I mac G5 PSU, ALS - odd voltages...

[trex66]

The guy I bought this Imac from re-caped the PSU, couldn't get it to work and just gave up. I found two dry solder joints and it brought the Imac to life. After a while it would just shut off. Sometimes after post, sometimes in a day, or a week or two. I couldn't repeat the shut down and there was no log on the computer of a shut down. I pulled the PSU and checked the voltages again and they were all a little high. the 3.3V were 47V when it was off. Also I noticed the pins that supply current when it is turned on continue to increase in voltage. As I write this I think I should have just re-caped this and re-tested. here are pics and voltages for the pins, any thoughts?

[retiredcaps]

Re: I mac G5 PSU, ALS - odd voltages...

the 3.3V were 47V when it was off

4.7 or 47V?

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

Yup, forty seven (47V). I wish I knew some history on this PSU.

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

Please stay out of the other thread until the PSU problem is fixed. Posting in both will only be confusing.

There is no 3.3v standby power. When the unit is plugged in and not switched on, the only voltages should be 5vsb & 24v. All others should be at zero with the exception of the "power on" line which will have ~5v on it, but there's no real "power" there. Just a potential provided by the power switch circuits.

Verify the (1) values on the caps that were replaced and (2) their positioning/polarity by using the other threads in here on those PSU. A few threads show the cap layout and values, and a few off-site links too. I have that board here in a unit that I will open later on and check too. The PSU model doesn't matter as much in this case as does the board number: API4PC47

The standby circuit provides 5vsb & 24v through the small vertical transformer. The 3.3, 5, & 12 come from the horizontal transformer next to it. If the 3.3v is 47v when off, then the other lines should be showing a voltage also. I think this points to a short or other connection to/from the 5/24 standby side. Check those traces for a solder bridge from the cap replacement job.

Note that 47v is about double 23.9v...

Toast

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

OK, the caps are right polarity and specs. I can't find any close ups of the end of my PSU to see what it is supposed to look like before it was butchered. I really need to see what this looked like so if anyone has a API4PC47 out of the case and can post a pic I would be eternally grateful. I'm not an expert but the 12V and 3.3V lines don't look like they are close enough for a solder bridge to short them. I traced the 47V back from the 3.3V wire out to a section I marked. Above it is, I think, a mosfet bolted to the copper heat sink. Can I test this in board or do I have to take it off?

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

You're missing the point. There is no voltage produced by that part of the circuit until the 'power on' wire is grounded. Whatever you are reading there is coming from the 5/24 volt circuit.

Is there any "power" to that 47v? Does it light a light bulb? I'm thinking it might be an induced voltage from T2 and that a ceramic or chip cap is damaged, has a cold solder joint, or missing. The cap would be across ground to this trace to suppress any induced voltage as these trace go top and bottom of the board, everywhere.

The top 47v is connected to the middle 47v by the vertical ferrite rod coil "in front of" the blue toroid coil. The middle 47v is connected to the lower 47v by the blue toroid to the center pin of the common cathode dual Schottky diode pack, D23.

Where are you measuring the 47v from? Where is your meter negative (-) connected to? I want to try and duplicate it here.

Toast

PS: Q1 ( marked:1P ) is bad. That's part of the start-up circuit. Underside at T1.

.

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

So that Schottky can't be leaking voltage, it has to be from the 24V circuit? For Ground, I used the ground posts that lead to the P1 connector (In the first pic above they are the 'G's). I traced the 47V back from the connector wire to the dual Schottky, and it seems to end there. I couldn't find any other contact that had any voltage close to that. My DMM reads 0.29mA from ground post to all of the 47V I marked above. I don't know what T1, T2, Q1, 1P are?

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

I'll check mine tomorrow to see. Your meter should never be placed -across- power when reading current. Inline only. Doing what you did could either blow the internal fuse (if your lucky) or burnout the meter. Fortunately the current is <1ma, so no harm done.

That 47v does seem like an induced voltage from T2 or something nearby that 24v section.

The part ID's are marked on the board. T1 - main transformer. T2 - 5/24 SB transformer. Q1 is under T1 on the solder side, next to R4. You'll need a 10x or better jeweler's loupe to read the numbers on it, but you'll see it's 1P. There are several more near it. That transistor has been discussed in the other thread and in some others. If you look at it, you'll see the difference and how I picked it out.

I'm working on finishing a system to ship out Monday, so I may not poke my head back in until late tomorrow. The hardware's easy. The software is just time consuming...

I think the same individual worked on yours and mine. The Sanyo caps are very welcome, but I'm jealous because I can't lay hands on them from the usual sources.

Also, plan to replace the 4 little caps on the 'HOT' side as they run & run with Q1 and the TOPSwitch.

Toast

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

I'll get the four caps, and I need to replace the P1 because the 'off' voltage is 24V and should be 20V? Do I also need to replace the 2F transistor, you said they work together?

The thing I was most concerned about was the rising voltages, is this normal for a PSU under no load? Something keeps making this PSU turn off, wouldn't a voltage that is out of range have the PSU shut down?

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

Hey Toasty, I tried and tried but I just can't figure it out, could you please point out which F2 NPN is paired with Q1

Also, I circled two areas I noticed that looks like it may be unintentional solder bridges. The only other pics I could find were on this forum and only the 20" PSU was high enough resolution for me to see. These solder bridges weren't there but it is slightly different.

[godonr]

Re: I mac G5 PSU, ALS - odd voltages...

Toasty- those Sanyos probably came from a guy named Jim Warholic (http://jimwarholic.com/category/capacitors).

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

Toasty- those Sanyos probably came from a guy named Jim Warholic (http://jimwarholic.com/category/capacitors).

Thank you.

Yeah, I forgot about his site. But, great caps don't make a good job. Both Trexx66 and mine look like crap soldering jobs. Likely inexperience and not a powerful enough iron. These boards suck heat big time.

Toast

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

I'll get the four caps, and I need to replace the P1 because the 'off' voltage is 24V and should be 20V?

It's 1P. That is not the reason. The transistor is -visibly- damaged, and is the source of problems as mentioned in other threads.

Do I also need to replace the 2F transistor, you said they work together?

I never said that.

The thing I was most concerned about was the rising voltages, is this normal for a PSU under no load? Something keeps making this PSU turn off, wouldn't a voltage that is out of range have the PSU shut down?

- No, not really.

- Yes, it would.

Also, I circled two areas I noticed that looks like it may be unintentional solder bridges.

No, they're correct. (factory)

You should be looking around where the caps have been replaced.

I would replace the transistor, the 4 small caps, and re-do the soldering on those caps that were replaced, just to clean up the lousy looking job. Just be sure to use a good iron of at least 45w with a medium or larger chisel tip.

Toast

[smason]

Re: I mac G5 PSU, ALS - odd voltages...

Thank you.

Yeah, I forgot about his site. But, great caps don't make a good job. Both Trexx66 and mine look like crap soldering jobs. Likely inexperience and not a powerful enough iron. These boards suck heat big time.

Toast

I had trouble with mine too. I checked my iron twice wondering WTF?

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

I pulled the four little caps tonight, they tested pretty close to specs.

2.2uf 50V = 2.3uf
33uf 25V = 31uf
100uf 50V = 103uf
47uf 25V = 47uf

Toasty, trying to do my homework - the reason I mentioned the 2F is in another thread you had said that P1 and 2F work together and advised to replace both. If you don't think I need to then I won't.

I've got the caps and transformer ordered. While I wait I can try to clean up the soldering mess. There are some places that I might have to wick up all the solder because I have no idea what's under those chunky fields of silver.

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

?? -1P- stop with the P1 already! Actually, you should be calling it by its circuit identity, Q1.

If you're going to talk about what I said in another thread, you need to quote it here, so I can review it.

What was the date on the other thread? How long ago? Do you understand how much time has passed and how many of these have passed through my hands and these forums??

Which 2F? They're all identified.

>>I've got the caps and transformer ordered.<<

Transformer? What transformer?

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

Sorry Toasty, it was late and I'm exhausted - I ordered the transistor @ Q1.

Here was your post(1/6/11):

Re: iMac P/S posts then shuts down

Yes. You need to do the research to learn to find these.

Look for a transistor marking code reference. There are several out there...
(This is actually one of the easier ones.)

Case style should be SOT-23-3.

Good Hunting!

Toast

PS:
While you're digging, look for the 2F transistor too, as that is the PNP complimentary one to the 1P.

[Toasty]

Re: I mac G5 PSU, ALS - odd voltages...

Which simply states that the 1P complimentary transistor is the 2F. Nothing about in this circuit, just a statement about the 2 transistors characteristics.

Transistors are said to be complimentary when they are the same in their packaging and in their electrical characteristics, except one is an NPN and the other is PNP. Or, in the case of MOSFETs, one is N channel, and the other is P channel.

Datasheets will often be for both transistors when they are complimentary.

Do some Googling to understand better.

[trex66]

Re: I mac G5 PSU, ALS - odd voltages...

OK, replaced the caps, transistor and cleaned up the solder a little. I get the correct voltages and they don't seem to be rising anymore.

But I have a new problem, there's no standby, the PSU is always on. The gray power on/off connection at the board is untouched. Not sure if it's something I did or replacing the components revealed a new problem.

[godonr]

Re: I mac G5 PSU, ALS - odd voltages...

I can tell you pin 15 (on/off) sits at 3.3v (off) and drops to 0 when the supply comes on.