iMac G5 PSU - no power

[seanc]

Re: iMac G5 PSU - no power

Extra caps are good, they can easily go in all of this junk I keep bringing it home.

I've fixed four LCD monitors, a 24 port switch, three 5v power supplies for 2wire routers so far. I may not have used the best caps in them, but I'm learning at least.

I'm linking the pictures, because then you can download them as they are and zoom in.
http://i100.photobucket.com/albums/m...7/P1000181.jpg
http://i100.photobucket.com/albums/m...7/P1000182.jpg
http://i100.photobucket.com/albums/m...7/P1000186.jpg

The last picture shows the area that I have the problem with.
I've not got Photoshop on here to highlight the area, if you can't identify it, I'll have a bash on another machine.

[Toasty]

Re: iMac G5 PSU - no power

Will look later.

Please don't use Photobucket. Links disappear shortly and next user won't have them to refer too. Just upload them from your comp to the site. If too big, open them in PaintShop or similar and "Save As:" a .jpg and use the option to set compression to 20%. Usually solves the "file too big" problem.

Toast

[seanc]

Re: iMac G5 PSU - no power

Do links disappear from posts as part of the forum mods? Interesting...

[Toasty]

Re: iMac G5 PSU - no power

Will look later.

Please don't use Photobucket. Links disappear shortly and next user won't have them to refer too. Just upload them from your comp to the site. If too big, open them in PaintShop or similar and "Save As:" a .jpg and use the option to set compression to 20%. Usually solves the "file too big" problem.

Cap is a decoupling cap between 12v & GND. Get it back in as best you can but don't be too concerned.

Clean the board pads with solder-wick and put just a drop on each of solder. Position cap and use a probe to hold it in place. Clean the iron and a touch of fresh solder to it then just "touch" the good end and remove iron. Repeat on opposite end. You should catch enough of what remaining plating is there and be fine. They are tough to kill.

Toast

[Toasty]

Re: iMac G5 PSU - no power

Links don't disappear unless the linked site does. i.e.- Geocities went away and several items followed off the boards. Always go "Advanced" and attach files. Unless the server takes a major dump and there's no backup, the pics will be here forever.

Cheers!

[seanc]

Re: iMac G5 PSU - no power

Thanks Toasty for your advice - it is very much appreciated.

Time to go spend some money with RS Components.

I got a sealed bag of 5 caps, with 4 in it. I mentioned it to them, expecting nothing, but got a free bag of 5.
I'm sure they don't really cost much to them, but it's the impression of good customer service it gives that's the important thing.

[seanc]

Re: iMac G5 PSU - no power

I have filled my cart, but one line gives me concern.

1000 / 10 - have P/ FM - 8x20 - 1560ma(2)- UCC KZE 7x18 & Taicon ? spec
EEUFM1A222
http://uk.rs-online.com/web/search/...oduct&R=5261165

You've suggested a 2200uf 10v rather than 1000uf 10v.

Typo? or are you suggesting I upgrade?

In case not, I've listed two I found below.

EEUFM1A102
http://uk.rs-online.com/web/search/s...duct&R=5261109

or

EEUFM1A102L
http://uk.rs-online.com/web/search/s...duct&R=5261115

I'm leaning towards the EEUFM1A102L since it's the same size as the old one and has 1560 ripple rather than 1790 - am I correct in thinking lower ripple is better?

[Toasty]

Re: iMac G5 PSU - no power

Yup, that was a cut-n-paste of info that I doubled somehow. Fat fingered it, I guess.

>>am I correct in thinking lower ripple is better?<<
No. HIGHER ripple current and LOWER ESR is the correct answer. It is the quest when repairing PSU's & mobo's.

If you can fit the 10mm, go with that one. If not, take the 8mm (L). They're both excellent.

Toast

[seanc]

Re: iMac G5 PSU - no power

Well, re-cap finished, seemingly, no change.
No lights on midplane, 5vsb seems to be 0, unless I'm blinking too much and missing it.
No explosion either... at least it would have done something!

Now I have an ESR meter, the caps I still have left, are all fine too.

So, can we continue to troubleshoot? I'm working on the assumption that my solder joins are good enough... for now - they're as tidy as I can get them at the moment.

[Toasty]

Re: iMac G5 PSU - Post recap tests - Part 1

Dang, sorry to hear that.

We've got to find the missing 5vsb FIRST. I was hoping the recap would get it back. It's usually a couple of caps that fail in that supply. Without the 5vsb, you can't go any further until that's restored.

Do you have an analog meter or a digital one? If you have the analog type, this goes faster.

Use the pics I've modded and attached as a guide. I may have one of the low voltages marked wrong on the back, so double check please. The voltages are all marked on the component side of the board.

Without the benefit of an isolation transformer, you need to be VERY careful working on this while connected to the AC Mains.

DO NOT TOUCH anything on the HOT side with anything except the meter probes.
Heatsinks and any exposed metal surfaces should be considered HOT.

1) Remove the PSU from its housing as you pictured it. Place it on an insulated, non-conductive surface.

2) Attach the AC power cord and plug it in to the AC mains. Do not connect the PSU to the computer.

4) Noting the pictures I presented, set your meter to the 500v DC (or Higher) scale and verify the power to the big mains cap.
- If the voltage is present as indicated in the picture, then proceed to Step 5.
- If it is not present, we need to find out why.

5) Check for any voltage present on the outputs measured from ground.

6) If the 5vsb is at zero (0), check the voltage across the 2 caps right next to its connection point. In your early pictures, one was a Taicon, the other a United Chemi-con. The 3 legged device between them and the transformer is what? Can you see its numbers? Check for voltage(s) around here. That transformer is the 5vsb transformer. Note that the circuitry on the other side of it is HOT. It is where the voltage and frequency is generated and controlled to create the 5vsb. If this is dead, then we have to find the problem on the HOT side.

7) Unplug the unit for at least 5 minutes. Please verify the voltage across the mains caps is at zero. If it is not, continue to allow the unit to discharge until it does. Confirm.

8) Set your meter to 50v (or closest) scale and verify no voltage is present on the 5 outputs to ground.

7) Once you're sure they're at zero, set the meter to the lowest resistance scale. Measure from ground to each of the 5 voltages: 3.3, 5, 12, 5vsb, 24v. If any of them show shorted, zero (0) ohms, such as when you touch the probes together, then we'll need to investigate why.

Wow! It's incredibly hard to do this by remote and not ever having this particular unit in my hands.



Toast

[Toasty]

Re: iMac G5 PSU - no power

Any luck or progress?

...Or am I talking to a fried pile of ash in the corner?

[seanc]

Re: iMac G5 PSU - no power

It's not fried ash.... yet, I've been distracted with family life recently - today was the good result 2/2

I did stick some probes on the main cap, it got to about 312v before I stopped watching it count slowly. So that part does work, we have power in, to a point.
ESR tested OK (in my opinion), about 0.50 ohms. I know it wasn't necessary... but I can now

I tidied out a cupboard last weekend and found a big industrial surge protector with RCD and on/off switches to each socket. I shall be making use of that.

Let me go find it, I may post some more results shortly.

[seanc]

Re: iMac G5 PSU - no power

Jumping to step 8 (I'll go to 6 in a minute) - what happens if I get a reading higher than 0, this is good, right?

[Toasty]

Re: iMac G5 PSU - no power

312v doesn't sound correct. It sounds low. I was expecting 390v or thereabouts.

What range(s) does the PSU label indicate as supply voltages?

Check the 2 resistors next to the fuse.
[ EDIT - should read 0.27Ω each ]

Check voltage at bridge rectifier. Upper right of full underside picture. 4 pins in a row, center 2 should be AC mains voltage, 240v. Outer 2 should be DC, 240v.

Check the 4 SMD resistors next to that bridge, R55A,B,C,D. They are a divider that supplies power to the startup circuit. Follow the traces and to put a probe on the one AC leg (closest in picture) of the bridge and the other probe on the last resistor R55D at the board edge, and get a voltage reading. Also, probe -across- those 4 resistors and you should get a low(er) voltage. That indicates some current is being drawn by the startup circuit.

Coming down along the same heatsink as the bridge, the next component (2 pins) looks like a rectifier. The one farthest down, directly across from the fuse could be a MOSFET. Discharge the unit and check the diode and the MOSFET.

Have Digital or Analog VOM?

Toast

[Toasty]

Re: iMac G5 PSU - no power

No, you want them at zero or you'll cook the ohmmeter portion of the meter.

Jumping steps is not good...

*B*A*N*G*

...as Sean's smoking remains hit the floor...



Roast Beast anyone ??

Toast

[seanc]

Re: iMac G5 PSU - no power

Well Results for step 5 are, 0. 6 are 0 for 5vsb, caps the same.

Can't get much useful out of that 3 legged device. It's packed in too densely to be readable, I can try taking it out. On a beep (continuity) test with el cheapo meter, I can get a beep on the outer two legs no matter which way round the probes are. Shorted, or because it's in circuit?

I can get a reading on the mains cap, so I'm not a complete probe wielding n00b...

In the previous post, I mean step 9, or (7 #2).

Just for reference I'm running with a cheap Screwfix DMM and an abused Fluke 8022a.

[seanc]

Re: iMac G5 PSU - no power

320v on the mains cap. My memory got confused.

[Toasty]

Re: iMac G5 PSU - no power

Yes, higher than 0 on step 9 (7 #2 - it was very late and I'd worked on that for an hour+), is a good thing.

No need to take device out. I just wanted to know for reference. I am fairly sure we are dealing with a loss of power on the HOT side now.

320v still seems low. See previous post #34.

[seanc]

Re: iMac G5 PSU - no power

Nearly midnight, I'll be back tomorrow.

[seanc]

Re: iMac G5 PSU - no power

In the middle on AC we've 231v as expected on the outer legs on DC we've 314v. (Big cap is rated at 450v).

Bridge leg (on the left side with R55a) to R55d gets me 310v

From R55a to R55d I get 105v

I'm pretty sure that the Diode D7 is shorted, either way gets me 0.003 whereas other diodes get nothing one way and a reading of 0.630, the other. This is how I expect them to work.

I'm not sure what to make of the two resistors, one outer one sits at 0.2 to 0.3 and the inner anywhere from 0.2 to 0.4 .

Mosfet seems good.

I've decided the Fluke meter is too suspect to be using. Switched to the meter in the below picture - it seems to be good so far. Found it stowed away in a box still new.

P.S. I've not been relying on the Fluke for any of my tests, I've been comparing the results between three (el 'cheapo from Screwfix, one pictured below and Fluke 8022a).