My Prototype iMac G4 "Q26" (DVT #6) [No Boot - Help!]

[MacTech68]

I saw that the usb host controller had a bulge in it today. I checked my mac mini, which uses the same usb controller chip from NEC, and there's no bulge.
Also, D26 had a hole in it, it's something to do with the audio ports.

Someone was kind enough to send me the schematic for the prototype, so I'm going to use the serial download port to check for life.

BTW, D23 was a 15V SOT23 3 prong thingy (D for diode?). It would be great if someone could send me a few

Also, i think it could use a new voltage regulator. It's labeled sc2602s, I couldn't find this chip on any other mobo, so I could use a few.

PDF for the SC2602 is here. Can't locate any just yet....

I'm intrigued as to what the failure sequence is. Sounds catastrophic.

 

[kbfr08]

PDF for the SC2602 is here. Can't locate any just yet....

I'm intrigued as to what the failure sequence is. Sounds catastrophic.

Just unsoldered the agere chip (southbridge?), there were burn marks under it. No big deal, since I've got a few of these chips from salvage boards. Also unsoldered q23, since it was sparking a bit, this could have been the source of the failure of the imac.

If you'd like the schematic PDF, I'll send it to you. The components list has been quite useful so far.

SC2602S is used in pretty much any product that uses RAM. You could probably find it on a graphics card.

 

[MacTech68]

Just unsoldered the agere chip (southbridge?), there were burn marks under it. No big deal, since I've got a few of these chips from salvage boards. Also unsoldered q23, since it was sparking a bit, this could have been the source of the failure of the imac.

If you'd like the schematic PDF, I'll send it to you. The components list has been quite useful so far.

Agere could be the ethernet IC (I would guess)...... but what do I know, I don't even have a board in front of me. Although, I think around that time, Apple was using Broadcom Ethernet controllers. I replaced a few on late slot load iMacs blown up by lightning strikes via ADSL.

And yes.... very interested.

EDIT: one source I can find for the SC2602 is a Thomson HD LCD TV sold in europe, model LCD42M61NF21 and LCD37M61NF21 which uses an SC2602 at U27. Thomson part no. 13-SC2602-S0B. The TV is circa 2006. Alternatively you might get Semtech to send you a "sample"

EDIT 2: I wonder if a G4 iBook might use one......

 

[kbfr08]

Agere could be the ethernet IC (I would guess)...... but what do I know, I don't even have a board in front of me. Although, I think around that time, Apple was using Broadcom Ethernet controllers. I replaced a few on late slot load iMacs blown up by lightning strikes via ADSL.

And yes.... very interested.

EDIT: one source I can find for the SC2602 is a Thomson HD LCD TV sold in europe, model LCD42M61NF21 and LCD37M61NF21 which uses an SC2602 at U27. Thomson part no. 13-SC2602-S0B. The TV is circa 2006. Alternatively you might get Semtech to send you a "sample"

EDIT 2: I wonder if a G4 iBook might use one......

I'll send an email to Semtech, I hope it's not discontinued.
I've only got a macbook, and a mac mini neither use the SC2602S. When I get my motherboards and graphics cards back from my friend, I'll check them.

 

[MacTech68]

I'll send an email to Semtech, I hope it's not discontinued.
I've only got a macbook, and a mac mini neither use the SC2602S. When I get my motherboards and graphics cards back from my friend, I'll check them.

There are two suppliers listed in the USA but they have what seems to be a full reel each so I'd imagine it would be an MOQ of a reel. That's 2500 ICs.

 

[kbfr08]

There are two suppliers listed in the USA but they have what seems to be a full reel each so I'd imagine it would be an MOQ of a reel. That's 2500 ICs.

Well, I requested a sample. Hopefully, they'll ship it.

 

[MacTech68]

The USB controller you mention is the IO controller (UDP720101) or the power switch (U1 - TPS2023)? I had to replace a USB power switch on a 233Mhz tray loader years ago (was only 14 months old). I had to buy a tube of 28! Sadly, I don't have access to them anymore.

The input to that IC has a 2Amp 6V fuse to the 5v rail. Worth checking that too.

Still wondering what the audio and USB will have in common.

 

[kbfr08]

The USB controller you mention is the IO controller (UDP720101) or the power switch (U1 - TPS2023)? I had to replace a USB power switch on a 233Mhz tray loader years ago (was only 14 months old). I had to buy a tube of 28! Sadly, I don't have access to them anymore.

The input to that IC has a 2Amp 6V fuse to the 5v rail. Worth checking that too.

Still wondering what the audio and USB will have in common.

It was the D720101F chip made by NEC. I have a spare from a salvage mini.

 

[zcream]

Excellent thread. I had similar problems with my imac 17-inch.
Lets see I upgrade to Leopard 10.5.8 and then I got stuck at the spinning circle. I opened it up and turned it on while it was still open - so the heatsink was not connected. Now there is no display but it chimes.

Simple q to start with.
How do you take out the logic board and debug it ? I think it needs to be connected to the rest of the system for it to even turn on. There seems to be an inverter near the AC connector. That connects to something in the body of the imac and there is a power connector that connects back to the logic board.
Its a bit strange to look at the logic board with cables dangling off it.

 

[toolbox]

Does it stay on long enough for you to get in and reset the firmware?

 

[MacTech68]

Excellent thread. I had similar problems with my imac 17-inch.
Lets see I upgrade to Leopard 10.5.8 and then I got stuck at the spinning circle. I opened it up and turned it on while it was still open - so the heatsink was not connected. Now there is no display but it chimes.

Simple q to start with.
How do you take out the logic board and debug it ? I think it needs to be connected to the rest of the system for it to even turn on. There seems to be an inverter near the AC connector. That connects to something in the body of the imac and there is a power connector that connects back to the logic board.
Its a bit strange to look at the logic board with cables dangling off it.

I always wanted to build a jig to work on these, they were cumbersome to work on opened (try a 20" display ). Apple did sell a cradle and extension cables for the earlier models.

It depends on whether you have a USB 2 G4 iMac, OR an earlier iMac G4. On the USB2 models, Apple moved some of the connectors on the logic board to one edge making it easier to diagnose with most things connected. The earlier ones are a pain, especially the video and the fan/speaker connector.

At the power connector is a mains filter (the silver box), this then goes up inside the dome to a Switch Mode Power Supply that is physically in two halves. This then supplies 12v to the logic board and 12v & 5v to the optical and hard drives. There is a thick-er black cable that carries video to the display, which often pops off if you're not watching carefully when taking it apart. If I recall correctly, it has a plastic cover that you have to carefully remove and the the connector slides apart sideways. Pulling on that cable could potentially wrench the connector off the logic board. This was revised in the USB2 models (thank goodness).

Your best bet is to establish which model you have, and then *ahem* getting the appropriate take-apart for it.

There are 4 possibilities.

imac.17-inch.pdf
imac_17in_1Ghz.pdf
imac_usb2.pdf - later updated to include the 20"
imac.flatpanel.pdf - for the original 15" 700 & 800Mhz model.

Plugging those into google should get you some nice resources.

Find the Apple designation of your model by plugging your serial number into here.


@toolbox. The problem here is that with the USB port not being activated at power-up, there is no way (that I know of) of getting into open firmware or even resetting PRAM. The only thing that can be done is to reset the Power Manager using the button in the RAM/Airport bay underneath and removing the battery on the logicboard. This is similar to the one I fought with a few years ago, where power to the USB port wasn't active until AFTER the OS had loaded (for some bizarre reason).

 

[kbfr08]

Does it stay on long enough for you to get in and reset the firmware?

No, I can reset the PMU from the button on the board. That's it. Any hopes of accessing open firmware will need to be done from the JTAG port. The port is J24, on the underside of the board, near the resistor labeled "101".

This port should give me access to diagnostics, and POST errors. The pins are as follows:
Pin 6: RXD
Pin 8: TXD

Pin 1 has a circle next to it, 6 and 8 are on the opposite side of pin 1.

Solder your jtag cable to those pins. RXD on the board goes to TXD on your cable. I'm using a PL2303 chip for this, any JTAG cable will work. Also, knockoff DKU-5 cell phone data cables use the PL2303. Use PUTTY to view the output.

I'm not able to jtag into the board until I get my tube of SC2602S chips.

Edit: L84 is burnt, it's a 250 ohm inductor on the +5v line from the psu. It's right by a screw, that would probably explain the clicking from the psu that I got when plugging the imac in for the 1st time.

 

[MacTech68]

...I'm not able to jtag into the board until I get my tube of SC2602S chips.

Edit: L84 is burnt, it's a 250 ohm inductor on the +5v line from the psu.

Hmmm If L84 is burnt, I'd be looking over the CPU OVDD. VR4 (EZ1852) and D19 (MBR8130T3 Zener possible shorted), Q4801 (FDC502P) and associated components (you get the idea). The 5V switch is also an oddity.

May I ask who is supplying the tube?

 

[kbfr08]

Hmmm If L84 is burnt, I'd be looking over the CPU OVDD. VR4 (EZ1852) and D19 (MBR8130T3 Zener possible shorted), Q4801 (FDC502P) and associated components (you get the idea). The 5V switch is also an oddity.

May I ask who is supplying the tube?

It's from Semtech.

Anyway, I replaced one of the 0 ohm resistors with a piece of wire, and the motherboard gets power now, it's also no longer hot to the touch. I'm gonna install the southbridge and the USB chip and see if it boots.

 

[zcream]

Interesting, I had no idea that resistors could go bad as well. I thought it was only capacitors.
This gave me an idea. I had a Macbook that its previous owner spilt coffee on. It has some rusting that I tried to clean with IPO.
Sometimes it would boot but not work well - and it was intermittent. Mostly there was no power to the board.
I thought it would be the capacitor - but it seems to have only smd capacitors.
Now I will test the resistors as well.
Is there a special probe I can use for testing small resistors ? /the standard one with my dvm still seems to be too big.

 

[MacTech68]

Interesting, I had no idea that resistors could go bad as well. I thought it was only capacitors.
This gave me an idea. I had a Macbook that its previous owner spilt coffee on. It has some rusting that I tried to clean with IPO.
Sometimes it would boot but not work well - and it was intermittent. Mostly there was no power to the board.
I thought it would be the capacitor - but it seems to have only smd capacitors.
Now I will test the resistors as well.
Is there a special probe I can use for testing small resistors ? /the standard one with my dvm still seems to be too big.

Well, if you have a liquid spill, boards become un-reliable. When doing quotes for insurance, if I can see evidence of a liquid spill, I will advise that that module be written off. Why? Because once corrosion starts, unless you can remove all the corroded material, it will "grow".

The big problem is that you can't see where the corrosion is. For example, under an IC where there is such a narrow gap that capillary action assists the flow of a liquid. Add to that that most boards these days are "Multi-Layered", where you have tracks sandwiched inside the board itself and the ability to remove corrosion becomes near impossible.

Sometimes, you can be lucky, and spot the surface corrosion, clean it and it will last for years, but if a customer was paying for insurance, the benefit of the doubt comes into play.

What you need to look for is a "rusty" brown, or fuzzy green. That'll be the killer. If it was powered when the liquid hit or shortly after, then it could be ANY component.

And yes, you can buy DVM probes with fine points. Just make sure you get a style of banana connector that fits your DVM.

 

[MacTech68]

It's from Semtech.

Anyway, I replaced one of the 0 ohm resistors with a piece of wire, and the motherboard gets power now, it's also no longer hot to the touch. I'm gonna install the southbridge and the USB chip and see if it boots.

Also, here's some high res images:
http://www.mediafire.com/?mlnimienznl

I scribbled some notes on the board, and the production date (which isn't accurate).

I envy your ability to to BGA re-work!

 

[kbfr08]

I envy your ability to to BGA re-work!

I'm an xbox 360 owner

 

[MacTech68]

I'm an xbox 360 owner

ah ha! That explains that! I chuckled upon reading that!

LATE EDIT -- anybody seeing the very late edits on my posts (several years late), my account was compromised and s-p-a-m links were posted to the end of many of my posts with a matter of 30 minutes.

 

[kbfr08]

Damn,
The agere chip separated from the pcb during reball. I can pull off the die cover if I wanted to, and it no longer lays flat. I'll see if I can get more of these chips, it's unlikely since agere no longer exists and these were made for apple.

If I do get another, it's going to be hard to reball it without any temperature curve to follow.

Anyway, I'll just wait until I get the SC2602s to continue.