Announcement

Collapse
No announcement yet.

MDD PSU Rebuild help

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

    MDD PSU Rebuild help

    Hi, was hoping to get some help from someone here... trying to fix a dead PSU for an old G4 Mac MDD (dual 867 from fall 2002). it's the ACBell 360W model. can't seem to locate the fuse it mentions on the pcb & having a hard time finding caps - locally & online.

    should I buy them from the badcaps.net store? the brands you folks recommend are there & the hard to find 10mm width sizes are there.

    also, toasty mentions one in socket C43 in another thread - haven't found this one yet. I've desoldered about 12 of them & written all the info down to buy them. everything else on the psu looks fine except for the caps the bulged out - damn teapos! - but not sure at all how to test the mosfets or anything else on it, kinda a noob

    thx

    #2
    Re: MDD PSU Rebuild help

    Don't worry too much about the ICs...yet. If it still doesn't work after the recap, you can try to replace those, but until then...

    The caps in the Badcaps store should work for power supplies as they have the specs you would want in switchmode PSU, but you want one of the capacitor legends to verify that before I send you on the wrong course.

    Comment


      #3
      Apple G4 MDD AcBel 360W Power Supply

      API1PC36 REV:A
      108W MAX on 5.0+3.3

      290PCB1 03H24-00061 REV:B 0302

      Board removal is a lot easier when the power jack is popped out. MDD goes by many names including Mirror Drive Double. Apple calls it Mirrored Drive Doors. The two 5.25" doors on the front have a mirror finish.

      Computer flashes power and the motherboard flashes red when the power button is pushed and won't turn on. When I connect the matching MDD monitor which is powered from the non functional tower the power button flashes then locks out. No more flashes anywhere when pressing the power button. The motherboard caps look good. I open the power supply to see UCC, Teapo, and LTec. The print on the Ltec sleeves has faded so I suspect heat damage. No bulged capacitors are visible so it's not obvious why this power supply doesn't work. An iESR scan may reveal where the where the problems are so I mark all the electrolytic capacitor negatives with a black sharpie dot and list out all the capacitors with what little specs I can read into the chart.

      D27: IR HFA08TB60 Ultrafast, Soft Recovery Diode
      M10: National Semiconductor LM339N Quad Comparator
      Q1, Q4: Toshiba 2SK2611 N-Channel MOSFET
      Q3, Q10: Infineon 20N60S5 Max247 N Channel MOSFET
      D28: MBR3045PT 30A SCHOTTKY BARRIER RECTIFIER

      iESR=in circuit ESR
      ESR=out of circuit ESR

      C37, C??, C??: UCC KMG 100uf 450v
      C53: Ltec TK 4.7uf 50v iESR=10.6 ESR=9.1
      C30: Ltec TK 0.47uf 50v iESR=12 ESR=11
      C45: Ltec TK 10uf 50v iESR=8.9 ESR>99
      C32: Ltec TK 47uf 25v iESR>99 ESR>99
      C43: Ltec TK 100uf 25v iESR=5.4
      C55: Ltec TK 0.47uf 50v iESR=25 ESR=25
      C56: Ltec TK 0.47uf 50v iESR=5.3 ESR=17
      C57: Ltec TK 4.7uf 50v iESR=10 ESR=9.4
      C41: UCC 35v iESR=0.01
      C59: UCC KY 1000uf 10v iESR=0.04
      C35: UCC KY 470uf 35v iESR=0.01 ESR=0.03
      C60: Teapo SC 1000uf 10v iESR=15 ESR=14.2 (bulged)
      C??: Teapo iESR=0.01
      C29: Teapo 2200uf 6.3v iESR=0.01
      C13: UCC KY 2700uf 6.3v iESR=0.01
      C10: Hermei 2200uf 16v iESR=0.01
      C6: Teapo SC 2200uf 6.3v iESR=0.01
      C7: Ltec LZG 220uf 35v iESR=0.09
      C8: Ltec TK 47uf 25v iESR=2.9

      PCB4 Sense Riser board:
      C401: Ltec TK 47uf 25v iESR=15 ESR=24
      C?: Teapo 470uf 16v iESR=0.07
      C?: Ltec 10uf 50v iESR=4.7
      C407: Ltec 10uf 50v iESR=6.0
      C?: Ltec iESR=5.4
      C405: Ltec TK 10uf 50v iESR=6.9 ESR=6.8

      PCB2 PWM Riser Board
      C?: UCC 3.3uf 50v iESR=18 ESR=24 WV=0.35

      C60 was the only cap bulged. It was under some wires so I didn't find it until part way through the iESR scan. The above chart may look complete now but that's only after many caps have been identified after pulling for replacement. Just after the iESR scan only a few caps were identified but all the iESR values were in and that revealed a lot of failed caps to be replaced. I initially considered only replacing a few small caps but the HV side seemed to be a hotbed of cap destruction compared to the LV end where the fans are so I pulled them all for testing. Good thing too since a few caps had a lot more ESR than what showed for iESR. Time for some replacements. I don't have much stock of these small caps so I had to choose low quality or different spec replacements.

      C32: Taicon VT 47uf 25v ESR=1.17
      C60: Panasonic FC 1000uf 10v ESR=0.06
      C53: Jamicon TK 4.7uf 50v ESR=2.0
      C401: Rubycon MH 47uf 25v ESR=1.18
      C45: Jamicon TK 10uf 50v ESR=2.2
      C30: UCC KME 0.47uf 50v ESR=18 (too high ESR for this replacement cap, not used)
      C30: Rubycon 0.33uf 50v ESR=7 (85*C too low for this replacement cap, not used)
      C30: Original cap is good enough and better than anything I can find
      C43: Taicon VT 100uf 25v ESR=0.80
      C56: Rubycon MH 1uf 50v ESR=2.0 (original=0.47uf)
      C55: UCC SME 1uf 50v 85*C ESR=2.2 (original=0.47uf, original=105*C)
      C57: Rubycon MH 4.7uf 50v
      C405: Rubycon MH 10uf 50v ESR=2.2
      C? on PCB2: Jamicon 3.3uf 50v ESR=2.2

      It works despite the subs. The small caps brought this one down. Keep those fans clear.
      Attached Files
      sig files are for morons

      Comment


        #4
        Re: MDD PSU Rebuild help

        >>also, toasty mentions one in socket C43 in another thread - haven't found this one yet.<<

        PS: Better to name threads with make ( Apple G4 ) & model ( API1PC36 ) in them. If I didn't know MDD was an Apple computer, I'd have passed on the thread.
        Attached Files
        veritas odium parit

        Comment


          #5
          Re: Apple G4 model API1PC36 MDD PSU Rebuild help

          See this other thread for more info on this PSU:

          https://www.badcaps.net/forum/showthread.php?t=14077

          Toast
          veritas odium parit

          Comment

          Working...
          X