iMac G3 Graphite 700MHz not booting, possibly cooked

[phlogios]

Hi!

I have an iMac G3 700MHz (Graphite, 2001) that will not boot or even chime. It just "crackles", and I see a green LED turn on on the mainboard near the RAM slots, that goes out if I hold down the power button or pull out the power cable. I bought a separate, similar iMac G3 blue 400MHz (also slot loader) so I can swap parts and see what part is broken. The blue iMac boots fine, so it is a good testing machine.
I have eliminated the typical suspects:
Both PRAM batteries work in the blue iMac.
Both Down Converters work in the blue iMac.
The blue iMac boots and chimes with or without every permutation of ramsticks from the graphite.
At first I suspected the flyback transformer (which I have no idea how to fix), because the graphite's display quality deteriorated in its final years before dying 2 years ago. But... that can't be it - because the graphite's Mainboard does NOT boot in the blue chassis (which has a known working flyback transformer). So - the issue is the mainboard. Does anyone know where to begin figuring out what can be wrong with the mainboard? An error that leads to:
No chime
No boot
Green LED ON
No HDD or CD Drive action

I suspect that the logic board has been fried by overheating, since the computer showed signs of having overheating problems before it died. The computer would freeze randomly and then I'd have to reboot it. Is there any way to tell for sure that the logic board is fried, and if it's fried, is there any way to tell if it's fixable?

 

[fred1212]

Did you look at the big 560 microfarad cap for signs of bulging?

 

[phlogios]

Thank you for your reply! I am away for a few days so I won’t be able to check now, but is the 560 microfarad cap on the logic board or on the downconverter? Because I’ve already verified that both downconverters work if the indigo’s logic board is connected to them. Only the graphite’s logic board seems to be broken. I will definitely check either way!

 

[lilliputian]

As I say to everyone who has one of these with these symptoms: leave it plugged in for about a week, testing periodically if you wish. I was able to revive 3 400MHz G3s with this method, no recapping required. In fact I had initially recapped one of them only to find it did nothing to fix the problem.

 

[phlogios]

Lilliputian: have you had the same symptoms? No chime, but green led on logic board?

 

[lilliputian]

When the power button was pressed, there was a brrzt as the crt degaussed, but nothing else seemed to happen. There was no other obvious activity as I recall.

 

[phlogios]

I took out my multimeter for the first time in my life and started testing according to the instructions in the service manual.

J9 shows +5V as expected
C10 shows 0 volts with power cable plugged in (expected is -1.2V)
C10 shows 0 volts when power button is pressed (expected is 21V) (EDIT: Wrong capacitor to check. C4 is correct one and measures -1.2V and +25V)
J7 shows correct voltages on pins 2, 4 and 14 as per the manual.
3.6V battery shows 0.06 volts, so it needs to be replaced again (!).

I doubt that the battery is the problem, since my indigo starts fine with this battery.
After testing C10 (my C10 is not in the same position as the picture in the service manual) I slipped and hit another one and there was a tiny spark
I did not check the fuse on the PAV board yet as I am scared of breaking the case even more.
Any ideas?

 

[fred1212]

I'm working on the same imac atm just waiting for parts. The big 560 cap was shorted. I'll let you know what happens

 

[phlogios]

Oh my god I found a bulging cap!! Is this the one you meant? Is that a 560 cap? The same cap on the indigo imac is completely flat so this could be the one!?

 

[phlogios]

The bulge is barely visible.

 

[phlogios]

I set my multimeter to "beep" and placed the black terminal in a screw hole, and the red terminal on each side of all the capacitors one at a time. All capacitors either made the multimeter go beep or at least show numbers on the display, *except* the slightly bulging 470uF capacitor which gave no response at all. I tried the same thing on the indigo's capacitor and it showed numbers on the display. At this point I'm quite sure that the 470uF capacitor on the Graphite is toast. I've ordered new caps on ebay, will update this thread when I've tried my first recap.

 

[fred1212]

I'm referring to the big cap on the power supply board you can see it through the clear case near the crt. you need to fully dissemble the iMac to get to it.

 

[fred1212]

This one

 

[phlogios]

But does it make sense that that one would be broken if it works with another logic board?

 

[phlogios]

I just now verified again that the indigo logic board will boot fine within the graphite case, with either DCO, with images showing up on the CRT.

 

[fred1212]

Good point, probably worth looking at changing anyway if you get the graphite LB to work. Frustrating things computers!

 

[phlogios]

Quick update:
I was reading the wrong manual. I found the correct one, for the Summer 2001 editions of the iMac G3. In the correct manual, it instructs you to check voltage at C4 instead of C10. The voltages are -1.2V when plugged in but powered off, and +25V when powered on (expected +21V). It could be within spec, I am not sure. Still waiting for my new capacitors to be delivered.

 

[phlogios]

It wasn't the caps. Most likely it is one of the big BGAs on the logic boards that is failing.

 

[phlogios]

I was not able to fix it with a heat gun and flux.

 

[aliasmk]

3.6V battery shows 0.06 volts, so it needs to be replaced again (!).

Your PRAM battery might be dead already if the PMU crashed at some point during handling, you didn't properly reset the PMU chip after changing the battery, or the PMU itself is dead somehow (possible culprit?).

From the iMac DV service manual starting on page 199 of the linked PDF:

Important: Be very careful when handling the main logic board. Remove the battery when handling the logic board so the PMU is not affected. The PMU is very sensitive and touching the circuitry on the logic board can cause the PMU to crash. If the PMU crashes, the battery life goes from about five years to about two days if the PMU is not reset.