Print

[paule]

I have a Lisa that doesn't power up.

Lisa Troubleshooting Guide (Level I Lisa Technical Procedures) spells out these steps:
    ...
    4. Verify secure connection of ribbon cable from motherboard to softswitch/keyboard assembly.
    ...
    7. Swap out the modules in the following order:
      a) power supply
      b) I/O board
      c) softswitch/keyboard assembly
      d) front interlock switch
      e) disk drive to card cage connector
      f) motherboard
      g) videoboard

I swapped out the power supply with a known working one without change.

Do the softswitch and interlocks need to be swapped out to be tested or are there shortcuts folks have discovered? (I can do it with parts from another working Lisa, but it seems like a fair amount of work to largely disassemble and reassemble two Lisas to test each part.)

Any other tips before I proceed?

[blusnowkitty]

I'd start here by pushing in the interlock switches and measuring for +5vsb on the COP421. Correct me if I'm wrong, don't the interlock switches kill even standby power if they're not set?

[fri0701]

I had a problem like this recently, and it turned out that one of the connectors that connects to the card cage when pushed into the back of the case was corroded and damaged. You might want to check that if your Lisa has battery damage (but beware - the pins are very delicate).

[compu_85]

Do you have a 2/5 or a 2/10?

Note that the COP won't turn the system on if it thinks it's already on. On my 2/5, battery leak damage had corroded the pin going to the COP which told the system it was on. If I forced the system to come on it would work, repairing that trace made it work normally.

-J

[paule]

I should have specified 2/10, so no battery damage per se. Could be that some other kind of corrosion has happened.

How did you "first the system to come on"?

[stepleton]

+1 to blusnowkitty's suggestion to check +5V standby power. It's unlikely that this has failed (it's a really simple linear power supply) but the show won't start without it, so may as well give it a look. You can plug the PSU into the wall and measure between component-side pin 20 (+5V standby) and component-side pin 15 (one of the ground pins). I don't think the interlock needs to be engaged for it to work, but if you don't see a voltage, try defeating the interlock first.

Maybe you'll get lucky and all you'll need to do is replace a 7805...

[paule]

Sorry, mistyped: computer_85, how did you "force the system to come on"?

[sigma7]

how did you "force the system to come on"?

The COPS generates the signal to turn the power supply on or off.

In order to do that, the COPS runs on the +5STBY supply, which is always available when the Lisa is connected to line power. Hence the recommendation to check that +5STBY is available (it should be a bit more than 5V when the Lisa is plugged in).

If the COPS & front panel power switch aren't functioning, but +5STBY is available, you may be able to turn on the Lisa by forcing the "turn-on" signal manually.

The "ON" signal from the I/O board to the power supply is pin 120 of the I/O board card edge connector, which is on the back of the board so not so easily accessed.

More easily accessed is R42, which is the resistor for the ON signal that pulls it up to +5STBY when the COPS allows it. By adding a jumper across R42 (ie. short circuit across R42), the ON signal will be forced high, which would typically force the power supply to turn on if +5STBY is working and the front and rear safety interlock switches are engaged.

Note that this is in contention with the output of the 7417 that normally drives the ON signal to the off state. In most cases one can do this with TTL signals without damage, but it may cause overheating or damage to more sensitive parts, so minimize the amount of time the jumper is applied.

[paule]

Thank you very much for that excellent explanation. Hopefully I be able to work on this some tomorrow.

[paule]

So, good news. The Lisa Troubleshooting Guide was good as gold. It turns out that just removing and reinserting the IO board brought the system back to life (I actually swapped in a good one, powered on, then put the original back in).

However, the Widget did not turn on or did not spin up at least, and one of the RAM boards is problematic (71 - parity error).

Diaganosis continues...

[sigma7]

just removing and reinserting the IO board brought the system back to life

Excellent progress!

That's probably an indication that the card edge connectors need cleaning, which may fix the error 71 problem too.

Sometimes it is enough to extract and reinsert each board a couple of times to wipe the contacts, but if the edge connectors haven't been cleaned in a few years then that's worth doing as it can resolve problems.

[compu_85]

If the Widget doesn't spin, don't leave it powered - this can burn out the motor driver. Try to spin the spindle by hand - it should move freely. I'd guess its bearings are shot and it's gunked in place.

-J

[paule]

If the Widget doesn't spin, don't leave it powered - this can burn out the motor driver. Try to spin the spindle by hand - it should move freely. I'd guess its bearings are shot and it's gunked in place.

-J

Good call comp_85 - I gave the spindle a good spin by hand (I didn't notice any resistance), then the drive spun up when I powered on the Lisa. Unfortunately, it got an error 82 trying to boot up from the drive. Baby steps...