Thank you to the guys at HEGE supporting Badcaps [ HEGE ] [ HEGE DEX Chart ]

Announcement

Collapse
No announcement yet.

All replacement capacitors in iMac G5 PSU burned up

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

    Re: All replacement capacitors in iMac G5 PSU burned up

    Originally posted by Toasty
    You will have to bend that piece of copper heatsink over the cap up a bit to remove the cap. The leads are just a bit too long to clear the board and allow easy removal.
    Yeh I had been wondering how best to attack that particular cap. Simply bending the heatsink had not occurred to me for some reason however.

    After replacing 5 bad Nichicon caps on this iMac's motherboard, I'm confident in my tool setup. Picking out replacements for this PSU though, not so much. Complete list of caps appears to be as follows (dimensional tolerance +-0.5mm):

    Code:
    Secondary Caps:
    	1000 uF / 6.3V / 8x14mm / 3x
    	4700 uF / 6.3V / 10x25mm /
    	2200 uF / 10V / 10x25mm / 2x 
    	1200 uF / 16 V / 10x25mm
    	330 uf / 35V / 10x16mm
    	1000 uF / 35V / 12.5x20mm
    	
    Primary Caps: 
    	100 uF / 35V / 8x11.5mm
    	2 uF / 50V / 5x11m / 2x 
    	10 uF / 50V / 5x11mm
    	47 uF / 50V / 6.3x11mm
    	560 uF / 200V / 18mmx45mm (Radial leads) / 2x
    It appears as NWD originally did, buying all Nichicon replacements from Mouser, is the easiest as they have nearly all of these. Don't know if that is the best way to go or not though.

    Also the 560uF/200V cap I couldn't really find anywhere. Was thinking of buying a Panasonic snap-in from DigiKey (P11615-ND) and just adding bigger leads (or even the old leads) to it.

    Comment


      Re: All replacement capacitors in iMac G5 PSU burned up

      I was asked this in PM where no one can see.
      Here's the short version.

      The Teapo SC is low ESR.
      330uF 35V 10mm - Ripple/ESR = 1050/.044
      Mouser doesn't stock a 330uF 35V at all with ESR that low.
      At Digikey lowest series with good enough ESR is KZE.
      KZE 330uF 35V 10x16mm - Ripple/ESR = 1430/.038
      They also stock Panasonic FM which are rated better than KZE.
      [Check physical size. FM are often big for their uF.]

      The Ltec TK series aren't low ESR. (But should be in a PSU.)
      VR would be just good enough but LXZ, FC, PW, or up would much be better.

      I never found a great match for the 560uF/200V either.
      These were the closest I've found.

      22x40mm 85c
      http://search.digikey.com/scripts/Dk...me=565-1196-ND

      25x40mm 105c
      http://www.mouser.com/Search/Product...Kzz3bGyg%3d%3d


      .
      Mann-Made Global Warming.
      - We should be more concerned about the Intellectual Climate.

      -
      Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.

      - Dr Seuss
      -
      You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.
      -

      Comment


        Re: All replacement capacitors in iMac G5 PSU burned up

        You replaced only 5 caps on the midplane? That's not enough! There are some 25 caps on that board. I see banks of 10, 4, and 5 or 7. If you just replaced the 5 obvious ones out of a bank of 10, those 5 are doing ALL the work and will eventually fail. If you just replaced 5 obvious ones across the whole board, you will have more failures. Again most likely the ones you replaced.

        Either way, you're applying a lot of stress to the PSU. Remember that these are only 180 watt PSUs. They don't have a great deal of reserve to begin with.

        See this thread for more info: https://www.badcaps.net/forum/showthread.php?t=785

        The PSU in that thread is somewhat different in construction. Plus there's a shot of that #$@!&*% coil in perfect condition.

        Although a better cap temperature wise, from a size consideration I do not recommend the 25mm primary caps. There's just not enough room for them, and the 22mm ones will be snug.

        You'll also have to carefully cut the gunk from around the green 1 watt resistor that stands upright next to them on the AC power plug side. Then, with your soldering iron on the body lead end, push the lead flush with the underside of the circuit board. That should allow enough lead length to allow you to gently bend it over just enough to get clearance for the cap. Do not re-gunk that resistor. It needs to dissipate the heat, not retain it.

        [Edit: Your cap count is 16, yet I only count 15...? This is an Apple AcBel PSU P/N: 614-0293, correct?]
        Last edited by Toasty; 03-07-2009, 02:56 AM. Reason: Cap count different
        veritas odium parit

        Comment


          Re: All replacement capacitors in iMac G5 PSU burned up

          Thanks for the guidance Toasty and PCBONEZ.

          Spent today desoldering and cleaning through holes. This PSU seems to be entirely Ltec, even the 330uF 35V. And you are right Toasty, only 15 caps. The 2uF 50V is listed as two instead of one by mistake.

          Probably order the new caps tomorrow and install next week. Thinking I'm going to go with PW series and Panasonic FMs if I can.

          Finally Toasty, don't worry about the motherboard, the rest of the caps on that were Rubycon MCZ. If you must worry about something, worry about the ground trace in the PSU I stupidly delaminated (on the second to last cap no less...). Going to take a bit of hacking on the re-install.
          Attached Files

          Comment


            Re: All replacement capacitors in iMac G5 PSU burned up

            >>You will have to bend that piece of copper heatsink over the cap up a bit to remove the cap.<<

            I didn't think I said "stand it up so straight you could hang a flag from it". Be REAL careful bending it back. It seems like fairly ductile copper, but you risk snapping it off.

            Geeze!
            {Cat rolls Rudycon gas grenade under your door for that one.}

            Okay for midplane caps then. You should be fine.

            Pic of delamination please.
            veritas odium parit

            Comment


              Re: All replacement capacitors in iMac G5 PSU burned up

              Sensitive readers may wish to avert their eyes from the carnage that follows (doubly so as we completely hijack NWD's thread):


              You will note the capacitor lead, still in its through-hole, still attached to its trace. To the bitter end for that cap.

              The trace measures about 90 mils wide. I don't know how to judge if it is single or double ounce copper. This is on the 5V rail I believe, so worst case 10 amps?

              Given the through-hole is gone, I'm thinking a bottom jumper wire from there to the big C6 cap where it normally connects. 18 AWG or bigger. Cut off the delaminated trace and pretend it never happened. Delaminated traces only happen to other people.

              Originally posted by Toasty
              I didn't think I said "stand it up so straight you could hang a flag from it". Be REAL careful bending it back. It seems like fairly ductile copper, but you risk snapping it off.
              Heh, yeh I expected it to be harder to bend. When it wasn't I said "lets make this easy" and just went for it all the way. It just has to survive a single bend more...
              Attached Files

              Comment


                Re: All replacement capacitors in iMac G5 PSU burned up

                Hang on. I'm going to look at the one here to better advise.
                veritas odium parit

                Comment


                  Re: All replacement capacitors in iMac G5 PSU burned up

                  First, let me say

                  Now, the first part of this is to get that thru-hole piece back into the board. Carefully get it realigned and back in. Don't flatten the trace or do anything else, yet.

                  You'll need -
                  a pin, the longer the better. (a hat pin would be great)
                  soldering iron on, cleaned and retinned ready to go

                  Once the piece is back in the board, hold it down in the center of the hole where the cap lead is stuck with the tip of a pin. Apply the soldering iron next to the pin and apply gentle pressure to the pin. You should be able to push the cap lead out and the hole should now be clear. The solder won't stick to the pin.

                  LMK when done and I'll get to next part.

                  Carefully get as much of the gunk off the other half of that caps area. Try not to chop the other traces.

                  Really is a hijack and should be in its own thread.
                  veritas odium parit

                  Comment


                    Re: All replacement capacitors in iMac G5 PSU burned up

                    The one that nwd did not replace originally because of it's difficult access is a TEAPO SC series, 330uF @ 35v, 10 x 20mm. According to the SC datasheet it should have an ESR of 0.044.

                    Out of circuit it reads 0.47 ohms on the ESR meter @ 320uF. Capacitance is okay (although I like them a little high 5-15%), but ESR is nearly 12x over spec.

                    That should say it all.
                    no, if you replace it and the coil and everthing is fine, it says it all.
                    esr of 0.47ohm in psu doesn't tell me much at all....could be a problem if it's on very sensitive place, but if it's just filtering, no, i wouldn't expect it to make problems...
                    that's why i said to change coil first(and monitor it's temp when loaded): if you change both coil and cap you won't know what was it...

                    will it be finished this month? change of one cap and coil, i mean...
                    <wink>

                    i like bonez avatar..i should make one from this footage...
                    http://www.youtube.com/watch?v=LZxmDJYBKKo
                    has more bang for the buck..hehe

                    Comment


                      Re: All replacement capacitors in iMac G5 PSU burned up

                      I will get to it after I get his new "used" PSU recapped. The info from that will start to appear in a new topic in a few days.

                      I will be testing the old PSU several ways, when I get the time. It is not a priority.
                      veritas odium parit

                      Comment


                        Re: All replacement capacitors in iMac G5 PSU burned up

                        Alright, here is my final update.

                        Here is a complete table of the caps in this PSU, any notes I have about them, and replacements (with more notes) that I used:

                        Code:
                        Primary Caps: 
                          560 uf  /  200V  /  18mmx45mm, 8mm   /  2x /  ECO-S2DP561BA (22x40mm, 10mm, Snap-In)  
                            ^ Radial
                          47 uF  /  50V   /  6.3x11mm, 2.5mm   /    /  UPW1H470MED (PW)
                          10 uF  /  50V   /  5x11mm, 2mm     /    /  UPW1H100MDD (PW)
                          2 uF   /  50V   /  5x11mm, 2mm     /    /  UPW1H2R2MDD (2.2 uF, PW)
                          100 uF  /  35V   /  8x11.5mm, 3.5mm   /    /  UPW1V101MPD (PW)
                            
                        Secondary Caps:
                          1000 uF /  35V   /  12.5x20mm, 5mm   /    /  UPW1V102MHD (12.5x25mm, PW)
                          330 uf  /  35V   /  10x16mm, 5mm    /    /  EEU-FM1V331 (FM)
                            ^ 0.044 ESR, 1050 ripple
                          1200 uF /  16V   /  10x25mm, 5mm    /    /  UPW1C122MPD (PW)
                          2200 uF /  10V   /  10x25mm, 5mm    /  2x  /  EEU-FC1A222L (10x30mm, FC)
                          4700 uF /  6.3V   /  10x30mm, 5mm    /    /  EEU-FM0J472L (12.5x30mm, FM)
                          1000 uF /  6.3V   /  8x15mm, 3.5mm    /  3x  /  ELXZ6R3ELL102MH15D (LXZ)
                        
                        Secondary Inductors:
                          6.5 uH  /  2100LL-820-H-RC
                        All of the above fit in the PSU, although some things are tighter then the originals, or need extras leads soldered on them. The inductor needed some amount of cleverness to get it to fit (see attachment).

                        In the end, the I got them all soldered in, got my previous mistakes fixed, and got the lid back on and closed. Plugged it in and verfiyed that all the pins were putting out their advertised voltages. Thus I declared the PSU fixed!

                        Sadly the iMac it goes in still will not fully boot up, maybe the rails are sagging under load, maybe something got nuked when the PSU had bad output, don't know. Cutting my loses at $20 worth of caps.

                        Anyhow, thanks for everyones help.
                        Attached Files

                        Comment


                          Re: All replacement capacitors in iMac G5 PSU burned up

                          Verify the voltages with it installed and running.

                          There are 2 separate grounds. On the connector, pins 1, 2, & 12 are the "floating ground" through the burned coil. Pins 3, 5, 7, 11, 14, 16, 17, & 18 form the other "common" ground.
                          veritas odium parit

                          Comment


                            Re: All replacement capacitors in iMac G5 PSU burned up

                            I've been following this thread for a while, since I have a 614-0293 that doesn't work. It had a few bulging caps and a toasty coil. I acquired a recap kit as well as the coil discovered in this thread. (Nichicons from Mouser and the coil from digikey).

                            Anyways, I recapped it & replaced the burnt coil. Shorted the power-on wire. All the voltages checked-out okay. I installed it in my iMac, and it wouldn't turn on. The power light blinked for a moment, but no boot sound. I reinstalled my known-good PSU, and everything was fine.

                            If there's a quick-fix, I'll keep trying, but we're way beyond my skills now!

                            So, clearly there's something else wrong with this. If you're interested in it, PM me.

                            Comment


                              Re: All replacement capacitors in iMac G5 PSU burned up

                              @iointerrupt & TahoeEd - If you cover the round trip postage and any required parts (with your approval), I'd be happy to check your unit(s) out. I have nwd's old unit here to compare them against.

                              If you still have the old parts, sending them along is a plus.

                              Reply here or PM if interested.

                              Toast
                              veritas odium parit

                              Comment


                                Re: All replacement capacitors in iMac G5 PSU burned up

                                Toasty -

                                Sure, if you feel like playing around with the PSU. I have nothing to lose at this point. If you prefer, email me your shipping address (iointerrupt -> gmail.com) and I'll send it out along with the old parts.

                                Comment


                                  Re: All replacement capacitors in iMac G5 PSU burned up

                                  Hey folks,

                                  I've got an iMac G5 17" power supply which also has a burnt L8. I recapped this supply a couple of weeks ago, L8 looked good but the power supply quit working and it's back in my hands again.

                                  Before I pull this supply apart, trace out a schematic and test every active component in the supply... has anyone figured out yet what destroys this coil?

                                  Comment


                                    Re: All replacement capacitors in iMac G5 PSU burned up

                                    Please confirm the model number on the supply. 614-0293 ?

                                    L8 feeds the 3.3 volt line to the mobo. From L8 it connects across C52 (1000/6.3) thru ferrite rod choke L10 and across C56 (2200/10) to the output.

                                    -What brand and series caps did you use to recap the supply?

                                    -Did you replace (sans mains) ALL 13 of the electrolytics?

                                    -Does the supply turn on with no load (jumper 15-16) and give the correct voltages?

                                    -Do you have any means, other than a system, to load test the supply? Low resistance-high wattage resistors?

                                    I have not yet had the opportunity to dig into the supplies members have sent. I do not know for certain why they take out that coil. The problem would seem to lie somewhere in the output circuit from that coil. But, given the components that lie beyond it, as mentioned above, it hardly seems likely.

                                    At this time, I would be looking at the supply circuit prior to the coil. If the chopped voltage is running at the wrong frequency, it could be what is causing that coil to saturate and overheat. The TOPSwitch chopper should be running above 60kHz and commonly is 130kHz.

                                    L8 is fed from the center leg of the Schottky barrier rectifier D33 (MBR2045CT) on the backside of the heatsink. That in turn is fed thru a nearly identical coil as L8, L6 which is supplied from the transformer T1 (the large one).

                                    I did a bit of research and am going to say that T1 appears to be operating as part of a push-pull converter. Starting with the 2 transistors Q12 & 13 in front of T1 which alternate driving T1's primary.

                                    FYI - chips on board:
                                    TOP244Y is a TOPSwitch switcher
                                    3845B is a PWM
                                    LM393M is a dual comparator
                                    WT7515 is a protection IC that monitors 3.3v, 5v, 12v for over/under voltage

                                    Good Hunting!
                                    Toast
                                    Last edited by Toasty; 05-18-2009, 12:14 PM.
                                    veritas odium parit

                                    Comment


                                      Re: All replacement capacitors in iMac G5 PSU burned up

                                      It's a 614-0293. I replaced all electrolytics on the secondary side of the supply, 9 in total. There are four other electrolytics on the primary side, plus the two mains caps, I didn't replace these. I used a mixture of Nichicon HE and Panasonic FM for the recap.

                                      Eyeballing the layout, it looks like the TOP244Y is used for +5.1VSB - this is a different circuit from L8's, so I'm not concerned about that.

                                      The main rails are run off the UC3845 switcher. I haven't traced out how the primary side works. On the secondary side, L1 is the coupled inductor used for +5V and +12V. It looks like they're taking the +5V winding off T1, running it through a series inductor (L6) to drop the voltage (I've seen this trick in Antec PSU's), rectifying that (D33) and L8 is the +3.3V output inductor.

                                      I'm still scratching my head how a failure can occur which causes *one* coil to pop. If the switching frequency goes off, I'd expect different carnage - too high a switching frequency and I'd expect snubbers to catch fire, main FETs/output schottkys to blow, etc... and they're fine. Too low a frequency shouldn't matter since this is a current mode supply, it'll most likely cut its switching cycles short, working fine under light load and dropping dead under heavy load. Regardless, switching frequency is set by a RC circuit on the '3845, this can be measured and checked easily.

                                      Since the varnish is fairly badly burnt off L8's wire, I'm thinking it was overcurrented. I'm wondering if the G5 itself has a crowbar circuit built in, which shorts +3.3V if it goes too high. With this happening, the supply may put out a high fault current through L6/D33/L8/L10, and perhaps L8 is the first thing to catch fire. Though I'd expect L6 to also get damaged...

                                      Anyway, that's my brain dump. As I dig into this thing over the next couple days, I'll post more.

                                      Comment


                                        Re: All replacement capacitors in iMac G5 PSU burned up

                                        UC3845 is dumb old SMPS IC and not fast enough to sense trouble. And came from same company who designed original UC3482.

                                        This type of crappy IC will keep trying (ticking) and blow up (not fast enough on some situations especially with bad caps) and do strange things too. There is electrolytic caps that is very important for proper function of the UC3845.

                                        The reason the coil burned is off frequency during Mac is in use due to damage or remaining bad caps not replaced. Low frequency will do that to a coil.

                                        For this reason I particularly prefer good SMPS that *do* not have electrolytic to filter DC just for the IC itself. Poly film or ceramic and cycle to cycle protection and cut out in a cycle and stay shut down and have to cut power to reset. ST and other good IC smps designs are available.

                                        Also WHY did apple use this particularly awful SMPS that won't let you fix without more troubles? This is seems to be one of those all or none PS that need fixed all the way to prevent repeats.

                                        Cheers, Wizard

                                        Comment


                                          Re: All replacement capacitors in iMac G5 PSU burned up

                                          Reflow top and bottom of board connections to T1 & L6.

                                          Let me know.

                                          Toast
                                          veritas odium parit

                                          Comment

                                          Working...
                                          X