Need help troubleshooting A2000 intermittent problem

[chfriend]

I have an A2000 rev6.2 that will occasionally not boot.  It has a GVP A2000-030-RAM16 with 4MB onboard plus one 4MB GVP SIMM, a TBC connected at an XT slot, a Video Toaster 2000 and an Indivision ECS.  The power supply was blown when I got it and has been replaced with a brand new 400 watt ATX power supply (I initially used an AT supply but replaced it to see if it fixed this issue).

There was battery damage to the cpu socket which has been replaced.  ROM has been upgraded to 3.1 from 1.3 (problem exists with either ROM).

Occasionally, the system will not boot at all, giving the Indivision screen and then not booting.  I can tell when it's going to do this because the HD LED on the front will not illuminate as the hard drive spins up.  Usually I can get it to boot by cycling power, but sometimes multiple cycles are required.

If I remove the accelerator, the system will always start to boot, but if I use a Workbench 3.1 disk, it drops into the CLI and I have to manually run LOADWB.  1.3 discs will load directly to the workbench.

Once at either workbench I can click on anything EXCEPT the floppy icon.  The second I click the floppy icon, even once, the system will freeze most of the time.  Every so often I will get lucky and I'll be able to open the floppy window but closing it and attempting to reopen could cause a lock.

Booting a game (Menace) directly from floppy seems to work without issues.

Any suggestions?  I'm at the point where I'm ready to part it out and use the proceeds to help upgrade the 3000 I have which is rock solid, but I'd like to keep it to play with the video toaster due to the number of slots available.

[Ral-Clan]

First thing I would try is re-seating the Paula, disconnecting and reconnecting the internal floppy drive or swapping the floppy drive.

My second guess is that the power supply is not giving enough juice - although that seems unlikely since you said it was new.

[chfriend]

Here are the specs from the power supply.  I divided the +12 rails up when I spliced the power to the motherboard:

+5 & +3.3v combined max = 180w
+12, +5 & +3.3 = 430w max
450w Max total

3.3v = 30A
+5 = 31A
-5 = .3A
+12 = 15A x 2
-12 = .5A

These should be sufficient and I have tested them with a multi-meter from power up.  They stabilize instantly and are all within 2% tolerance at load.

The last time it stopped working completely I did reseat Paula to solve the problem.  Maybe I should pry it out and reinsert it completely.  I also wanted to add that when the computer boots with the accelerator, it will work as long as it's powered on.  No floppy issues or freezing.

[Xanxi]

Are you sure this ATX PSU matches the +5V and +12V requirements of the A2000?
Is the 3,3V line linked to something? It seems that when the 3,3V is not used on an ATX PSU, the other lines may be unstable.

[chfriend]

Are you sure this ATX PSU matches the +5V and +12V requirements of the A2000?
Is the 3,3V line linked to something? It seems that when the 3,3V is not used on an ATX PSU, the other lines may be unstable.

Those specs should overshoot what the original supply had.  This issue also happened with the AT power supply I originally used, and I replaced it with the ATX one that I had floating around because I wanted to make sure it wasn't the power supply.  

PS- The power supply was brand new in the box, but I have had it for a couple of years.  It is also one of the heavier supplies, not one of the super cheap ones that float if you toss them into the air because they don't weigh anything.

[Azryl]

Go back to basics... pull all the expansions out...  check if its booting. If all good, start re-installing your expansion cards one at a time, check it boots properly between each install and keep going until the machine fails to boot. The last card is the problem

Also swap the CIA chips, just for good measure..

Az

[chfriend]

Just an update.  I've gotten the system stable with the accelerator again.  It appears one of the CIA's and possibly Paula had worked themselves partially out of the sockets, however, the Workbench freeze problem when the accelerator is disabled or removed remains.  I am assuming at this point that there is damage to one of the traces, pins, etc... on the 68000 or one of the pins is still corroded enough to cause an erratic connection on the 68000, since if it was one of the custom chips, these problems would continue with or without the accelerator connected.  

Hopefully it will stay working like this as beyond removing the 68000 and cleaning it again, I don't know what else to do.  If anyone has any additional insight I'd appreciate it.  This issue happens even with everything disconnected, including the TBC card.

Chris

[chfriend]

Just an update.  I found two problems:

1) The SCSI cable I was using had a small slice in it, probably from when I was reassembling at one point and it rubbed against part of the case

2) I reseated all of the chips

Now it works perfectly when the accelerator card is installed.

[gertsy]

Just an update.  I found two problems:

1) The SCSI cable I was using had a small slice in it, probably from when I was reassembling at one point and it rubbed against part of the case

2) I reseated all of the chips

Now it works perfectly when the accelerator card is installed.

Well done...!