Last revision 700MHz iMac G3 not turning on

[philgxxd]

Hi again.
During the process of rebuilding my old iMac G3 (400MHz from year 1999) i purchased another graphite one just in case my repair attempts would fail (i changed the flyback transformer and a resistor).
This graphite one is actually working but it came with two other non-working ones in a bundle.
After i could succesfully repair my old iMac and could even resolve the non-working Firewire ports thanks to member @MacTech68 i'm now trying to restore another spare one.
It's the said 700MHz model from the title.
I tried yesterday to turn it on and it wouldn't give any sign of life.
I took it apart and began replacing the big C905 from the PAV board for a slightly bigger on (400v 560uF OLD to 400v 680uF NEW) and the potentiometer VR901.
I then saw a busted inductor. See attached images.
Can someone guide me to how to read the value of this one to get an replacement?

On that last picture you can spot where it sparked from the other side of the board.

Now i think that an inductor rather fails by itself so i'm aware that probably there are more to it.
Thanksful for any help with troubleshooting...

 

[MacTech68]

I was never too successful on the iMac Slot loading analog board. Time pressures way back meant that a replacement board with a full video adjustment was faster and an easier quote. Gone were the days when a module had a 10 year life span, and troubleshooting faults with these became a nightmare in the ever-changing pace of a personal computer's lifespan.

But enough of that. If I'm looking at it correctly, this looks like it's in series with the deflection coils. One horror I saw but never read much of at the time was rubber wedges under the deflection coils sweating a fluid and overheated shorted deflection coil windings, putting a heavy load on the deflection circuit - usually resulting in an immediate shut-down.

Have a look at the rubber wedges. If they have little globs of liquid on them, I'd be very concerned.

I'm not sure what caused it but I saw it on two machines from different sources. Part of the yoke plastic was melted.

Otherwise, this could be an under-powered component that heats up and causes the dry solder joint.

IIRC at the time I decided that so many faults were different causes on these analog boards that finding each one was near impossible with the fast turn-over of machines through the workshop. Poor quality soldering seemed to be a major cause of most problems, that, and the heat generated inside the (fanless) case in our climate here in Australia.

As for the component, if you remove the plastic wrap and don't see bad discoloration of the copper windings, you may like to risk cleaning up the solder joins by removing the component completely, and using desoldering braid and NEW fresh solder to clean the pads. If you've not done this before for a dry joint that has ended up like this, it's critical that you clean both the pads and the end lead(s) of the component.

It's just possible that the dry joint came first, but grew worse with arcing heating/shrinking the plastic wrap. Perhaps I should have said that first !

The component seems to be A HLIN-001 (Horizontal LINearity coil) manufactured by "FEC" with a date stamp of 2001 week 22 (0122). If I find out the who and what I'll update this thread.

 

[philgxxd]

Thanks a lot @MacTech68 for your attention!

I will take a look at the inductor tomorrow as i don't have the mac at home. I measured the inductor/coil and it measured less than 1 ohm this morning. Does that mean it's shorted?

 

[MacTech68]

Without a microhenry meter (measuring inductance), that is all you can test. Your result with an ohm meter is to be expected for a working part, but is NOT a true indicator.

 

[philgxxd]

Today i cleaned the inductor AND board and resoldered it in again but there still is no sign of life.

I tried to measure the inductance with an function generator (just a 1MHz sine wave generator) and an osciloscope in xy-mode but the resonance frequency seemed to be out of range (above 1MHz) so i got no real clue if the inductor is still good.

Don't know what to try next as it's difficult to measure much on the PAV board when the iMac is half assembled.

1. There you can see the windings from the inductor.
Just where the kink is i think the insulation is a little bit gone but there doesn't seem to be a short.


2. The other side. The insulation seems fine.


3. The electrolitic cap i replaced (the big black one) and
the blue cermet variable resistor.

 

[MacTech68]

The Linearity Coil looks fairly good - of course that means very little but it doesn't seem to be burnt.

When you replace the B+ adjust pot (VR901) do you measure the original out of circuit and adjust the replacement? Take into account that the centre and middle are soldered together on one side. I'd usually measure the old one (making it warm first) and then adjust the replacement. This fault (bad B+ ajdust pot) also occurs on the "14" Color Display".

Don't know what to try next as it's difficult to measure much on the PAV board when the iMac is half assembled.

Yes, that's one issue that made these a nightmare to diagnose. I never had the time to build a 'troubleshooting rig' to make this easier.

Try resoldering ALL components attached to heatsinks on the Analog Board. Look for and resolder any components that exhibit a dark ring in the solder joint and/or any joints near areas where the board itself looks darker/heat affected.

You'll need to go through the steps in the service documents to be sure the 5 volt standby rail is running and that the 'Down Converter On' is changing state and the 'Down Converter' outputs appropriate voltages.

When re-assembling, be sure that all connections are attached, especially the one in the top-middle of your last photo. IIRC it carries the supply voltage to the 'Down Converter'.

 

[philgxxd]

When you replace the B+ adjust pot (VR901) do you measure the original out of circuit and adjust the replacement? Take into account that the centre and middle are soldered together on one side. I'd usually measure the old one (making it warm first) and then adjust the replacement. This fault (bad B+ ajdust pot) also occurs on the "14" Color Display".

I measure in-circuit AND out-of-circuit (bridging with one probe the pins which are connected together in-circuit) and take note of both values. Then i take an average value of the two and adjust that when i have soldered in the replacement pot.
How do you warm up the component?

Yes, that's one issue that made these a nightmare to diagnose. I never had the time to build a 'troubleshooting rig' to make this easier.

I have to look into that...Could i just leave the actual screen tube disconnected to check voltages on the PAV board or would high-voltage sparks come out of the flyback transformer if i get to close...

Try resoldering ALL components attached to heatsinks on the Analog Board. Look for and resolder any components that exhibit a dark ring in the solder joint and/or any joints near areas where the board itself looks darker/heat affected.

Will do that!

You'll need to go through the steps in the service documents to be sure the 5 volt standby rail is running and that the 'Down Converter On' is changing state and the 'Down Converter' outputs appropriate voltages.

When re-assembling, be sure that all connections are attached, especially the one in the top-middle of your last photo. IIRC it carries the supply voltage to the 'Down Converter'.

Will do that also!

Thanks again for your valuable tips!

 

[MacTech68]

How do you warm up the component?

Usually a few seconds with a soldering iron.

Could i just leave the actual screen tube disconnected

Not recommended at all - not only is there a possibility of damaging the Analog Board, but leaving 20,000 Volts free to arc to the nearest ground path may be detrimental to your health (permanently). Please DON'T

 

[philgxxd]

I finally managed to do the measurements on the LogicBoard.
I find the indications from the service manuals (i tried to compare different revisions) a little bit vague.
Unpowered only having the AC applied i get the 5V trickle Power at Pin1 from J9.
Then its getting a little bit confusing to measure the DCO voltage at C4 (in another paragraph it says C10?).
It should measure -1.2V to 1V in unpowered state and jump up to 21V after powering on.
I could measure that but it only raises up to 19V and seems to stay there. It's not coming down again (i'm not sure if it should be like that).
As soon as i power it on it lights up a little led called 3.3V on the motherboard nearby the IDE port but i don't get any bong sound nor does the led from the power button do light up. Only when i force a shut down the led on the power button does light up briefly.
Furthermore i can measure the 12V on Pin2, 5V on Pin4 and ONLY 1.97V on Pin14(should measure 3.3V?) on connector J7.
Does these voltages from J7 come from the PAV board?
Should i just try another down converter board?

 

[MacTech68]

The DownConverter seems to get AC input via the J7 connector itself, via Pin 19 and 17.

The DCO voltage test point should raise to 21 volts and stay there when the machine is powered on.

IIRC, the DownConverter receives only an AC input from the PAV board - if you are getting 12volt and 5volt on the J7 connector, this would seem to indicate the DownConverter is basically working.

However, the 3.3volt on pin4 sounds like a place to start. It's possible that the Downconverter has a very strange fault, or the iMac logic board is pulling it down.

The way Apple routed power on the CRT iMacs drives me crazy. I must say I've never seen a faulty downconverter - but it's not impossible for a near 20 year old machine.

If you can easily swap the downconverter from a working machine, sure - give it a try.

I would also suggest disconnecting the hard drive and optical drive data & power cables AND removing all RAM modules for testing.

Also, check the PRAM battery is measuring at least 3 volts (in circuit) and press the PMU reset button - ONLY ONCE - see the service manual for the PMU reset procedure.

 

[philgxxd]

Thanks again @MacTech68.

The way Apple routed power on the CRT iMacs drives me crazy. I must say I've never seen a faulty downconverter - but it's not impossible for a near 20 year old machine.

Yes. I think i saw that the AC line is routed through a long way on the Logic Board with thick traces. Kind of crazy.

I'll have to add that during troubleshooting i still didn't put back the HDD/DVD cage nor did i put the battery nor RAM sticks back.
I did however the PMU reset.

I will put back at least the PRAM battery and go on checking this evening.