing much in terms of what I would consider capacitor electrolyte damage, but I'd definitely love second (or third, etc) opinion.

 

[Unknown_K]

Wire up the wires that melted and check for shorts (DO NOT POWER THE BOARD UP).

I looked at my spare IIx board and there is nothing in that are to short out. I don't see any major leakage on top of the board that has leaked into vias.

The thing is that board shorted long enough to completely vaporize a heavy trace and a smaller one next to it along with charring a big ground plate and also melting the solder in a via close to it.

Was there a card in the Nubus slot?

 

[Kipio]

Do you mean shorts between the traces that melted? Or shorts elsewhere? I did generally try to check that -12V wasn't shorted to anywhere it shouldn't be. But so far I haven't been able to find a schematic of the IIx, so I'm unclear exactly where that bus is and isn't supposed to go.

There WAS a card in one of the NuBus slots. Unfortunately I'm not sure which slot. This was a Monochrome video card and TBH I haven't inspected it yet. I didn't think of that and I absolutely need to take a look at that card to see whether it could be shorting the -12V bus to something. Thanks for that suggestion!

 

[Kipio]

I checked my Nubus card, which is a "Macintosh II Monochrome Video Card". It actually DOES have some electrolyte leakage on its two electrolytics, so I definitely should plan to fix those.

However, I don't see any evidence that the -12V is used on that card at all. I see no traces going to either of the -12V NuBus pins. And I tried measuring for continuity between them and pretty much every other component on the card and I'm not seeing anything. So I doubt the issue was with the NuBus card itself.

 

[Kipio]

As suggested, I went ahead and wired up the broken traces and tested them for continuity, and here is what I found:

None of the 4 impacted traces are shorted toeach other.

I measured what the -12V trace DOES have continuity to. As expected, the (2) -12V Nubus pins on all Nubus slots. No connection to any other Nubus pins on any other slot.

Middle pin of Q1, which I believe is the voltage regulator that generates the -12V. Negative side of C3. 7 pins of UB2! This seems like a lot, but I don't know what this chip is or what it is for. Middle pin of D3. 2 pins on Q3. 2 pins onD4. 2 pins onD5. 2 pins on D6. 1 pin onUB6. Positive side of the capacitor on the battery board, plus battery terminals. 4th pin down (starting from the inside of the board) on edge connector J1.

Some of these seems suspect to me. Why would there be continuity to the battery board? Why so many pins on UB2? But without a schematic or spare board, I'm kind of at a loss to know what is and isn't correct. I looked at a Mac II schematic I found, but I didn't find a lot of these components on that schematic.

Any further assistance would be greatly appreciated!

 

[Kipio]

Since I'm not finding much in the way of shorts, and I think I've exhausted the ideas of the various contributors (thank you ALL for your suggestions), I'm thinking my next step is to try to replace the power supply.

I've thought about maybe replacing the innards of the PSU with a PicoPSU or something like that. However, the specs on my PSU are 18 amps on the 5V rail! I imagine that these older systems probably didn't have much if any power regulation on board, hence why so much 5V... But the PicoPSU's I'm seeing only supply maybe a max of 6 amps.

So I guess my question is, what to do about this? How much current do I actually need to be able to supply on the 5V rail? If it is on the high side, is there a good alternative that would allow me to replace the PSU innards?

Any and all thoughts would be appreciated!

 

[Juror22]

Since I have just completed re-recapping my IIx, and I (finally) have it running and starting normally, I have a couple thoughts, but first, I'd like a bit more information - which PSU do you have (Sony or Astec)?
Also, it looks like they did not replace the axial caps (they need to be replaced on the II model boards).
The board does not look 'clean' and there is a haze across a lot of areas in the photos that could be lighting and the look of the board or could be something that did not get cleaned off in the re-capping process (see around UB2 in the photo PXL_20240718_005546510.jpg). Maybe check these areas out a bit. Once we sort out which PSU you have, then we can determine next steps.

 

[croissantking]

As suggested, I went ahead and wired up the broken traces and tested them for continuity, and here is what I found:

None of the 4 impacted traces are shorted toeach other.

I measured what the -12V trace DOES have continuity to. As expected, the (2) -12V Nubus pins on all Nubus slots. No connection to any other Nubus pins on any other slot.

Middle pin of Q1, which I believe is the voltage regulator that generates the -12V. Negative side of C3. 7 pins of UB2! This seems like a lot, but I don't know what this chip is or what it is for. Middle pin of D3. 2 pins on Q3. 2 pins onD4. 2 pins onD5. 2 pins on D6. 1 pin onUB6. Positive side of the capacitor on the battery board, plus battery terminals. 4th pin down (starting from the inside of the board) on edge connector J1.

Some of these seems suspect to me. Why would there be continuity to the battery board? Why so many pins on UB2? But without a schematic or spare board, I'm kind of at a loss to know what is and isn't correct. I looked at a Mac II schematic I found, but I didn't find a lot of these components on that schematic.

Any further assistance would be greatly appreciated!

A IIcx schematic would be more useful as it’s basically the same design in a more compact form.

 

[Kipio]

croissantking: Good to know about the IIcx schematic! I'll take a look at that. And I own a working one which might come in handy.

Juror22: The PSU is a Sony. I have taken a look at it physically and I don't see any obvious leaking capacitors or anything like that. As I mentioned, I can't get it to start reliably disconnected from the board and under a dummy load. The -12V seems to "float" up to around -16V in the brief period before it shuts itself off. (FWIW, the -12V isn't under load... only 12V and 5V)

I do see what you mean by the haze in some areas. I tried cleaning one spot with IPA and it seems to come off. So I could go through the board and clean that stuff up.

Looking at the axial caps, I concur that they appear to be original. Those solder joints don't look like they have been redone. I could definitely replace those axial caps.

 

[Kipio]

I went ahead and replaced the axial electrolytic caps on the motherboard. I also replaced the SMD caps on the monochrome video card for good measure, since those ones looked like they were starting to leak.

Regarding the haze around the board, I went around and cleaned every inch of the board with IPA, but that haze didn't really come off. It almost seems like it is some artifact of the conformal coating or something like that? It doesn't strike me as leaked electrolyte or anything like that, but I'm not really sure what it is. Aside from that haze, the board seemed really clean... I didn't get much off of it.

I have purchased a TFX form factor PSU which I think I might be able to make fit into the casing for the old PSU. So barring any other things I should be attempting, my inclination is to start work on trying to retrofit connectors and whatnot to work with this modern PSU. But I am all ears if there are other things I should be looking at on the motherboard, or if there are things I should be doing to attempt to diagnose or repair my PSU.

 

[Juror22]

It sounds like you have taken care of everything you can and I was actually hoping for you to have a different PSU, since there is a recap kit for the ASTEC one, which makes sourcing the caps a lot easier. Having a list is one thing, but trying to look them all up, get the right diameters and such is so painful (at least for me).

Trying to retrofit an internal replacement should get you to the same place (a working PSU), so that may be the better way to go - if you could post pictures along the way?. So far I have lucked out with my Sony PSUs and have gotten away with just taking them apart and giving them a REALLY good cleaning. Looking out from the other direction, if there is anything that you need from me (readings, etc.) I can do that as long as its not overly destructive.

 

[Byrd]

I've come across several bad/dead Macintosh II/IIx/IIfx PSUs over the years (mostly Astec but one Sony) even when relatively new and certainly wasn't caps that were the issue back then. None had obvious signs of failure. Sounds like you've gone as far as you can, it gets expensive too replacing suspect parts so agree finding an ATX replacement will be a good move.

 

[Kipio]

Well, I can't believe it, but I was able to wire up that TFX ATX PSU in a temporary setup, plugged everything in, and it booted on the first try! No sign of any magic smoke or any other issues. So I'm tentatively confident that this WAS a power supply problem. I guess I should still try plugging in the old HDD and FDD just to make sure, but everything else that could cause a short was connected up.

No for the challenging part of getting everything to fit back in, in a tidy way. I have a plan that I've already started working on, but I think it will be fairly challenging for my skill level.

Thanks again to everyone who helped with this. I am really overjoyed that it is finally back up and running again!