iMac G3 Lime Green (M5521) power supply

[jantoman]

Hi. A friend asked me to fix his old iMac that stopped working. I opened it up and found out that the SMPS IC was burned so badly that it was impossible to identify the type. After a lot of research I discovered it was a KA5S12656: I bought some of them from AliExpress, keeping my fingers crossed. Now I get the correct voltage on the main cap and on pin 3, but nothing happens when I push the front switch: I guess there should be a control voltage sent from the logic board to pin 5 (soft-start) but I can't see nothing there. Following the service guide's Troubleshooting suggestions, I checked the logic board looking for trickle 5v on J9 and DCU -1,2v on C10: nothing at all.
I checked several caps on both analog and logic boards, and the values seem correct. I didn't find a proper service manual with schematics, just the aforementioned "take-apart" service PDF.
Any suggestion and help would be highly appreciated. Thanks in advance.

[R_J]

Re: iMac G3 Lime Green (M5521) power supply

Do you have voltage on the drain (1) How much is the VCC voltage you are getting? Is pin 2 connected straight to ground or through a resistor?
Check if there is a diode and resistor coming from a small winding on the transformer that also connects to the VCC (this would be the run voltage supplied by the transformer)
The feedback pin should connect to an optocoupler (photocell) and the led side would connect to the secondary voltage (likely a 12-14v supply)
Its likely the sync input connects back to a small capacitor and then to a small isolation transformer which is supplied by the horz drive signal

Something else to check is on some of these models there is a second small bridge rectifier that supplies another circuit.

These are all based on a power supply that uses a similar ic

without a schematic, can you post a clear picture of both sides of the board for the power supply section. high resolution, clear enough to follow the traces

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

Thanks a lot for your support, RJ.

- on pin 1 (drain) there's the same voltage I find on the main capacitor: they're connected by a transformer winding, pin 2 and 3 (1 is ground). About 335 volts rising to 360... then I turned it off

- pin 2 is connected straight to ground

- on pin 3 (VCC) about 43 volts

- pin 4 (feedback) is connected to the optoisolator's pin 4 through R904 (15 ohms), there's 1,0-1,2 volts on it. Pin 3 is connected to ground and pin 1 and 2 are connected to the secondary section. I checked the optoisolator, it's ok.

Its likely the sync input connects back to a small capacitor and then to a small isolation transformer which is supplied by the horz drive signal

Yes, it looks like that. Less than 1 volt on it. Please note that the CRT is disconnected right now, otherwise it would be inpossible to work on both analog and digital boards.

Check if there is a diode and resistor coming from a small winding on the transformer that also connects to the VCC (this would be the run voltage supplied by the transformer)

The other winding has two pins, marked 49V and 19V on the PCB: they're not connected neither to the other winding nor ground, as far as I can see. I can draw the circuit connected to them and post it.
I can't see any other rectifier bridge: just two big diodes (D901 and D903) located under the big 33 kohm resistor.

Here are some pictures, let me know if you need more of them. Thanks a lot!

[stj]

Re: iMac G3 Lime Green (M5521) power supply

some shiny resoldered joints there - and some shit ones.
which did you do??

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

The IC and some capacitors. The contacts are good, anyway, no matter how they look (the crappy neon light doesn't help).

[R_J]

Re: iMac G3 Lime Green (M5521) power supply

What ground are you using to measure the VCC? the ic is rated max VCC 30 volts. You need to use hot ground
Check R912, it should be low Ω (5~10Ω?)

The two windings 49v and 19v will have continuity to ground. the 19v winding is used to supply the vcc voltage to keep the ic operating, if either winding is open, the transformer is likely defective, double check its resistance
The 49v circuit supplies voltage to Q901 which acts as a voltage regulator for the vcc. Check ZS904, its the reference on the base for Q901 and its emitter is the regulated output voltage that feedds the VCC via D906

If you do not have continuity between 19v and ground or 49v and ground, I suspect you found your problem.

C905 or C918 could be defective (open or no filter capability)

The sync input is not that important, its used to sync the power supply frequency to the horz frequency to prevent a beat pattern on the screen.
The small bridge rectifier may not be used on this model.

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

What ground are you using to measure the VCC? You need to use hot ground

Of course: look at the short wire soldered on the main cap's negative pin.

Check R912, it should be low Ω (5~10Ω?)

15 ohms.

If you do not have continuity between 19v and ground or 49v and ground, I suspect you found your problem.

No continuity with ground, neither one of them. I guess you nailed it. I'll check the other components you mentioned, but the transformer seems definitely faulty

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

Good news. I removed the transformer from the PCB and guess what? look at the 49v pin...
I fixed it with a short wire and soldered the transformer back in its place. Next thing to do is to check the other components of the startup circuit, then I'll try to turn on the iMac and see what happens, keeping fingers crossed.

[R_J]

Re: iMac G3 Lime Green (M5521) power supply

That would do it, I would replace C905 or C918 also, they are small capacity caps and usually dry out causing no or poor startup.

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

I replaced all three el-caps, C906 - C918 (47 mF) and C916 (1 mF). Then I mounted the analog board on the chassis, connected the CRT tube and plugged the iMac through a 70w lamp, an useful trick against short circuits. In fact, when I switched it on, the lamp flashed quickly then went off...
The fuse didn't blow, though. I'll check the rest tomorrow: here it's past midnight and I better go to sleep.
Just a question: do you think I could check the analog board without connecting the CRT?

[R_J]

Re: iMac G3 Lime Green (M5521) power supply

If you don't connect the crt, be aware that the high voltage lead will still develop high voltage, If you leave the anode lead connected to the crt, you MUST also connect the crt ground (aquadag) to the chassis ground. (if not the whole crt will be at 15~20kv)

If you have the crt completely removed from the chassis, keep the high voltage lead far away from everything, and your self. It can arc 2-4 inches

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

Right. Maybe it's a better idea to remove the flyback transformer to work safely on the boards, do you agree? Maybe it's defective as well, like another iMac 233 I have on a shelf waiting for the spare transformer... (different kind, of course)

[stj]

Re: iMac G3 Lime Green (M5521) power supply

no, what you do is isolate the B+ line between the psu section and the flyback,
it's the 100v(or more) feed - it usually goes through an inductor so it's simple to lift one end.
if you do that, put a 40w lamp between the B+ and ground to act as a load.

if your really lucky it may even route through a connector - philips tv's used to loop it through the connector to the scan-coils so if you pulled that plug it killed the HV for test purposes.

[R_J]

Re: iMac G3 Lime Green (M5521) power supply

Follow the +127 volt line between the power transformer and the flyback transformer, there is likely a jumper wire you can remove to shut down the high voltage to test the power supply section.
Just remember that without a load on the 127 volt line it will remain charged to 127 volts so remember to discharge it befor you reconnect it. you can use your light bulb to discharge it.

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

Yes, there's a jumper to the +B line heading to the flyback transformer. Tomorrow I'll be busy, but Friday I'll try to check the PS again.
Thanks guys for your help: it's about 40 years I'm messing up with electronics, but I always tried to stay away from CRT devices and I don't know much about them.

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

Hi: here I am again, I've been busy for a while and I had to put the iMac on a shelf. A couple of weeks ago a friend of mine came visiting my lab, saw it and told me "I've got one of those, but I don't remember if it's still working: I can give it to you, if you like". Of course I said "YES!"
He brought it here yesterday: as I inserted the plug I heard a relay click that was missing on my "patient"... first thing to check. Then I pushed the start button and it booted right at once... great! I turned it off and unplugged to discharge the caps.
In the morning I put the old analog board on the workbench to check the relay area and found out that protection diode D939 was shorted. I replaced it but still no luck. So I opened the working iMac and started to measure the main PS on both boards, comparing the results: I noticed a big difference on the main capacitor, and after further investigations I discovered that the ICs I bought (see first post) were just useless pieces of crap...
When I measure the "good" IC desoldered from the working board, keeping the positive probe of the digital multimeter (continuity mode) on pin 2 (ground) these are the readings I get:
1 (drain) 452 Ω
3 (VCC) 535 Ω
4 (feedback) 570 Ω
5 (soft start & sync) 575 Ω
Measuring the fake ICs I get no continuity at all, on any pin. What a great deal
I just bought another KA5S12656 from an Italian eBay store: quite expensive (21 euros including shipping fees) but at least there's a purchase protection if it turns out to be another fake.
I'll post an update as soon as I get the spare IC.

[jantoman]

Re: iMac G3 Lime Green (M5521) power supply

This morning I found the spare KA5S12656 in the mailbox: the measure readings matched with the good IC, so I soldered it, reassembled the iMac and turned it on keeping my fingers crossed... Well, it finally works. Hooray!
Thanks for your assistance and support, guys. I don't think I'd have any chance to get it working without your help. Respect!