Announcement

Collapse
No announcement yet.

Powermac G5 late 2005 power supply troubleshooting

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Powermac G5 late 2005 power supply troubleshooting

    Note to mods: please remove the first thread i posted. It had crappy title. Hopefully this will make some powermac expert look at it more likely

    So i bought PowerMac G5, which worked for around an hour. Then i decided to partially strip it down and re-paste CPU, after which it still worked. However just after i added DDR2 memory into it and connected it to 220v, the mac made a loud pop noise, diagnostic leds on the front side of motherboard shortly blinked and since then, the power mac has no signs of life at all

    PSU is API4fS13. There is no visibly broken component on the PSU. No bloated capacitors, no exploded mosfets, nothing.

    I tested several stuff:

    5V standby line seems to have some sort of voltage spike after i plug psu in (my multimeter on 20V setting displayed overrange for split second) and then settles to around 0.05V or lower.

    Bridge rectifier clearly works. High voltage capacitors are getting charged and after i disconnect the PSU, they keep their high voltage for a very long time (tens of minutes), which probably means, that there is nothing to drain them to. I'd expect, that if the standby line was operating properly, they'd drain a lot faster.

    I tried shorting PS_ON on the PSU to ground. Nothing happened

    I tried applying 5v to standby rail on motherboard. LED lit up informing me, that i should insert CPU so motherboard very likely isn't toast

    I tried manually triggering the relay, that seems to enable the non-standby lines (PSU always clicked two times after i plugged it in, but since the failure, it no longer does that, and PSU always clicked when i turned the mac on). Relay worked as expected. I could see, that the positive output from bridge rectifier goes to the relay. I presume 5v standby line just bypasses the relay.

    I measured internal diodes on multiple MOSFETS and all reported pretty much same values. I also found two IGBTs (20N60C3), that have got G and E pins shorted, which is weird.

    5V and 25V standby lines are next to each other, which complicates stuff quite a lot. It seems like both rails are controlled by one MOSFET and one transformer.

    Should i try posting the pictures of the board and try to annotate my findings? Did anyone ever try fixing this PSU? I found one guy on this site, but he had faulty bridge rectifier and shorted MOSFETs, which i'd have detected already.
    Should i post pictures of the board and try to annotate them with what i had checked?

    **

    #2
    Re: Powermac G5 late 2005 power supply troubleshooting

    I found two IGBTs (20n60c3) that have extremely low resistance between gate and emitter pins. After desoldering one, the igbt is perfectly fine and short remains. Is that normal? Also my transistor tester labels it as N MOSFET and i also found datasheets for 20n60c3 where it's labeled as mosfet and not igbt. If that's true, gate and source are shorted out.

    I found even weirder thing. None of the transformers have any expected resistance of their windings
    When i measure two pins, i either measure dead short between them or infinite resistance. Could the transformer failure cause the loud bang, but at the same time not be hot enough to burn through the yellow tape insulation?
    Last edited by powermacenjoyer; 09-04-2022, 04:47 AM.

    Comment


      #3
      Re: Powermac G5 late 2005 power supply troubleshooting

      20n60 is mosfet
      transformers are copper, the windings should have low resistance.

      so, when you removed the mosfet, was the short on the fet or the pads on the board?

      Comment


        #4
        Re: Powermac G5 late 2005 power supply troubleshooting

        Originally posted by stj View Post
        20n60 is mosfet
        transformers are copper, the windings should have low resistance.

        so, when you removed the mosfet, was the short on the fet or the pads on the board?
        I could never measure resistance bigger than literally dead short.

        The short was on the pads of the board between G and S pins of the MOSFET. MOSFET itself is fine and measured in my LCR tester as N MOSFET

        Comment


          #5
          Re: Powermac G5 late 2005 power supply troubleshooting

          look at the gate driver circuit - if your lucky it will just be a shorted diode or ceramic cap

          Comment

          Working...
          X