The powering down and powering up problem is typical of the the bad capacitor problem. I always say, you’ve got to fix what you find bad. In this case, having bulging caps on the MOB can in fact cause the problem you describe. I would definitely fix the MOB with new caps to begin with. Then, if you still have a problem, the original bad capacitors on the MOB could have caused something to go out in the PSU other than the PSU caps themselves.

Refer to my other Apple articles for more information.

Regards,

Jim

Excellent information!

All electrolytic capacitors have a finite life:

The voltage across the plates of the capacitor, in normal use, causes electrolysis in the electrolyte solution. i.e. breaking down the solution into its constituent atoms/molecules. This process ‘consumes’ the solution because the solution is converted into various gases. The rate of consumption is highly dependent on the voltage applied (higher voltage=more electrolysis!).

As the capacitor ages, this process begins to accelerate and eventually sufficient gas is produced to cause a pressure build-up within the capacitor, eventually causing the capacitor to ‘vent’ (pop or explode!). The score-markings on the top of the capacitor are there to ensure the capacitor ‘vents’ in a controlled manner.

A partial venting will also cause electrolyte solution to leak out, which further accelerates the capacitor’s demise.

Testing capacitors:

While measuring the capacitance of a capacitor seems like the right thing to measure, actual measurements on bad capacitors will either show a good reading, matching the value stated on the side, or, most likely a higher reading. Strangely the capacitance actually INCREASES as the capacitor is dying!

In DC filter applications higher capacitance is normally a good thing because it provides more filtering of the high-frequency ripple from the SMPSU. However in the case of comparing an electrolytic capacitor’s stated value to the measured value, this indicates a dying capacitor – This provides a poor-man’s capacitor test: Look for higher than normal readings (in addition to basic visual tests)!

The real parameter to check is the Equivalent Series Resistance (ESR). This resistance is in series with the capacitor and so cannot be measured with an ohm-meter – A special ESR meter is required (I use a Peak ESR60 £90).

The ESR of a capacitor is important because it IS the parameter that increases as the electrolyte is consumed. The ESR is also represents the power loss in the capacitor (Ploss = Resr x I^2), this power loss causes heating of the capacitor, which further accelerates the electrolyte consumption.

It is very normal to see capacitors in a SMPSU that look normal (no signs of bulging, etc) but actually have a high ESR reading and are in fact faulty – Always test the ESR of all capacitors (individually out-of-circuit).

Again a poor-man’s test for ESR is to simply feel the temperature of the capacitor. If it is hot, then there are significant losses caused by the ESR.

Finally, increasing the capacitance values of replacement parts may seem like a like a good idea to reduce ripple currents, but be warned that going too high may mess-up the dynamics of the SMPSU causing it to become unstable!

Hope that helps, sorry if it’s already been said (haven’t waded through the comments!!!)

Si

I have a question for you, I can see on the underside a small SMD diode has blown. Is there anything that can be done about this or is it time for a new PSU?

Thanks very much in advance for your response

Nj

It is always a good idea to replace all the caps that are included in the PSU capacitor kits. Just because you do not see any others bulging, does not mean the other caps are good.

It is possible you have something more than blown caps as being the problem on the PSU. The fact the voltages are bad is a sign of a faulty PSU. If you are not careful when soldering on the bottom of the PSU circuit board, it is possible that now, one of the other components is damaged. I would inspect your soldering work carefully to make sure there are no solder bridges or cold solder joints.

Psu…1 obvious dead cap. Replaced about 5 of them so far, very hard to get out….all caps in the MOB look perfect. Result

Light1 always on….never 2 anymore (btw lcd pulses sometimes.)

Testing voltages on PSU. Sometimes when jumped get results but very off from the list. ad it seems the grounds yield very different results too.

So there it light scorching inside thePSU im sure this is the problem. Surely if it chimed ever, (fans spun, HDD spun then nothing) then the MOB is good? Please set me straight as everything in the PSU now looks good.

The Apple trouble shooting guide trips up a lot of people in this area of the power supply and logic board diagnosis.

The first LED is only an indicator of trickle voltage getting through from the PSU to the MOB. The Apple trouble shooting guide is off quite a bit on their flow chart in regards to this issue with LED number 1 as it relates to the power supply unit.

If LED number 1 is lit, AND the PSU capacitors are bulging, which means the power supply is bad, then the power supply is a good candidate for capacitor replacements to fix the problem. However, if LED #1 does not light when plugged in, then something else in addition to the capacitors is probably blown on the PSU. Some folks have trouble shot down to either a blown inline fuse (soldered in to the circuit board of the PSU), or there are some surface mount transistors on the bottom of the board which have been known to go out. I do not have the spec on the transistors, but I know some folks have taken two PSUs and made one good one by replacing the surface mount transistor from one to the other.

Hope that clarifies the number one LED being on or off on the MOB when the PSU is plugged in.

Regards,

Jim

First of all, compliments on the amazing work and info on iMac G5 caps and PS issues, diagnostic and fixes.

I have read a fair amount of the relative info on your site and am working on a dead iMac.

It had the motherboard replaced a few years back, then started getting flaky a few months ago.

Shuts off by itself. Intermittent start up, but would always eventually, until….
Two days ago power in my house went off for a short while now it wont start up. Outlet has power :^).

The main board looks good (caps). I yanked the PS and opened it. Numerous bulging caps with a couple with small indication of brown guts leaking.

Here’s the question:
Apple site says that if first indicator light (LED) on board light up, the PS is OK, if the second does not the board is bad. Is it in your experience possible that the PS indicator light can be illuminated but the power supply is bad causing the system to not start up?

I’m considering delving into replacing the PS caps if it seems like the path to resolution.

Thanks very much in advance for your response,

Ted (SJ California)

All the grounds are common so it doesn’t matter which one is probed with the meter right? The only voltage read is from ground to pin 11 at 21 volts. I thought that I was doing something wrong because measuring between a ground and any other pin has no voltage.

Do you have any ideas on where to start looking to fix with these readings?

Thanks again,

JJ

Basically, once the power supply is turned on, by jumpering the appropriate pins [15 (gray) and 16 (blk/gnd)] as outlined above, then you would proceed to measure the voltages from ground pins to the voltage pins using the multimeter set on DC. You could also check to see if there is any AC ripple on each of the voltage pins referenced to ground. To do this measurement simply set your meter from DC to AC readings. If there is excessive ripple, it will show up as an AC voltage reading on the DC pinout.

Hope that helps explain it a bit more.

I got the thing out and jumpered the pins. I started testing with my multimeter between pins 11 (gnd) and 22 (24v). I got the 21.3 volt reading. How do I test the other pins? With one end at pin 22 or at ground?

When I test between pin 22 and any other pin I get the same 21.3 v reading. When I test between 11 and any other pin I get zero. Could you explain a little more on how to test for voltage on each of the pins?

Thanks and keep up the good work.

J

Does anyone have a schematic or a pin out diagram of a power supply PA-3241-02A1 from a IMAC A1225; I want to test and repair that power supply and for that I need a schematic or at least a pin out to identify the power on pin.Thanks

Now, it is possible that just replacing the visually bad caps on the MOB has now progressed to the next weak link. I always recommend replacing the entire two types and banks of capacitors on the MOB. You simply can not be sure the other ones are good and will forever be second guessing yourself as to are they or aren’t they good caps.

By the way, if you are hearing the hard drive, then something is turning on. Maybe you need to reset the PRAM and the SMU. It could be the PSU is faulty under loaded conditions too.

Cheers for the swift reply and advice.
Very decent of you to help people in this fashion.

The power supply which doesn’t work at all will go into my junk electronics box from where i salvage bits if required.

The other PSU I will test the voltages when not under load and see if it still cuts out after 60-90 seconds and then post back with the results.

Cheers once again…..

Cid.

Thank you for the comments about the website.

I was thinking about your missing voltages problem child PSU, and think that there is probably something else blown out completely, closer to the front end of the PSU. There are all types of other components that could get taken out when capacitors become so bad that it draws excessive current and causes a loading effect internally. This is why it is always a good idea to replace bad caps as soon as they become visually apparent instead of waiting for the entire PSU to die completely.

The other PSU probably has one of the surface mount transistors or control chips blown out. I’d be curious if the PSU will power up by itself, and stay on under no load. It would be interesting to note what the voltages are in this condition.

You also might be able to check the surface mount transistors on the bottom of the PCB and see if any of the emitter base or base collector junctions are shorted or open. If you find any that do not compare the same as the others, then you might be able to replace it from one power supply to the other. Just look for the same p/n on the surface mount transistor, though you might need a magnifying glass to read it.

Regards,

Jim

Cheers for all your hard work on the website.

I have replace the caps in a couple of G5′s and managed to get working units for my children.

I have picked up another G5 and I’m pretty sure that the logic board is working fine as it was a good unit that a customer salvaged the working PSU from.

I have another couple of PSU’s on hand which I’m trying to get running with limited luck.

One of the PSUs is a DD type PSU. It was only putting out 24VDC on pin 24. No other VDC on any of the other pins. It did have a couple of popped caps, which I replaced, however the unit still won’t turn on when I jumper pins 15 and 16. I think maybe this PSU is a dead…

I have another PSU which is behaving a little weird. It will power on fine but after approx 60-90 seconds it will power off completely. So that none of the diagnostic lights on the logic board are lit. If I take of the mains and leave it for 60 seconds, then plug it back in, the unit will fire up again for approx 60-90 seconds.

This PSU unit doesn’t look like any on your website.
The sticker reads Delta Electronics INC
Model DPS-180QB A
Apple PN 614-0329

I’m half thinking this might be a replacement unit as opposed to an original Apple part.

None of the caps look bad, so I’m not sure if there is anything i can do to fix it. It’s a little frustrating as the unit does power on. It seems as if the load on the unit causes it to trip out.

Any ideas or comments would be welcome. Otherwise I will wait for another salvage G5 to come my way.

Cheers
Cid.