General Category > Lisa Troubleshooting and Repair
CPU 42 error...
sigma7:
--- Quote from: bmwcyclist on March 22, 2025, 04:51:00 pm ---R8 - 151ohm.
VROM pin 14 to R8 close side I have connectivity with no resistance
VROM pin 14 to R8 far side I have 151 ohm.
--- End quote ---
Those readings are correct, so no problem there.
--- Quote ---Could I have 2 bad VROMs? both have stickers on them, so I assume they are burnable?
--- End quote ---
You could have a bad one, but they aren't a typical cause of failure. If you did have a bad one (or installed a PROM that is not a video PROM), the video display is likely to fail.
The Video PROMs are "One Time Programmable". They can't be erased and reprogrammed.
These are bipolar fuse programmable parts. They have literal fuses in them that are blown open during programming in the desired pattern. The fuses can't be restored. Although one can blow additional fuses, that's not usually what's needed.
(Some other OTP parts are actually EPROMs in windowless packages to save the expense of the windowed package; they potentially could be erased by x-rays or something like that, but the video PROMs are not like those. The Lisa doesn't have any of those.)
So you'll need to do more detailed troubleshooting... are you interested in purchasing an inexpensive logic analyzer or can you borrow one?
bmwcyclist:
--- Quote from: sigma7 on March 22, 2025, 05:20:19 pm ---
Those readings are correct, so no problem there.
--- End quote ---
Ok, thanks!
--- Quote from: sigma7 on March 22, 2025, 05:20:19 pm ---You could have a bad one, but they aren't a typical cause of failure. If you did have a bad one (or installed a PROM that is not a video PROM), the video display is likely to fail.
--- End quote ---
I tried the known good one from my other board and no change so its not the VROM.
--- Quote from: sigma7 on March 22, 2025, 05:20:19 pm ---So you'll need to do more detailed troubleshooting... are you interested in purchasing an inexpensive logic analyzer, or can you borrow one?
--- End quote ---
My friend is going to give me his unfinished Vintage Chip Tester kit. But if you think that is barking up the wrong tree, then yes, I will need to buy one.
.
sigma7:
--- Quote from: bmwcyclist on March 22, 2025, 08:41:59 pm ---My friend is going to give me his unfinished Vintage Chip Tester kit. But if you think that is barking up the wrong tree ...
--- End quote ---
I think it is very unlikely the chip tester will help you figure out which chip to test, and you won't want to de-solder more parts than you have to, so IMHO that is not the place to start. But someone else may have another approach in mind... anyone?
A couple of other simple tests that may or may not give clues:
1) When you get error 42, if you "continue", can you boot into MacWorks and if so, does it operate correctly?
2) If you have a set of 3A ROMs + the video PROM for the square pixel screen modification kit... if you install those on the problem board, does it still give error 42? Since those don't have a serial number, error 42 would only be generated by a vertical retrace interrupt fault.
3) If you side-by-side compare the bad board with a good one, is there any visible evidence that the bad board has had any damage (repaired or not), or any parts replaced?
4) I don't know how much tolerance there is in the serial number timing, but I wonder if the crystal Y1 (near coordinates F1) was replaced with something not sufficiently close to 20 MHz... is it stuck to the bad board the same as on the good board?
5) The new 68000 CPU you installed in the previously dead board... if you put that CPU in the good board, it works and does not generate error 42, correct?
edit: added #5
bmwcyclist:
--- Quote from: sigma7 on March 23, 2025, 01:07:47 am ---
1) When you get error 42, if you "continue", can you boot into MacWorks and if so, does it operate correctly?
--- End quote ---
No.
I made a video that includes the startup tone and a walk-through of trying to boot.
https://youtu.be/GJHmPptvQhc?si=y2Up7WxvseUwC9oT
--- Quote from: sigma7 on March 23, 2025, 01:07:47 am ---2) If you have a set of 3A ROMs + the video PROM for the square pixel screen modification kit... if you install those on the problem board, does it still give error 42? Since those don't have a serial number, error 42 would only be generated by a vertical retrace interrupt fault.
--- End quote ---
I only have H ROMs
--- Quote from: sigma7 on March 23, 2025, 01:07:47 am ---3) If you side-by-side compare the bad board with a good one, is there any visible evidence that the bad board has had any damage (repaired or not), or any parts replaced?
--- End quote ---
The board looks pristine (see attached photos) except for some scratches on the small capacitors bodged on one of the chips (see attached photo)
--- Quote from: sigma7 on March 23, 2025, 01:07:47 am ---4) I don't know how much tolerance there is in the serial number timing, but I wonder if the crystal Y1 (near coordinates F1) was replaced with something not sufficiently close to 20 MHz... is it stuck to the bad board the same as on the good board?
--- End quote ---
It appears to be original but has no freq numbers that I can see. (see attached photo)
--- Quote from: sigma7 on March 23, 2025, 01:07:47 am ---5) The new 68000 CPU you installed in the previously dead board... if you put that CPU in the good board, it works and does not generate error 42, correct?
--- End quote ---
I am loath to pull and swap the CPU out of the good board. From a previous discussion here, I learned that the CPU sockets can fail after repeated use, and I don't know how much use these used boards have already had.
I do have a 3rd CPU, new and unused, which I can try.
stepleton:
I'm sorry I don't have the time to pitch in to this diagnostic session --- I've got a lot on my plate at the moment. But here's a remark about the Retro Chip Tester Pro:
I have one, and I love it --- it's a wonderful device as well as a fun kit to assemble. What's become clear to me in using it, though, is that you can only really trust it if it says that a chip is bad.
The RCT Pro will tend to exercise a chip pretty well. But I suspect it will do this with voltages that (for TTL at least) are well within the band of legitimate TTL signal voltages and not something closer to the margins. It will also only change a chip's inputs at a certain frequency or within a certain frequency range. Both of these testing conditions may differ from the operating conditions for the IC, which the computer may drive at higher frequencies or with signals that are not as well-formed as the ones the tester uses. If the IC is marginal, in other words, the tester may not catch it. Therefore, a chip that tests well may still be too broken for the computer you want to use it in.
On the other hand, if the RCT Pro says that an IC is bad, then in my experience it is either really broken or at least too marginal for anything I'd like to use it for.
I don't regret getting the tester for a second, but as far as detecting problems goes, be prepared for the thing to yield a fair number of false negatives.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version