Suspect Bad PS Apple AcBel API 1PC36

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Check all the connections to drives, leds, motherboard, etc. for a tight fit. The unit is -very- sensitive to power interruptions. Senses an imbalance in the load.

I believe you hit it Toast. I've solved the over heating problem. Had a non-functioning fan and put new thermal paste on the cpu chips. The unexpected shut downs however are still an issue. I'm 98% certain the problem is in the PSU connector at the mobo.

I've visually inspected and probed the pins in the connector from the PSU and they 'appear' fine. Some vertical movement but only slight till each hits a stop. The mobo side is rock solid pins and receptacle.

After starting it up, I used my probe to gently press laterally on each wire and found 5 or 6 that would stop the machine. A couple of them I only barely touched and it shut things down.

Hoping I'm right, and thinking I have an idea. Just not sure how to implement it. I think I need to crimp these female ends somehow. Make them hug the male end better. There was enough 'poking and prodding' on these that they may have widened and become sloppy.

Any ideas?

In any event, thanks again for all your help.

[Pyr0Beast]

Re: Suspect Bad PS Apple AcBel API 1PC36

Use a needle and 'crimp' those ends.

Or rotate the connector for a whole circle and plug it in, this will provide some torque on the pins, making them connect well

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

I like the 'rotate idea. I used a small jewelers screwdriver that fit and pried the connectors tighter. They are like a small square tube that is split. I noticed on a few of them they had spread quite a bit. The prying idea helped but don't think I can get it back like new. Now if I can throw a twist in it - the torque may add the part I can't get.

I was at a point I could not close the case and then start the computer. Could only start it opened and then only after manipulating the wire bundle. Once started I couldn't close it because it would 'short' and quit. After prying the connections tighter - at least now I can close it and start it. Tried opening it while running and it 'shorted' - so my statement that I can't get it back like new.

Thanks Pyr0Beast.

[Pyr0Beast]

Re: Suspect Bad PS Apple AcBel API 1PC36

Uff. This is pain then. I've repaired many that way, by squeezing the pins, it however works only once or twice and afterwards it is the same or not worse than before. Problem is that many cheaper connectors just don't have the 'spring' in them - cheap material and will eventually make the connection loose. Best thing to do would be just to cut it off and solder a new one there.

Oh, and if you did notice. You may help it a little by individually pushing the wires inside the connector further down, but that is of no guarantee for long term work stability.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Yeah - pushing the wires in tighter helps, but only momentarily. Slightest movement and they 'wiggle'. I may work the connector ends again. If I could torque them a little to the side so one side bypasses the other would help. Prying till they meet is marginal but better. Haven't tried your twist idea yet.

[Pyr0Beast]

Re: Suspect Bad PS Apple AcBel API 1PC36

I would personally call it 'a shitty design'

Older AT connectors were MUCH better IMO. Had their own springy connections, like molex connectors have for fans.

Hdd conectors cause similar issues. Floppy drive connectors - never seen one do that.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

It's always a matter of them saving a few cents here and a few there - such as it is today with everything today

I reworked the pins and threw in the twist. Started it up with the case open. Once it was booted I closed the case. Hurray!!! it's still running.

Because of Toasty's efforts and the rest of you at Badcaps I have my computer back. Of course I had a cooling issue to iron out, but fixing a stopped fan and some new thermal paste on the cpu chips - this computer has never been this cool and quiet.

[Pyr0Beast]

Re: Suspect Bad PS Apple AcBel API 1PC36

Sweet. Let it serve you well.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Use a straight pin or push pin on one side then the other of each female connector pin. Between the housing plastic and the connector. You want to do it on either side of the "split" not on the "split" side or the side opposite it.

Squeeze it this way ---> [_] <---

Using a pin gives you much greater control than a pair of tweezers or pliers.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Nice pin graphic Toast.

I've worked the pins twice now, each time with a tiny jewelers straight slot screwdriver. I think I have them as good as they are gonna get.

I'm 'pressure testing' the computer as I type this. I was told a computer vs computer chess game x10 will ask everything possible of a processor. It's in it's third match and the processor temp is steady at ~60º so it's doing great!!!

It's appearing this computer is up for another round of work (or play) . I've had it since it was new back in 02 or 03 and I think it's working better now than when I got it.

There's a ton of info in this thread, hoping it will pay dividends to someone someday. You guys here at Badcaps deserve a salute.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Just keep the A.E.D. nearby....

CLEAR !!!

[uffda_oi]

Re: Suspect Bad PS Apple AcBel API 1PC36

These caps will arrive tomorrow and my AcBel is desoldered and ready to receive. This list was clipped out of the Digikey confirmation file and doesn't give the specs. I used this thread as a guide for cap specs and hope I got it right.

1 493-1578-ND UHE1V221MPD CAP ALUM 220UF 35V 20%
1 493-1547-ND UHE1E470MDD CAP ALUM 47UF 25V 20%
1 565-1532-ND EKY-160ELL222MK25S CAP ALUM 2200UF 16V 20%
1 493-1473-ND UHE0J272MPD6 CAP ALUM 2700UF 6.3V 20%
2 493-1472-ND UHE0J222MPD CAP ALUM 2200UF 6.3V 20%
1 565-1651-ND EKZE100ELL332MK25S CAP ALUM 3300UF 10V 20%
1 565-1694-ND EKZE350ELL471MJ20S CAP ALUM 470UF 35V 20%
1 565-1579-ND EKY-350ELL681MJ30S CAP ALUM 680UF 35V 20%
1 493-1548-ND UHE1E101MED CAP ALUM 100UF 25V 20%
2 493-1496-ND UHE1A102MPD CAP ALUM 1000UF 10V 20%

Only C60 was bulged. The decapped AcBel board is pictured below. The wiring at the bottom of the image (angles down from left to right) looks like it might have carried a surge. It may be nothing.

[uffda_oi]

Re: Suspect Bad PS Apple AcBel API 1PC36

p.s.

I learned a little more "caps-speak" this a.m. So all you savvy and fluent caps-speakers know I bought UCC (495) series KY or KZE and Nichicon (493) series HE caps.

[uffda_oi]

Re: Suspect Bad PS Apple AcBel API 1PC36

I replaced the 12 caps recommended in this thread as follows:

c7 (220/35), c8 (47/25), c10 (2200/16), c13 (2700/6.3), c16 and c29 (2200/6.3), c21 (3300/10), c35 (470/35), c41 (680/35), c43 (100/25) and c59-60 (1000/10).

The leads to the yellow box connect to the plug posts as do the brown and blue wires. I marked the color code on the components and re connected in the original pattern.

The light bulb test on the AcBel psu produced nothing but 4 bright blinks per second. I expected it to go bright, go dim, blink a few times then go out. Confidence is low for testing the PSU in the computer.

A view of the resolder job is included (pretty ugly in close up--gonna redux).

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

What wattage bulb are you using? A 60w won't turn this on. It needs to be large enough to charge the main caps. I start with a 100w. If it blinks as you say, then use 2 in parallel for 200w. You can also use a combination of wattages to step up to it slowly. Blinking is a good thing. It means it is trying to work.

Toast

[uffda_oi]

Re: Suspect Bad PS Apple AcBel API 1PC36

Toasty

I used a 100 watt bulb and will start stepping up the wattage (in parallel).

Thanks for the suggestion.

[uffda_oi]

Re: Suspect Bad PS Apple AcBel API 1PC36

Re: #115 and #116

I am pleased to report that my AcBel (in Apple G4 MDD) is now working like a champ, mainly thanks to Toasty's tutoring style and expertise. The search tool is a friend, indeed. So is observation of previous sessions with other folks with similar hardware problems. However, my noobie path was perilous, even ridiculous at times.

I assembled a two-bulb parallel circuit to persuade the AcBel to start, forgetting that the PSU itself was one of the parallels in the circuit. I fastened the clips together as I had done to complete the series circuit and was rewarded with a loud pop in my power strip followed by deadness. The pop came from rapid vaporization of about a 1/4" of circuit trace inside the strip.

The second dry run of the test circuit succeeded using another power strip. Instead of steady blinking as in the serial circuit, the bulbs shined steady bright except for a barely perceptible blink about three seconds into the test. Since nothing fried, crackled or smoked, I put the PSU in a computer I knew worked (not like the last test).

The PSU work fine. Had I done this in the first place, a couple of R&Rs could have been avoided. But, my R&R speed is much better and my recap success rate is now 3:4 PSUs.

This experience also gave me a chance to repair a PCB in the power strip, also successful. It is now time to segue to an Apple G4 MDD Samsung thread to find out why PSU #4 is still not working.

[thestitch]

Re: Suspect Bad PS Apple AcBel API 1PC36

Hi Toasty,

considering the same problem as buz in this thread #99
(Suspect Bad PS Apple AcBel API 1PC36), followed your tutorial,
replaced all the mentioned caps and have a working PSU again :-)
But sometimes the PSU won`t start up. It just makes a tickling sound
if pluged in.
Can you tell me which cap was faulty or replaced at buz`s PSU ?

Thanks in advance
willy

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Nope. Don't remember.

Looks like there are only 3 choices anyway - C30, 32, 45.

C30 - 0.47@50v
C32 - 47@25v
C45 - 10@50v

Toast

[thestitch]

Re: Suspect Bad PS Apple AcBel API 1PC36

Hi, Toasty !

That is a fast response.
I`ll replace the 3 caps and give feedback.

Thanks
willy