SE/30 was working one moment, then it wasn't

[JeffC]

I was using my SE/30 earlier to run some benchmarks with and without a Daystar '030 40mhz accelerator. The machine has a MacSD and recently recapped logic board. The battery has been removed. It was working fine for a couple hours. I was running Speedometer 4.02 to compare some benchmarks, and when I tried to quit the program I received an error, something about a bus error that I didn't think to write down, with a restart button. When I clicked restart, nothing happened. I powered the machine off and back on, and there was no chime, just a grey screen. I removed the accelerator to see if somehow that was causing a problem, and there was no change. I've attached a photo of the screen.

Does anyone have any suggestions where to start trouble-shooting? I was thinking of re-seating the RAM and ROM, but ran out of time. I know this simasimac screen usually rears its head when there is battery or capacitor leakage, it seems strange that it happened to a working machine in the middle of use. Thanks for any suggestions!

 

[Kai Robinson]

Hmm - that could be a video RAM failure as well - still, worth re-seating everything and checking the caps aren't leaking everywhere.

 

[retr01]

Hi @JeffC!

I am sorry you are experiencing problems with your beautiful SE/30.

I would like to help you to troubleshoot.

1. Let's go back to the basics. Remove the Daystar and all PDS cards you may have in the SE/30.
2. Disconnect any external drives or peripherals you may have connected to your SE/30.
3. With just your internal HDD (or BlueSCSI or any other SD adapter you may use) as the startup disk, power on your SE/30.
4. Repeat the steps you did when you ran the benchmarks.
5. If the issue is recreated, remove the internal HDD or BlueSCSI. Reseat the ROM SIMM and all RAM SIMMs.
6. You will need to get your emergency bootup floppy disk and the benchmark disk.
7. Power on your SE/30 with the boot disk inserted. Then at the desktop, swap the disks to the benchmark disk.
8. Repeat the steps you did when you ran the benchmarks.
9. If the issue is recreated, then if the logic and analog boards have been recapped, it may be one of the chips on the logic board. If the boards have not been recapped, then please consider doing that.

1. If you did not recreate the issue when running off the floppy, then reseat the ROM SIMM and all RAM SIMMs.
2. Reinstall the HDD or BlueSCSI.

1. If you did not recreate the issue when running off the HDD or BlueSCSI, then reseat the ROM SIMM and all RAM SIMMs.
2. Reinstall one PDS card at a time.

Those steps should help you narrow down the offending device and address that. Sometimes reseating and reinstallation does the trick. If you have done all those and the screen still behaves crazy like that, then please let us know here. Someone can help you to narrow down which chip is offending.

 

[retr01]

Hmm - that could be a video RAM failure as well - still, worth re-seating everything and checking the caps aren't leaking everywhere.

I agree with @Kai Robinson. Check the logic board for leaks when you are about to yank out the PDS cards and reset the ROM SIMM and RAM SIMMs.

Please let us know if you catch a cap bulging or leaking. That could mean it happened after a long period of repeated instruction sets that the benchmark is pushing through the boards repeatedly. If not, all the caps were recapped. That might be why. I hope no caps are bulging or leaking.

 

[JeffC]

Thanks for the suggestions, I'll try to work on it tomorrow and report back!

 

[YMK]

I was thinking of re-seating the RAM

That's where I'd begin.

 

[JeffC]

Things seem to be working fine now. Here's what I did:

• Yesterday after sitting for a couple hours I plugged the machine in and tried again: no chime, displayed simasimac screen, same behavior as previously.
  • At this point it had no PDS card, and the floppy and MacSD were connected.
• This morning I unplugged the floppy and MacSD, powered on, everything worked fine. I didn't think to try first with the drives connected.
• I plugged the floppy in and then the MacSD, and everything worked fine.
• I ran benchmarks for about 20 minutes, no issues.
• I removed the ROM and RAM SIMMs, cleaned with deoxit and alcohol and replaced, everything worked fine.
  • I also cleaned around C7 at the same time since it looks like it might have leaked a bit.

So it seems things have fixed themselves, which is a mixed blessing... Awesome it works, but now I'm wondering why it had the problem. I looked at all the caps on the logic board, none were swollen, none had leaked except possibly C7. I've attached a photo, but it's not great quality since it's just with my cell phone.

I do keep this machine on a shelf, so any time I want to use it I need to pick it up and move it. I'm always very careful, but there is a bit of jostling when picking up and setting down so maybe that played into it.

Thanks for all the help, I'll update this thread if things go south again.

Edit: none of the testing above was done with the accelerator installed. Since the symptoms happened yesterday with and without the accelerator I doubt it's involved, but I'll do some more testing with the accelerator installed when I get some time, maybe this afternoon.

 

[retr01]

Morning, @JeffC!

I agree with you that this is mixed news, with more good news than bad news. I inspected the C7 area in the photograph you provided. Here is what I noticed:

​

The Motorola LS166D at UE8 is probably a 8-bit serial shift register chip. It is not fully connected as designed due to that path that looks like breaking up.

Can you look on the back of the logic board (LB)? On the back of the LB, you need to look at the solder joints. You may find some that are cracked. Cracked solder joints look like this:

​

Your Mac's logic board has been recapped before, right?

 

[JeffC]

@retr01 thank you for the very thorough reply! Yes the LB has been previously recapped, though I don't know how long ago, it was done before I purchased the board. When I get home I will take a close look at the back of the LB for cracked solder joints. I'll also see if I can get a better photo of the damaged trace.

 

[retr01]

@retr01 thank you for the very thorough reply!

No problem!

Yes the LB has been previously recapped, though I don't know how long ago, it was done before I purchased the board.

Uh oh. Sometimes it's a person who just recaps before selling. Even if that person is a professional recapper, sometimes an area on the board is simply overlooked. I suspect the logic board was previously damaged somehow prior to cleaning and recapping. Some people forget to inspect the circuit lines. Relying on eyes alone does not cut it, better to use a device to be able to see it up close clearly.

 

[JeffC]

I took some photos with my microscope, it's not great quality but it does better than a phone camera. 001 - 004 are the area around C7. 001 & 002 are zoomed out, and 003 & 004 are the same area but zoomed in. 005 & 006 are taken at an angle to try to see the trace under the edge of C7. 007 is taken with the light turned up just to show the area of what I think is capacitor leakage (after I cleaned the area with alcohol and a swab). 008 is taken on the back of the board, those pins are for the SIMM sockets, I took that one just to show there is some corrosion on the back of the board as well. I took a look for cracked solder joints, however my magnifying headset isn't strong enough to get a great view, and my microscope isn't good enough quality to get a great view. So I can't make any solid statements about the presence or lack of cracked joints at the moment.

I knew when I bought the board that a large piece of solder mask over the ground plane by the rear metal bracket had come off when the previous owner had tried to blow-dry the board, and that he had brushed on some solder mask. I've included a photo of that, 009. This may indicate that the corrosion is worse than I had thought.

My soldering skills are just enough to replace surface mount caps and do some basic trace repair, but I'm definitely not an expert. If this is all bad news, don't hold back telling me, better to know and figure out a path forward!

 

[retr01]

The SMD capacitor at C7 needs to be replaced. I am concerned about the circuitry path. You are using 99% IPA to clean that area, right?

On the bottom of the board, it does not look normal.

In the last picture, it should have been cleaned up by the previous owner.

You have options. That area needs a deep cleaning, resoldering of the flow, recapping, and restoring the circuit path or wire jumper from that point to another like a detour. If you want it done right, I recommend @AmigaOfRochester. Otherwise, you can get guidance from someone here on Tinker Different.

 

[AmigaOfRochester]

Is it even chiming? I agree, remove anything and start fresh. Ram and rom only

 

[retr01]

Oh, @JeffC, if you have Rominator ROM SIMM from Big Mess O' Wires, GGLABS ROM SIMM, or Garrett's Workshop (GW) ROM SIMM, you can boot from ROM on your SE/30 without any startup disk.

 

[JeffC]

The SMD capacitor at C7 needs to be replaced. I am concerned about the circuitry path. You are using 99% IPA to clean that area, right?

On the bottom of the board, it does not look normal.

In the last picture, it should have been cleaned up by the previous owner.

You have options. That area needs a deep cleaning, resoldering of the flow, recapping, and restoring the circuit path or wire jumper from that point to another like a detour. If you want it done right, I recommend @AmigaOfRochester. Otherwise, you can get guidance from someone here on Tinker Different.

Yes, always using 99% IPA. Thanks for the help and the referral! I've looked at the Rominator from BMOW, and just recently learned about GW's ROM SIMM. Never heard of GGLABS. That's something I wouldn't mind doing down the road once I have the rest of the system dialed in.

 

[JeffC]

Is it even chiming? I agree, remove anything and start fresh. Ram and rom only

Yes, it works fine, except for the temporary issue I had yesterday where I had the no-chime and simasimac screen. Do you have any thoughts on whether it's worth repairing, or just run it while I search for another board?

 

[retr01]

Yes, always using 99% IPA. Thanks for the help and the referral!

You are welcome, @JeffC!

I've looked at the Rominator from BMOW, and just recently learned about GW's ROM SIMM. Never heard of GGLABS. That's something I wouldn't mind doing down the road once I have the rest of the system dialed in.

I recommend BMOW's Rominator ROM SIMM for the SE/30. The nice thing about booting is that it skips memory checks making boot time faster. I have it in my IIci that has 128 MB of RAM and will put it in my SE/30 that I am restoring. My SE/30 will have 128 MB of RAM, too.

 

[AmigaOfRochester]

Yes, it works fine, except for the temporary issue I had yesterday where I had the no-chime and simasimac screen. Do you have any thoughts on whether it's worth repairing, or just run it while I search for another board?

Well if its working now, it is more likely it was a bad connection perhaps due to heat expansion. I'd check the analog board, make sure theres no cracked joints. The recap looks okay, not great but unlikely your issue. Looks like someone went over the various chips already. I'd try to reproduce it. Likely the ROM became unconnected in some way resulting in a crash. If it's working again, it's possible but unlikely to be a bad trace, usually that's permanent in my experience. Ghosts can be a bit annoying to track down.

 

[JeffC]

Well if its working now, it is more likely it was a bad connection perhaps due to heat expansion. I'd check the analog board, make sure theres no cracked joints. The recap looks okay, not great but unlikely your issue. Looks like someone went over the various chips already. I'd try to reproduce it. Likely the ROM became unconnected in some way resulting in a crash. If it's working again, it's possible but unlikely to be a bad trace, usually that's permanent in my experience. Ghosts can be a bit annoying to track down.

Thanks for the info. I'll replace that leaking cap at C7. Do you think it's worth repairing the corroded traces, or leave it alone along as it's working? I'll check the AB for cracked joints, maybe reflow them out of caution.

 

[AmigaOfRochester]

Thanks for the info. I'll replace that leaking cap at C7. Do you think it's worth repairing the corroded traces, or leave it alone along as it's working? I'll check the AB for cracked joints, maybe reflow them out of caution.

c7 is definitely not leaking, that's a new cap. That's flux. Like I said, an okay recap not great. I would have exposed those traces and covered with something.
I don't see any clearly broken traces, but that'd involve using a meter to check traces.