8

My old Amiga 500 has been living in its box for the past 20 years and does not function as it should. When turned on initially it seems to work fine, loads the Workbench disk and allows me to open/close windows and look at the contents of the disk. At some point, usually between 30 and 60 seconds of turning it on, the system will freeze. This always seems to happen when I click on something. The mouse pointer still moves, but I can no longer interact with any of the windows or pull the screen down.

If I now reset it (or power-cycle it) then it won't boot. The drive seems to read the boot block and then hangs. Removing the disk and resetting gives the "insert disk" animation (KS 2.0), inserting the disk makes it read the boot block and it then hangs. I have waited an hour before powering it back on and it still would not boot. However, when I come back next day it will boot into the Workbench again (until it hangs again).

It happily boots and runs the "World of Commodore" demo by Sanity. The demo reloads at the end (NOT by rebooting/resetting) and I let it go for about an hour during which it never crashed or froze. When I do reset it then it refuses to boot again.

I also booted into an "Amiga Test Kit" disk. Just like the demo, I left the memory test run for an hour (no problems found). After that I ran the CIA timing tests, again no problems found. Resetting the system via the menu made it hang on the boot again.

The main board looks absolutely clean, no trace of leaking caps. It is not an A500+, so no RTC and no battery.

Things I tried to no avail:

  • I have removed the trapdoor memory.
  • I have removed the Kickstart switcher.
  • The PSU came originally with this A500, but has been used with my A1200 since I bought that new, the A1200 has seen a lot of regular use in the past 3 years and is accelerated so should pull at least as much power as an A500 does. I therefore assume the PSU to be fine.
  • I have swapped the two CIA chips between the two sockets.
  • I swapped between Kickstart 1.3 and Kickstart 2.0.

Added on 12/Feb/2020

  • I have tried an A1200 PSU, with the exact same result as the A500 PSU.
  • I have pushed down on all socketed ICs. No difference unfortunately.

Added on 14/Feb/2020

  • After booting I opened a shell window, in which I could type stuff for at least 20 minutes. I started and stopped some programs from disk and all continued to work. I accidentally hit both Alt and L-Amiga, which simulates a left-mouse click and the system immediately froze.
  • When the system refuses to boot I can still go into the KS2.0 boot menu and click on things there. That continues to work fine although none of the options will allow the system to successfully boot again.

I have ordered a DiagROM, should arrive early next week.

Anybody any idea what to fix, or what to try next?

Edders
  • 861
  • 5
  • 18
  • 1
    Are any of the chips on the motherboard getting really hot? – snips-n-snails Feb 10 '20 at 20:01
  • 1
    the PSU could be the issue. A1200 PSUs are smaller and less powerful. So theorically a bad A500 PSU could still power a A1200 properly but not a A500.

    My A500 PSU is 4,5V / 5A, when my A1200 PSU is 4,5V / 3A. So A1200 is less demanding. What's the model of your accelerator board?

     – Jean-François Fabre Feb 10 '20 at 20:08
  • What Workbench disk are you using? Have you tried booting from another disk, or imaging a new disk from an adf? If it boots demos but not Workbench, then it seems like Workbench may be the culprit. And if your computer hangs after trying Workbench, and shows similar issues after a Ctrl-A-A, maybe your WB has a virus on it! – Carr Feb 10 '20 at 20:55
  • 2
    OP states: "I have waited an hour before powering it back on and it still would not boot. However, when I come back next day it will boot into the Workbench again (until it hangs again)": probably not a virus. A virus doesn't resist to a long power off. Hardware issue all right – Jean-François Fabre Feb 10 '20 at 22:10
  • get hold of another PSU. It's worth trying as it's simple and it's the only thing that you didn't swap – Jean-François Fabre Feb 11 '20 at 07:41
  • @snips-n-snails: None seems to be getting hot. – Edders Feb 11 '20 at 09:03
  • @Carr: It is a copy of the Workbench 1.3 disk, which works fine when I boot the A1200 with it. Also tried the Workbench 2.0 disk with the same result. – Edders Feb 11 '20 at 09:03
  • @Jean-FrançoisFabre Fabre: It is an original A500 PSU, the one that came with the A1200 was indeed always too feeble and has never been used. The accelerator in the A1200 is a GVP 1230 II (68030 40MHz and 68882 FPU). – Edders Feb 11 '20 at 09:04
  • if you still have your A1200 PSU, that would be worth testing with your A500. Costs nothing. – Jean-François Fabre Feb 11 '20 at 09:11
  • @Jean-FrançoisFabre: Yes, that is worth a try. I'll dig it out and report back. – Edders Feb 11 '20 at 19:48
  • 2
    A simple thing to check - make sure that all the chips are firmly pressed into their sockets. Perhaps one of them has worked slightly loose, and when the computer warms up metal expansion pushes it slightly further out of the socket. Failing that, search eBay for Amiga DiagROM - replaces the Kickstart chip and runs extensive diagnostics. Should only cost around £10/€10/$10! – Richard Downer Feb 11 '20 at 20:51
  • @Jean-FrançoisFabre: Tried the original A1200 PSU with the same result unfortunately. I have updated the question with this. – Edders Feb 12 '20 at 19:24
  • @RichardDowner: I have pushed down on all the socketed ICs and it unfortunately made no difference (and have now added that to the Q). Ordered a DiagROM on eBay, hopefully that gives me a clue when ti arrives! – Edders Feb 12 '20 at 19:26
  • "This always seems to happen when I click on something." - so if you don't click on anything it keeps going forever? What happens if you break out of the startup sequence with ctrl-d and try typing in some commands? – Bruce Abbott Feb 13 '20 at 11:40
  • @BruceAbbott: Good suggestion, with interesting results. I've updated my question with the new information. – Edders Feb 14 '20 at 19:45
  • Try cleaning the DIP leads and sockets with contact cleaner: https://retrocomputing.stackexchange.com/questions/11022/preventing-socketed-chip-corrosion – snips-n-snails Feb 14 '20 at 23:21

2 Answers2

5

After posting the question I got a lot of helpful responses. This made me purchase a DiagROM on eBay. The DiagROM initially ran all tests fine, but after about 15 minutes it would hang when starting the 2nd graphics test. All other tests still worked fine without a problem. One of the other suggestions was to use contact cleaner to clean all the DIP sockets, which I did for Denise (and I sprayed some for good luck on Agnus as well).

After this treatment the system has been working fine for the last couple of hours, so I consider it fixed at this point. Of course I will clean the rest of the sockets as well now.

Many thanks to everybody who helped with this!

Edders
  • 861
  • 5
  • 18
0

Get it recapped, a lot of issues with these machines are oxidised contacts (remove chips, clean and put them back again) and Caps ESRs being off range for the specs).

Bartek Malysz
  • 4,204
  • 11
  • 39