SGI Indigo 2 Impact PSU repair

Northsky

Reindeer Whisperer
Feb 8, 2019
23
26
13
DISCLAIMER: Opening a power supply and performing repairs to it could be deadly! Do not attempt any repairs if you do not know what you are doing.

Hello all,

Thought I should write something about my most recent repair project here. Currently I have one almost dead Indigo2 Impact PSU (060-8002-001 Rev A) on-going repairs.

At the first round, I decided to try to follow a guide pointed out to me by Ian (http://www.sgidepot.co.uk/i2psu.html). I managed to successfully replace the mentioned capacitors on the High Voltage side of the supply, but that did not help. So now this week, I have started to finally investigate the Low Voltage side of the supply. This is probably the most difficult PSU I have attempted to repair so far. As Elf's great reverse engineering post indicates, the LV and HV sides are soldered into each other with thick slices of copper and it makes it difficult to work with.

So, for the Low Voltage side. I found one SOT-23 packaged component that looked suspicious when I did optical inspection of the LV board.
1579294152219.png

Also, on the opposite side of the board were some signs of dust and solder seemed to have blackened on a resistor. When cleaning that section with gas duster, the said resistor learned to fly, so to make space to put it back I replaced the closest capacitors to it. Also, as the contacts on the board and on the component were blackened, I used a scalpel to carefully scrape off the oxidized layer and applied some fresh solder on them. Here is a series of pictures of that process:
1579294198237.png

1579294530910.png

To remove the capacitors, I used a soldering iron with chisel tip and 350 Celsius degrees set as temperature along with gel flux, leaded solder, and most importantly, a solder wick.
1579294887901.png

1579295011470.png

1579295303228.png

1579295684483.png

1579296078701.png

1579296372667.png

1579296432511.png

To be continued on next post!
 
Last edited:
  • Like
Reactions: Elf

Northsky

Reindeer Whisperer
Feb 8, 2019
23
26
13
1579296814814.png


Now that I got the resistor / capacitor issue solved I've started to investigate the suspicious SOT-23 component further. I have cleaned it and measured it as follows:
1579297176268.png


In diode measurement mode, 1 -> 2 yielded 0.603 V and the reverse was 1 V. Resistance was 5.567 kOhm. 2 <-> 3 was Open in both modes and directions. @Elf , your Indigo 2 Impact PSU is a bit different model, but this section seems identical. Could you please help a bit by measuring the component on your PSU so that we could compare results? If the results are the same, I can conclude that the part does not need replacement. However, if the measurements do not match, then this component needs to be identified so that a replacement could be ordered. :)
 
  • Like
Reactions: Elf

Elf

Storybook / Retired, ex-staff
Feb 4, 2019
792
252
63
Mountain West (US)
In diode measurement mode, 1 -> 2 yielded 0.603 V and the reverse was 1 V. Resistance was 5.567 kOhm. 2 <-> 3 was Open in both modes and directions. @Elf , your Indigo 2 Impact PSU is a bit different model, but this section seems identical. Could you please help a bit by measuring the component on your PSU so that we could compare results?
I'll just renumber quickly to match up with the SOT-23 package:
1579321164150.png


Referring to the pin numbers above in the diagram (as distinct from the ones on your photo), basically what I see is:
  • Pin 2 open to everywhere (like yours) -- and also appears to be unconnected on the circuit board, suggesting perhaps a diode
  • A diode with anode at pin 1, cathode at pin 3, with 0.62870V drop (at the tester's current level)
I did measure resistance:
  • With the positive lead of the ohm meter on pin 1 and the negative on pin 3, the resistance changes with range (as expected) due to the diode behavior, so the numbers don't mean much
  • In reverse with the positive lead on pin 3 and the negative on pin 1, I read 1.15938kR -- likely some resistance across the diode somewhere else in the circuit
The top of my SOT-23 package is marked "J1 p." At least one manufacturer has a BSS138 MOSFET in that marking, but I think this may be misleading due to pin 2 being unconnected, and the 1 > 3 diode being in the wrong place for the BSS138's body diode. So I still think it is more likely that it is a diode that also has some resistance bridging it elsewhere in the circuit. Your part is probably ok?

If you did want to replace it there are many diodes in such a package that meet that (e.g. BAS16 of which there are a nearby on the board from this, but they have different markings). But it will be hard to figure out which one just given the slim pickings for markings on SOT-23 packages :(
 

About us

  • Silicon Graphics User Group (SGUG) is a community for users, developers, and admirers of Silicon Graphics (SGI) products. We aim to be a friendly hobbyist community for discussing all aspects of SGIs, including use, software development, the IRIX Operating System, and troubleshooting, as well as facilitating hardware exchange.

User Menu