Adventure in fixing a Lisa DataPower PSU

[berskyboy]

Hi,

I ordered 7 of the little caps and a dummy load. 

It powers up shows 5V then zero, and the other rail 7-8V then zero

here's a photo of my shenanigans

https://www.dropbox.com/scl/fi/cqwv7x3u49sj5sipfc4uw/Lisa-with-Dummy-Load.HEIC?rlkey=du47osr524mfqqnb1n58igijy&st=26thql5v&dl=0

AI told me to replace the following:

- 296-49723-ND
LM393PE3
Texas Instruments
IC COMPARATOR 2CH 8-PDIP
(here at DijiKey): https://www.digikey.ca/en/products/detail/texas-instruments/LM393PE3/2596857

- 4N35MLT-ND
4N35M
LITEON
OPTOISO 3.55KV TRANS W/BASE 6DIP
(here at DijiKey): https://www.digikey.ca/en/products/detail/liteon/4N35M/388380

I can't see where to put this one, but wanted to get more info from the group

thanks

[sigma7]

here's a photo of my shenanigans

Thanks for the update - can you post a pic of the current wiring on the other side?

I ordered...

With your next order, please get a card-edge connector so that the wiring is more easily seen/verified.

For the Lisa PSU, that is 0.156" pitch, dual row aka dual readout, 22/44 positions, with solder eyelets for a 0.062" thick circuit board.

eg. https://www.digikey.ca/en/products/detail/edac-inc/355-044-500-202/1295268

If you don't have a junk-box with scavenged/spare parts where you can find a free one, you might also get a switch that you can use to turn on/off the PSU (acting like the signal from the I/O Board & soft switch).

eg. https://www.digikey.ca/en/products/detail/adam-tech/SW-T3-1A-A-A3-S1/15284460

Those two parts suggested as among the lowest cost of what is currently in stock at DigiKey, there are lots of alternative parts that will work if those are out of stock. You could even use a household wall switch if you have one of those.

AI told me to replace the following:

LM393PE3
IC COMPARATOR 2CH 8-PDIP

4N35M
OPTOISO 3.55KV TRANS W/BASE 6DIP

I can't see where to put this one, but wanted to get more info from the group

AI is hallucinating, the Lisa PSU doesn't have an optoisolator and replacing the LM393 is very unlikely to be the solution in this case.

Keep in mind that desoldering parts isn't like disassembling something mechanical. It is easy to damage a circuit board (delaminating the fibreglass, lifting traces, etc.), and the damage accumulates, so replacing parts that don't need replacing may result in creating an unreliable end-result or even scrapping the board entirely.

[berskyboy]

With your next order, please get a card-edge connector so that the wiring is more easily seen/verified.

For the Lisa PSU, that is 0.156" pitch, dual row aka dual readout, 22/44 positions, with solder eyelets for a 0.062" thick circuit board.

eg. https://www.digikey.ca/en/products/detail/edac-inc/355-044-500-202/1295268.

If you don't have a junk-box with scavenged/spare parts where you can find a free one, you might also get a switch that you can use to turn on/off the PSU (acting like the signal from the I/O Board & soft switch).

eg. https://www.digikey.ca/en/products/detail/adam-tech/SW-T3-1A-A-A3-S1/15284460

Those two parts suggested as among the lowest cost of what is currently in stock at DigiKey, there are lots of alternative parts that will work if those are out of stock. You could even use a household wall switch if you have one of those.

I have this in my BOM at DijiKey.  (I added the capacitor as I ordered a wrong part the first time)
(how will I use this CONN EDGE DUAL FMALE 44POS?)

https://www.digikey.ca/short/b4rwqz2f

Here's a recent photo.

https://www.dropbox.com/scl/fi/zkmb2xrtpdenb2k7mu5gr/LisaPSU-Bottom-2026ASept05.jpeg?rlkey=eli19l3jrqhjdj9ih26ed0dnq&st=kf66l5zm&dl=0

https://www.dropbox.com/scl/fi/rnek80a8q0bm78civyp5p/LisaPSU-Top-2026ASept05.jpeg?rlkey=q69vo78b9po0hzs3xg4zmwm3p&st=8lxcjac4&dl=0
(I recapped all the caps that are missing)
https://www.dropbox.com/scl/fi/b5x7w6wy0jsux7xby6pm6/LisaPSU-Top-2026ASept05b.jpeg?rlkey=x3mmfyvw11bjlz7mqts2biw2h&st=s2g4mbwk&dl=0

I had a short after reflowing all the joints on the bottom.  I have new fixed that and getting 5V then zero and 7-8V then zero on the 12V rail.  I think the replacement cap coming will fix the 12V rail.

I did rejumper the jumpers wires too

[sigma7]

how will I use this CONN EDGE DUAL FMALE 44POS?

Your PSU under test will plug into the new edge connector (it is the same connector as inside the Lisa chassis). You will attach the switch and load resistors to the edge connector rather than soldering wires directly to the PSU. Some advantages are that you can replace components without jumper wires in the way, access all the output voltages, see the load wiring from the top, and so on.

I did rejumper the jumpers wires too

Please post a picture showing the current wiring... the wires disappear under the board in the latest pictures.
Here are some things to check - please ask questions if anything is unclear:

Unplug the power supply and wait 5 minutes.
Check the DC voltage across R33 (from one end to the other, not to ground) & across R2. (These are two large resistors beside the large black capacitor C8.)
If they are not below 0.1 VDC, wait longer.

Once the voltages are below 0.1 VDC:

Detach the load resistors (one lead of each is sufficient, you can leave one lead connected or remove entirely)

Put your digital multi-meter (DMM) into the "Diode Test" mode.

Locate CR22 along the top edge of the board. Put the Red DMM probe on the top lead of CR22 (the end with the light coloured band) and the Black DMM probe on the bottom lead of CR22. Once the display settles, the DMM should indicate open-circuit (same display as when your DMM probes are not touching anything). Reverse the probes so black is on top and red is on the bottom of CR22. When the display settles, the reading should be around 0.4 V.

Repeat for CR21 beside it, expecting very similar readings.

Check CR23 and CR25 in the same way, these two should show around 0.4V in one direction and a bit less than 0.4V in the other.

Check CR24 and CR26 with red on the bottom and black on top. Each of these should settle to about 0.4 V. You can try the other direction, but it will take a long time for the voltage to settle since they have big capacitors attached, if the voltage increases past about 0.5V then you can stop waiting; we're looking to confirm it isn't 0.0V indicating a shorted diode.

Locate CR20, which is a big silvery component (often marked NSD3040) with a cylindrical top on a black heatsink. With the red lead on ground and the black lead on the metal case of CR20, the DMM should settle around 0.1 or 0.2 V (not 0.0V). If you reverse the leads, the reading should start low and increase... you may remove the leads when you reach 1V or so, you don't need to wait for it to settle; we just want to make sure it isn't near 0V.

Return your DMM to the DC Voltage mode if that is separate from the On/Off switch.

Report your findings.

... continued

Assuming nothing suspicious was found checking the diodes above (ask if you're not sure), proceed with these live measurements:

1. Connect the +5, +12, and +33 loads, making sure that "+5 Sense" is connected to the +5V output.

2. If your test setup has an on-off switch to control the PSU, set it to the off position.

3. Plug in the PSU... be careful what you touch as there is now line voltage present on the board.

4. Check the DC and AC voltages from pin 3 of the daughterboard connector to ground. This should be about 16 to 18V DC, and should have an AC voltage of less than 1V. If the DC voltage is about half or double, remove power and investigate whether the 115/230V jumpers are not installed correctly or damaged or not making proper contact. If the voltage is zero, check the fuse. Remember to set the meter back to DC volts after measuring the AC volts.

5. Check the DC voltage across R33 (from one end to the other, not to ground) & across R2. (These are two large resistors beside the large black capacitor C8.) They should be about 150-170 V each (positive or negative, depending on which way you probe them). If not, then the primary rectifier, line filter, and associated components are suspect.

6. Check the DC voltage at pin 12 of the daughterboard connector to ground. This should be about 5V. If not then Z2 on the daughterboard is suspect.

7. If your test setup has an on-off switch to control the PSU, set it to the On position. Make sure the interlock switch has been engaged.

8. Check the +5 and +12 output voltages. If they are close to those values, then what follows is the wrong troubleshooting procedure... this is for the case where those voltages are close to zero (perhaps after briefly being at the correct voltage).

9. Check the DC voltage at pin 8 of the daughterboard connector to ground. If this is low (around 0 VDC), then some part of the circuit is telling the PSU not to run. That could be the interlock switch, the +12V overvoltage crowbar on the main board, the +5 or +12 voltage sense on the daughterboard, the over-current sense on the daughterboard, or the soft-power control circuit. If it is high (around +5 VDC), then the daughterboard is attempting to run the PSU.

10. Check the DC voltage at pin 13 of the daughterboard connector to ground. If this is low (around 0 VDC), then the brownout detect is telling the PSU the voltage is not high enough to turn on, or it thinks the soft power switch is indicating turn-off.

11. Carefully (be careful the probe does not slip and cause a short circuit) check the voltage at the + side of C14 on the daughterboard. If low (around 0 VDC), then the brownout circuit is suspect.

12. Carefully (be careful the probe does not slip and cause a short circuit) check the voltage at the lower end of R30 on the daughterboard, if high (around 5 VDC) then the soft-power circuit is suspect.

Report your findings.

Note this is a new troubleshooting procedure description, so there may be errors. We will double check conclusions and discern further steps depending on what you find.

edit: clarified resistor location, typo
edit 2: continuation
edit 3: clarified DMM, "high" & "low" voltages
edit 4: added reference to +33 load

[sigma7]

Here's a recent photo.

https://www.dropbox.com/scl/fi/b5x7w6wy0jsux7xby6pm6/LisaPSU-Top-2026ASept05b.jpeg?rlkey=x3mmfyvw11bjlz7mqts2biw2h&st=s2g4mbwk&dl=0

The pictures show that at least one of the trimmer potentiometers (R11) on the daughterboard is askew. I suspected this is just a manufacturing slip, but now wondering if it may indicate that it suffered some trauma or other misadventure, breaking something (either a trace or solder joint on the circuit board, or the potentiometer itself).

I don't think we've seen a picture of the back of the daughterboard, so:

• please post a pic of the back of the daughterboard so we can look for damage, and
• wiggle the trimmer potentiometers to see if they are rigidly attached to the board or at all loose.

[berskyboy]

Here's a recent photo.

https://www.dropbox.com/scl/fi/b5x7w6wy0jsux7xby6pm6/LisaPSU-Top-2026ASept05b.jpeg?rlkey=x3mmfyvw11bjlz7mqts2biw2h&st=s2g4mbwk&dl=0

The pictures show that at least one of the trimmer potentiometers (R11) on the daughterboard is askew. I suspected this is just a manufacturing slip, but now wondering if it may indicate that it suffered some trauma or other misadventure, breaking something (either a trace or solder joint on the circuit board, or the potentiometer itself).

I don't think we've seen a picture of the back of the daughterboard, so:

• please post a pic of the back of the daughterboard so we can look for damage, and
• wiggle the trimmer potentiometers to see if they are rigidly attached to the board or at all loose.

Hi Sigma7, I did order the parts from DijiKey, but in the meantime here's the photos of my daughter board.

https://www.dropbox.com/scl/fi/cno0vcd618l7x5om0bj45/LisaPSU-Daugher1.jpeg?rlkey=f6jctwfti5tmbdklgnbu3mpk8&st=o711m088&dl=0

https://www.dropbox.com/scl/fi/q08payw5ehw7si9gln1hw/LisaPSU-Daugher2.jpeg?rlkey=n6oanpg7c4rkdthew3ibv3f0p&st=cokxhfes&dl=0

2.  They seem solid.  I was using a screwdriver to adjust some of the values and they were moving.

[berskyboy]

how will I use this CONN EDGE DUAL FMALE 44POS?

Detach the load resistors (one lead of each is sufficient, you can leave one lead connected or remove entirely)

is this R33 and R2? detaching, them measuring the diodes.  When do I reattached these resistors?  Or NOT yet, continue with steps 1-12?

[sigma7]

Detach the load resistors (one lead of each is sufficient, you can leave one lead connected or remove entirely)

is this R33 and R2? detaching, them measuring the diodes.  When do I reattached these resistors?  Or NOT yet, continue with steps 1-12?

In this case, the load resistors are the light bulbs or aluminum housed power resistors or the like that are external to the PSU. ie. they are resistors that are not permanently part of the PSU. They simulate the load that is presented by the real Lisa hardware when the PSU is installed in the chassis.

You need to disconnect the external load resistors to check the diodes as they will affect those measurements.

After checking the diodes, you attach the load resistors in step 1 of the second part: "Connect the +5, +12, and +33 loads, making sure that "+5 Sense" is connected to the +5V output."

[sigma7]

here's the photos of my daughter board

https://www.dropbox.com/scl/fi/q08payw5ehw7si9gln1hw/LisaPSU-Daugher2.jpeg?rlkey=n6oanpg7c4rkdthew3ibv3f0p&st=cokxhfes&dl=0

The circle in the attached pic. is around what looks like it may be a broken solder joint.

Put a finger on the questionable solder joint, and wobble the body of the potentiometer to see if you can feel the joint moving.

Regardless of whether you can feel anything moving, resolder the 3 legs of that potentiometer.

Ideally, you would desolder the 3 legs, fully seat the potentiometer against the board, and resolder. This would reduce the chance of the same thing happening again.

[berskyboy]

Detach the load resistors (one lead of each is sufficient, you can leave one lead connected or remove entirely)

is this R33 and R2? detaching, them measuring the diodes.  When do I reattached these resistors?  Or NOT yet, continue with steps 1-12?

In this case, the load resistors are the light bulbs or aluminum housed power resistors or the like that are external to the PSU. ie. they are resistors that are not permanently part of the PSU. They simulate the load that is presented by the real Lisa hardware when the PSU is installed in the chassis.

You need to disconnect the external load resistors to check the diodes as they will affect those measurements.

After checking the diodes, you attach the load resistors in step 1 of the second part: "Connect the +5 and +12 loads, making sure that "+5 Sense" is connected to the +5V output."

OK that makes sense.  And should I have those jumper wires also connected when doing the diode tests.  And reload the dummy load and jumper wires for test 1-12 for live measurements.  (I think the parts I bought from DijiKey should make adding those loads and the jumper wires much easier and I can put jumper wires to the male connectors of the Card Edge Connectors).

[berskyboy]

here's the photos of my daughter board

https://www.dropbox.com/scl/fi/q08payw5ehw7si9gln1hw/LisaPSU-Daugher2.jpeg?rlkey=n6oanpg7c4rkdthew3ibv3f0p&st=cokxhfes&dl=0

The circle in the attached pic. is around what looks like it may be a broken solder joint.

Put a finger on the questionable solder joint, and wobble the body of the potentiometer to see if you can feel the joint moving.

Regardless of whether you can feel anything moving, resolder the 3 legs of that potentiometer.

Ideally, you would desolder the 3 legs, fully seat the potentiometer against the board, and resolder. This would reduce the chance of the same thing happening again.

OK, I will resolder those potentiometer.  I did a resolder of the main board yesterday also.  Once I do that I will re-send the photos of the bottoms of both cards.

[sigma7]

should I have those jumper wires also connected when doing the diode tests.  And reload the dummy load and jumper wires for test 1-12 for live measurements.  (I think the parts I bought from DijiKey should make adding those loads and the jumper wires much easier and I can put jumper wires to the male connectors of the Card Edge Connectors).

I believe the jumper wires for the soft power switch and the +5 Sense shouldn't affect the diode check, so you could leave those.

Yes, being able to plug and unplug everything at once with the card-edge connector is convenient.

[berskyboy]

how will I use this CONN EDGE DUAL FMALE 44POS?

Locate CR22 along the top edge of the board. Put the Red DMM probe on the top lead of CR22 (the end with the light coloured band) and the Black DMM probe on the bottom lead of CR22. Once the display settles, the DMM should indicate open-circuit (same display as when your DMM probes are not touching anything). Reverse the probes so black is on top and red is on the bottom of CR22. When the display settles, the reading should be around 0.4 V.

Here's the readings if I've done CR22 correct

https://www.dropbox.com/scl/fi/8cwfuee42nvnn3gwvw3gl/CR22_1.jpeg?rlkey=iqhlzg6a3wwvir2uvbpd66eez&st=914epwqo&dl=0

https://www.dropbox.com/scl/fi/766m25yt8jibe72yms6t5/CR22_2.jpeg?rlkey=stnxp2q8b6perldu17kkq3ayz&st=gys6p8mv&dl=0

https://www.dropbox.com/scl/fi/o7p1vgwum3edotlcs5z70/CR22_3.jpeg?rlkey=lkrds1cduqm576b4d6rdca0pf&st=pbi8eb84&dl=0

CR23

https://www.dropbox.com/scl/fi/ceeth99alhm5w0zqqqkoe/CR23_1.jpeg?rlkey=ejej1etyeiv5w3nlpyjxg687o&st=3pvwxm28&dl=0
https://www.dropbox.com/scl/fi/p40kxcrnsrrluxfh2jwlm/CR23_2.jpeg?rlkey=rsyxfm0ivitw0wfvj6fba7x7i&st=0t26p1jh&dl=0

and the rest of the CR values in the area

https://www.dropbox.com/scl/fi/q310mpvrycrgaetpg9xwf/Other-CR-Values.HEIC?rlkey=ivho0ghzd6zv8odqe188ha8bj&st=bcz8m4hz&dl=0

so it seems like CR23 and CR25 have values other than 1 (458 and 536)

CR20

https://www.dropbox.com/scl/fi/oapr7jy1eamhjydykcu42/CR20.HEIC?rlkey=3iyh3xovugcqohuh9ddxaqcic&st=ol29exls&dl=0

and climbs with the leads on the opposite polarity.

[sigma7]

Here's the readings ...

Everything looks fine there.

Your particular DMM has a couple of quirks:

For the diode test at least and probably resistance too, "Out of range" or "Open Circuit" is shown as a "1" at the left with no other indicators lit.

For the diode test, the decimal point is not shown, so the display shows "461" when the measurement is 0.461 V.

This isn't unique to your brand of DMM, and other brands have different ways of showing the same thing.

[berskyboy]

Here's the readings ...

Everything looks fine there.

Your particular DMM has a couple of quirks:

For the diode test at least and probably resistance too, "Out of range" or "Open Circuit" is shown as a "1" at the left with no other indicators lit.

For the diode test, the decimal point is not shown, so the display shows "461" when the measurement is 0.461 V.

This isn't unique to your brand of DMM, and other brands have different ways of showing the same thing.

So I've put the new capacitor into C24 with higher voltage (CAP ALUM 4700UF 20% 35V RADIAL), and put the edge connector and now it's doesn't do anything when turning on, but when I turn it off, it makes a noise and stays at 0.5V.  Here's a video of it all

Any help would be amazing.  I'm even thinking of shipping it to someone to fix?

https://www.youtube.com/shorts/OFgUFjGKWMQ