Suspect Bad PS Apple AcBel API 1PC36

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

You're right about it being the green wire. Here's another pinout for future reference. http://www.jcsenterprises.com/Japama...EB2C04619.html I think I can take the 'jump' from here.

Having a strange boot sequence now .... as in long and slow boot (like 5 mins). In the process of checking ram atm.

Is there any way you can think of that the PSU could impact the boot process?

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Hey Toast .... just checking my resources .... the computer crapped out again. Checking the pinout and it points back to the PSU.

The computer ran for (idk) maybe 10 - 12 hrs but did have some issue on start up. Turned out the Firewire was dead and after removing the associated .kext files it would start up normally again. Then one day noticed the cpu temperature was on steady climb and before I could shut down it shut down on it's own. Blink - quit. Opened it up and a spare molex was jambing the main fan. Got a new fan and installed it and no go - computer wouldn't start - identical as before, mobo light flashes, fans twitch.

I took the PSU out again and (what do I know) it 'looks' ok. There isn't anything standing out.

Any thoughts?

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Will it run if it is disconnected from computer and jumpered on?
Are the standby voltages present?

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

I have already removed it and disassembled it which included removing the plug in to get the board from the case. If necessary I'll solder the AC plug back on and try the jumper test.

I checked the voltages prior to removing it and got the 25v and 5v results but none of the rest. Used this guide .... http://www.jcsenterprises.com/Japama...EB2C04619.html

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Again, will it run if it is disconnected from computer and jumpered on?

If it runs outside, then the problem may be the machine, not the PSU.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Ok - I'll get the plug soldered back on and try a jumper test on the bench (desk).

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Till I can reassemble the plug to the board I can't do the jumper test and my amateur soldering kinda fried the insulation on the leads from the board to the plug. Not to mention the 'connectors' on the plug are a mess of solder. I'd replace the plug and leads if I could find a replacement plug and it isn't any worry to replace the leads. (?)

Otherwise I could 'cobble' the plug and wires together to test, but it would certainly be outside the box.

Been all over town and no wick anywhere, gonna have to order it online.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Ok .... jumper test result!! I managed to 'cobble' things together and ...

Plugged it in ... .5 to 1 second later the fans start .... fans spin for about 1 - 2 seconds and stop. Unplugged and repeated half a dozen times .... same result each time.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Check opto M8. Measured from output ground, the first pin on the "cold" side (square pad on solder side, or lower left in your above picture) should read 5v. Pin next to it will read 5v until the supply is jumpered on. It should then read 0.1v or so. If it reads near 5v (i.e. 4.3v), then the OV/OC protection is kicking in.

That would indicate that perhaps one of the replacement caps is bad. Do go over all your solder work to be sure a "cold joint" is not causing this problem.

If the fans in the supply were functioning, I don't see how the system fan blockage would cause a problem, unless the whole unit got very hot...?

Since you are working without an isolation transformer, please remember that everything on the other side -and- to the left of those opto's is "HOT" in regards to mains power.

Toast

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Toast .... let me know when you've reached your limit on 'hand holding' or if you've put enough time in to this. I appreciate your help but lack considerably in 'understanding' electronics etc. That said ....

I attached a photo of 'opto M8' that I think you are referring to. I'm unable to understand the instruction for checking voltage on this. Is this test done on a 'hot' board? or at the connector where it's jumpered? (first pin / cold side / square pad?)

I've re-examined my soldering. Is a cold joint where you have a lead soldered solid but no solder down the hole? No question some of my joints are not as uniform as the originals - maybe I should touch them up (a little heat and a touch of solder).

The fans were working without question. I checked the warm output many times while the machine was running. I had a thermometer in the case placed right at the intake to the PSU and the hottest it ever got was 118ºF. The software TempMonitor was reading the CPU and it was at 72ºC when it shut down. I sat and watched it rise from normal to 'hot' - so it wasn't 'hot' for more than a few moments.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Yes, hot board.

Measure Pin 1 & Pin2 on the opto when hot but -not- jumpered on. Then jumper the supply on and check Pin 2. It should be around 0.1v or so. If it goes high as I said, that could mean the OC/OV has kicked in or that the OC/OV chip has failed.

Cold joint = poor solder joint. It moved after you soldered it, it cracked, it wasn't hot enough to flow solder correctly, the solder didn't stick to something... Pick one...

Just check all the joints you did.

Try to monitor the 3/5/12v lines when you jumper it on. See if anyone looks low or high. I know it's tough with a DMM and such a short time span. Most react too slowly.

Also, make sure you didn't put a cap in backwards. It happens.

Toast

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Ok, now I know which pins you want checked - I'm not sure where to check from (i.e. where does the black lead touch to)

Duh!!! ... just a little dense so .... hot board no jumper pin 1 5v and pin 2 4.65v.

hot board jumpered pin 2 4.44v

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Post 49

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

idk .... guess we (I) could be at a dead end. I have gone over the solder joints and touched up any that appeared to possibly be troublesome. The voltage checks on the 2 pins remain as before - before the jumper pin 1 is 5v and pin 2 is ~4.4v then after jumping pin two remains at ~4.4v.

My soldering skills still remain a variable in this ..... /shrug ..... although initially after the new caps it did work for 10 - 12 hrs (rough guessing).

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Okay. Do you have a 1kΩ resistor about? Something in the 680Ω-1kΩ range is fine.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Is this something I could find at Radio Shack

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Umm, yeah.... but don't you have some old boards lying about?

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

All I have laying around is some video and pci cards that still work. I have a buddy who owns a small computer shop and I'll be in his neighborhood this morning, I'll stop in and see if he has something laying around.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

So I struck out and wound up at the Shack .... this is what they had --

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Fine.

We're going to 'fake out' the PSU for testing.

Carefully solder a piece of wire, 8" or more, to pin 2 of M8 on the solder side of the board.
Solder the other end of the wire to one end of the resistor.
Jumper the supply on as usual. (Pin 11 - green wire to ground)
Touch the other end of the resistor to any of the black (ground) wires on the PSU output plug.
Measure the voltages for 3.3/5/12 and report the readings.

If they all are fine, then the supervisor IC has likely failed.

On the "daughter board" on the output side (where the fans plug in) is where the supervisor chip is. Looking at the daughter board from the side, notice that the it is attached to the main board with a single line of 17 pins. Measure the voltage on the 5th pin from the left, with the resistor jumper connected and disconnected.

Report.

Toast