Announcement

**TheLaw** · 04-02-2011, 12:21 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Pictures would be nice.

I don't have a definite answer, but if your standby voltage is off, that's definitely not good.

**sylphhawkins** · 04-02-2011, 07:34 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Definitely, I'll get a close up photo of the underside of the board, and post it up as soon as I can. ^_^

As an update, after a little bit of hunting around online, I found a US ebay seller, selling "refurbished" PSU for this model of iMac, to match the same EEE code, for only $39 inc. postage!, so I've bought one, and it's hopefully on its' way.

I really want to:
a) try it, and hope that the iMac boots!, and that it doesn't turn out to be a logic board issue, =p

and b) to inspect the refurbished board closely, and see what components they have replaced/repaired?

In the mean time, I'll get a photo up tomorrow.

^_^

**TheLaw** · 04-02-2011, 08:50 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

The underside of the board is alright, but the topside of the board would be even more helpful. That's the only way we can see if there are bad capacitors, other components. Usually it won't be the solder traces, unless you've recently dropped it off a cliff or something...

**Toasty** · 04-02-2011, 09:10 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

A quick perusal of eBay shows there are a few numbers that these can go by:

614-0403
614-0420
614-0426
661-4670

The odd voltage on pin 5 -sounds like- it's the on/off control line to the PSU and I would not be concerned. The G5 PSUs vary in a similar manner. The ifixit chart indicates it at zero (0) volts when "on".

Pin 4 may be a temperature signal to a mobo circuit to vary fan speed as the unit temp changes.

Pin 6 might be ground/return?

Refurbished may only mean they blew the dust out of it and tested it.

Found a *HUGE* clear pic on ifixit of the back (attached).
The markings indicate only 2 - 12.1v sources.

The picture appears "clean" around the components you indicated "Q33, D41 and D45". If the -board- is actually discolored, I would be concerned.

The 2007 date puts it past the "bad cap formula" period, but certainly does not exclude it, as the problems continue to the present.

You say you've tested the PSU caps? By what method? On the board or off?

Monitor the voltages closely on the 12v lines when the unit cycles. Results?

Toast

Attached Files

**sylphhawkins** · 04-03-2011, 05:00 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Thanks for your reply Toast, that's an awesome amount of info and resources that you just put forth. ^_^

In regards to the board caps, I removed and tested them (just a voltmeter capacitance test), and then replaced them with new spares that I had on hand anyway, since they were out, and although I managed to do a terrible solder job (my 45w iron just didn't cut it), getting the new caps on the board hasn't changed the situation at all.

You're right about Pin 4 being the PSU temp, i've seen a few notes about it around the net, and it seems to range from ~2v to 3.6v, and vary the fan speed accordingly.

The strange voltage on my Pin 5 is around 2.6v on standby, but it does drop to around 150mv when "power on", so I gather that the mobo is obviously managing that correctly.

I've had a hard time finding out what Pin 6 is for. What do you mean by ground/return, as it's not connected to the PSU ground (Pin 1 + 10), or the other mobo grounds, but lives on an "island" of its own, connected to a 12v rail (the same rail that goes to Pin 9), via 2 16v 680uF caps and a resistor that I haven't been able to measure)

--
I'll have a closer look at the exact 12v voltages tomorrow, and report what I find out. =)
I can clearly see the 2 different 12v rails now that you point out the notes on the board. Pins 2,7,8 on one, and Pin 9 on the other.
Would the motherboard really be affected in this way, by a slightly higher voltage on the 12v lines?, or are you thinking that it might just be showing up as 12v with no load, and drop out under load?

--
On a side note, While troubleshooting it a few days ago, I thought, what the hey, I'll bypass the PSU, and wire up the mobo connectors straight a 12v regulated source that I have, and the board responded in exactly the same manor as with the PSU feeding it, so I initially wrote off the PSU as the problem, but then, after some further thinking and noticing that things were not quite right with the board either, I've rethought that maybe it had something to do with Pin 5 or Pin 6?, which were the only Pins that I didn't connect to anything during my PSU bypass experiment.

I'm wondering if Pin 6 needs to sense some kind of voltage or current on the line in order to kick the mobo into the next stage of POST?

In terms of comparing the brown/discolouring of Q33, D41, D45 + Q32 (I took another look over the area, and realised that Q32 was also affected), on my board, there is a very obvious distinct burnt flux/brown residue on the "south side" of all the components, so: looking at the board as it is in the picture, with the Pin rail at the top, and the sticker at the bottom, the bottom left leg of Q32, the bottom of D41, the bottom of Q33 and the bottom of D45 are all brown / discoloured.

I'll get some high quality close up photos of my board tomorrow, and post them up for comparison, =p

I'm embarrassed about my bad soldering job though, =p

^_^

Thanks for your awesome help everyone, =), I really appreciate it,

- Sylph

**Scenic** · 04-03-2011, 09:59 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

as for the soldering.. a 45W iron should be fine, except on huge ground traces on multilayer motherboards/logicboards.
a good soldering iron doesn't help though if you're using solder without a flux core. flux-less solder will just end up in a mess, as the solder just wont flow properly

i'm using a 35W iron most of the time. works well with leaded (Sn60-Pb40) solder.
the solder i'm using is Felder ISO-Core (simply because it's available locally.. no wonder as it's made here

)
probably not available "down under" though..
http://ecx.images-amazon.com/images/I/41wPbkgw53L.jpg

**sylphhawkins** · 04-03-2011, 08:48 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Ok, Hey again everyone,

I've managed to get some photos close up of my PSU, but unfortunately it seems that my camera just isn't able to get a detailed photo of the whole board, so there's both the wide photo, and a close up photo of the top right hand side components which have the burnt flux colour residue on them.

On the close up photo, you can clearly see the burn out colourings on Q33, D41, Q32 and D45.

What do you guy think?, i've gone over the board and re-soldered any "cold-solder" looking joints, and have also lightly resoldered the legs of the components mentioned above.

Note: the terrible, terrible soldering job on the top left corner of the board, re-soldering the new caps after testing the old removed ones, =p
still, I checked the continuity of them, and everything is attached to where it's supposed to, =p

"On another note, after some more reading, I stumbled across a few posts on the net from iMac users of various models (G4's - up to 2008 intel's) which talk of strange boot and POST issues with failing SMC batteries (or I suppose they could still be called NVRAM batteries), one user noted that he swapped out his failing (it measured 3.08v) cell, for a new one (measuring 3.35v), and that it helped the imac at least get past Post?,
Surely this would still only indicate the early stages of a failing PSU or some kind of voltage control system for the NVRAM battery yeah?,

I vaguely remember articles about PSU that apple produced in various iMac models that required the 3v sense from the mobo battery to kick start the PSU's "power-on enable" system, but I thought that when apple changed from Open firmware to EFI that they stopped storing boot required info, etc in the battery.

Still, it's a lead, so I'll get a new battery and report back.
^_^

- Sylph

Attached Files

**Toasty** · 04-03-2011, 09:40 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Nothing wrong there. Just factory sloppiness.

Some alcohol on a toothbrush or similar will clean that flux up.

Burnt would appear as a dark color -in- the board, not the flux.

**sylphhawkins** · 04-05-2011, 01:16 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Ok great, thanks Toast, at least that rules that out.

I tried a new NVRAM / SMC battery today, and it hasn't made any difference to the problem. A "refurbished" PSU is on it's way, so I'm hoping that the problem is the PSU, and not the motherboard, but I guess I'll find out when the new one arrives.

I'll get in an check the voltages tomorrow too, and post up what I find running on the 12v rails.
^_^

- sylph

**sylphhawkins** · 04-10-2011, 01:33 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Sorry for the delay in getting back to the topic guys,
I've had a few developments in the iMac's non-booting, no-chime, but power on problem

After a long and tedious process, =p , I discovered that I could actually get the iMac 2007 to boot with chime, when the SMC reset "switch" point was shorted out on the logic board (it's on the underside, next to the CPU, and also has a friend, simply named "SYS reset", both are blank mounting points, but can just be shorted out across the terminals (it seems to just short the pin to ground anyway).

This would allow me to boot the imac for various amounts of time, but it would ALWAYS freeze, lock up, etc.., and require reboots, which would slowly "reduce" the effectiveness, until it was back to "no boot, no chime".

I found that removing the PRAM battery all together, would help to seem to keep the rebooting "fresh" after shorting out the SMC reset points, but still, the system would eventually always lock up, and deteriorate, over a period of mins or hours, but I never got it to last longer than 2.5 hours.

The next strange power-related issue that I noticed was that upon plugging in the iMac to the wall, and turning it on (at the power point, not the power on switch on the mac), it would automatically turn it self on, and go through the "no boot, no chime, reset, try again, still nothing" process, however I could manually turn it off using the power switch on the imac, and it would stay off, so this got me thinking about a possible short somewhere, or something confusing the SMC controller to trip into "power-on" mode automatically.

----
Where this project took a new turn, was when I thought about checking the GPU die and heatsink for any "overflow" of thermal paste that might have leaked down into the chips and be causing some capacitance/shorting issues.

When I pulled the MXM video card (this imac has an ATI radeon HD 2600), and removed the heatsink, I found that (as expected with manufacturers these days), it was LACED with thermal paste and "spongy" foam.

The GPU die has a "plastic spongy rim" around it's edges, which was stopping any mashed thermal paste from overflowing onto the board, but it needed cleaning up off it's own on-chip components,

but mainly, the DDR3 RAM chips (I'm assuming that's what they are) were covered in a white spongy foam, which had squished out over most of the components around them on the board.

So I cleaned up everything and removed the spongy white foam from the board and heatsink, reapplied some arctic silver, and replaced the heatsink and GPU onto the logic board.

This time, when I powered it up, I was shocked that it booted quickly, and chimed straight away!. I wasn't expecting that, so I still had the LCD off. I thought that I would quickly turn it off and put the LCD back on to test it properly, and noticed a slight burning smell as I went to switch it off. (coming from the GPU board)

I took the GPU and heatsink apart again, and noticed that I hadn't taken off all the white spongy foam (this is all that I can think off as being the issue?), which was still caught between a few components which connect to (what I'm assuming again,) are the power-in pins?

(See the attached photos of the GPU card below. The bits of white spongy thermal pad goo were still lodged among various components around C718, on picture#2 (named: "power-in section?.jpg)

and I traced the burnt smell to around L36 on picture#3 named "burnt smell - L36?.jpg", but it could be any of those components?

----

After cleaning it up, and using a thermal paste solvent to remove everything off the board properly, I put it back in, and the LED diag situation had changed.

Now, LED #1 would stay lit as normal when wall power is on,
but instead of LED#2 and #3 lighting up, and it going through the "no boot, no chime" routine, when I pressed the power-switch on the imac, LED #2 and #3 flash for a second, and then turn off, back to just LED #1 remaining on.

The fun part?
Is that If I remove the GPU all together, and re-install the logic board, it powers on perfectly!, LED's #1, #2 and #3 light up, and it chimes. I can remote login (I put in a HDD to test with 10.5 on it), and it's working fine as far as I can see,

---

So now, it's even more confusing. =p

I have tried to get the iMac to go back to its' "no-boot, no chime" routine, but I can't, it will POST, chime and boot perfectly every time, so long as the GPU card is removed.

, Initially, I thought that it may be the heatsink shorting out something along the SMC circuitry that it runs over (albeit with spongy foam protectors and what looks like a static free sticky tape over the components), but even with the heatsink removed, and just the bare GPU installed, the same problem occurs.

--

Just to finish off reporting on my findings thus far, ^_^

It was still having the "automatically turning itself on when plugged into the wall power" problem, even with the GPU out, so while I had out the solvent for removing the thermal paste from the GPU, I dipped an old toothbrush in the solvent, and scrubbed the logic board up near the MXM video card socket (it has a few polymer caps mounted on it on the otherside, as well as various surface mount chips).

Now, the "automatically turning itself on" problem is gone, and it sits there and waits for me to use the power switch as it should. ^_^

--

So, this has changed my whole approach as you can imagine.

What i'm faced with now, is still wondering if the PSU is at fault, and perhaps is simply not able to provide enough stable power to the logic board when the GPU is drawing a load as well?
but I am still waiting on a refurbished PSU to arrive, so when it does, I can
hopefully test it out, and either jump with joy that its' going, or at least rule out the PSU as the fault.

What also has me thinking, is that initially, the board booted and chimed with the GPU in after taking off the heatsink and removing most of the GPU thermal paste/goo, but then I smelled burning from the GPU card, and it would not work again after that, so I also wonder if in fact, the whole problem has been caused by the leaky thermal paste on the GPU, causing issues with the other on-board components, and that if I had removed it all correctly in the first place, that it would have solved my problem??

Has anyone had issues like this with their imac GPU's or similar, and can provide any suggestions?,

I've attached a photo of the GPU board (top and underside), and pointed out where I pin-pointed the burning smell on the board.

Do you think that maybe it's just a transistor that's blown? (there's no obvious burn or blown marks on any of the components that I can see), or maybe has broken down, and now isn't switching correctly?

--

I'm really hoping that at the end of the day, I can repair this iMac to a perfect working condition, not only for myself, but also as a reference guide for everyone else who has these problems (or the symptoms of them), and is actually interested in FIXING it, not "replacing the logic board", which seems like the cop-out solution to an obviously wide-spread problem, =p

^_^

Thanks again for your help everyone,
-

Attached Files

**Scenic** · 04-10-2011, 02:56 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

congrats. sounds like you've successfully killed that HD2600.

those white thermal pads are there for a reason. removing them and just applying thermal paste results in no proper contact to the heatsink..

but even with the heatsink removed, and just the bare GPU installed, the same problem occurs.

**sylphhawkins** · 04-10-2011, 04:58 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Hi Scenic,

Maybe, but the iMac hadn't been on for long enough for the GPU to heat up to a point that would damage the DDR3 chips, especially as they were still cool to the touch when I removed the card to re-look over it again.

Ironically, more than the issue with the HD2600, I'm excited that I've been able to get a bit more consistent control over the whole iMac problem, and work it down to a fairly consistent (if this : then that) way of replicating the Power and Boot issues.

Also, given that the logic board seems to be working great again now, headless without the GPU, perhaps the GPU has always been the culprit at fault, and may have been defunct from the beginning?

**Scenic** · 04-11-2011, 02:02 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

hm.. well it might be the main problem.
given the fact that these things run pretty hot anyways, the HD2600 really doesn't help in there.

one problem might be to get a replacement though. on tower macs, you can't just put a normal PC video card. they need to have a special BIOS flashed to the card before it'll work.
if the same applies to those MXM cards this could get really tricky (or expensive) :/

**Numnutz2005** · 04-12-2011, 12:16 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Ooohaiii guyzz

Ya im registering to post in this topic so no flamin please...just wanna share my info on a few things in case it will help u because me helpin u reach ur goal will actually help me cause im havin the same prob.

Ok so rrrrr uuuuu readyyyyy??? Lol jk lets get down to bizz harr

I have 4 of these units that i got from a lot govt auction along with a ton of 24 and 27in imacs some new some used. Some wprked great but others had problems. I will start with the 8.1 24in model. Specs for them were 3.06 c2d 4gig of ram and nvidia 8800 video card. There were about 10 units that would KP right wen booting up. Apple claimed it was a bad board and a bad video card...they wanted to charge me over 1k per unit to fix em. I said fuck that and she laughed i joked andwent home. I work on computers for a living...i may only be 22 but over the years i have learned what im told isnt always how things are so i wasnt going to accept apples answer so i took them apart and scanned ever inch of the imac for flaws and found nothing. I got so frustrated with my machines i almost threw them away. I took the video cards out of each unit and removed the white foam...they load it up like crazy and it looks sloppy but to be 100% serious do NOT remove it thinkin it will cause damage. The foam is made to be like corn starch and water in the sense that it forms around chips that need to be kept cool but it forms when you arent messing with it...meaning if you mess with the cornstarch mixture a lot it gets more and more like a solid but when u stop it gets sticky and liquidy but rhe foam on the video cards is made to form but not leak into sensitive areas because it always stays at a particular consistancy. The smell u mentioned wasnt from the thermal foam...it was from ur board heatin up too much and much have caused the plastic part of the housing to start melting...does this mean ur card is dead?? If i was apple i would tell u yep and try to get 500 bucks from u but im not apple so i tell u no....before we get into how to fix the vid cards...do not skip the foam. When u dont have it the heatsync doesnt make contact with the chips it needs to hit to properly cool tge card and if u think the slight smell was bad just try letting it run overnight not knowin what will happen...had one buyer do that shizz to one of my imacs and i told him he was sol he shouldbt have messed with it.

Now to fix the cards...to be plain simple and blunt bake the video cards in a fully preheated oven running at 400 degrees for 8-10 mins. Make sure u remove all the thermal paste and foam before hand and when its done bakin for 8-10 mins let it sit on the counter for 30 mins. Dont
Put it in the fridge or over...if u try to cool a hot glass bowl too fast what happens?? Do u really wanna risk it?? It sounds like it wont work at all but it heals micro solder joints that u cant see and i have resurected a good 30 imacs so far just by doing this...i might even make a youtube vid about the whole process for people that cant find info on how to fix this stuff.

Ok next part is about my 20in 2.0 imacs...these r the 7.1 models. I ordered a new psu and replaced the old one and it had no diff the lights lit up before and did after so that didnt help me. I did the same thing as i described above to the vid card and added new thermal paste to my proc still didnt fix the issue. When i press the power buttons i dont get ne beep codes and only the hdd fan spins so 1/3 works. It isnt overheating...it isnt getting too little power...its hard to comprehend why so many units in my batch of computers died with the same exact issue. Hell maybe it is dead but im like u man...if we get out machines working we dont lose out on money that we worked to get...now i have been following this thread for a bit now and just read the stuff about ahortting things out and gettin a response from ur imac so i havent tested that so if i can get some kind of life instead of a black screen i know its revivable and not dead like apple says lol.

Also when u say remoted into ur imac...did u have a remote app loaded before it stopped working?? If so does that mean i cant remote into it unless i did that?? I have literally never used a mac to remote into a mac...i use my mac to remote into servers and thats easy to do because im use to it but im still and osx noob and will be for the next 6 years haha. Sorry for ne typos...aderol + no sleep + iphone = a fail post but what i habe given is better then nothing and i hope ur psu helps you more than mine did.

**Toasty** · 04-12-2011, 12:27 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Let me light the flamethrower...

Please post in ENGLISH. No shorthand or texting-type talk.

Welcome to the forums.

Toast

**sylphhawkins** · 04-12-2011, 03:12 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Wow, epic post Numnutz2005, thanks for contributing what you've been finding working on the same model imacs, that's awesome. =)

I have thought about baking the GPU to reball the BGA, but I haven't been convinced that its the chip BGA thats the problem, although, I suppose that it could make sense given that it stopped working after removing the heatsink, perhaps the heatsink was adding a good amount of pressure on the GPU, which, when removed, took the pressure off the BGA and allowed a few balls to separate from the array.

I was watching a bid for a faulty imac, but the price went up wayyyyyyyYYY! above reasonable, so I'm back to looking for a whole imac to get the GPU out of to test, or to find a relatively cheap new/used GPU for it.

My PSU should hopefully be arriving before the end of this week, so I'll get it and try it out. I haven't been holding out hopes for it to fix the problem though, as I originally ran the logicboard straight from a 12v regulated source that I have, with more than enough amperage for it, and it still showed the same signs and problems, (although I bought the new PSU because I couldn't find what PIN 6 was doing on the PSU / motherboard, and wondered if there needed to be some kind of flyback or sensing circuitry between the PSU and the logic board to get it going, but I doubt it).

I think that I'm going to wait for the PSU to arrive, and test that with the GPU as it is in it's current state, and then, if that doesn't change anything, I'll give the card a reflow attempt. (I've done this a few times for other various products, but never succeeded,

, but there's always got to be a first time for success, =)

--

Also, in terms of remotely-logging into the iMac with the screen dead, I didn't know how to do it either at first, so I did some reading around, and in the end, this was the solution for me:

- Mount the iMac's HD (in target mode, or in an external enclosure, not in a network link)

- Get the finder to "show all hidden files" by typing the following into terminal. (note, that they are 2 separate lines of code, so "enter" after the first line, then on a new line, enter "killall Finder")

defaults write com.apple.finder AppleShowAllFiles TRUE

killall Finder

- Navigate to the mounted external HD (the iMac's HD), and into the folder /Library/Preferences (from the root of the HD, not the user's /Library/Prefences folder)

- make a new blank file in /Library/Preferences called com.apple.ScreenSharing.launchd

- then unmount the iMac HDD, and restart the iMac with the HD back inside and connected up.

From here, it should "enable" screen sharing on OSX 10.5, and if you navigate to the iMac when it pops up in the Finder side-bar (once their networked together), it should have an extra option under the "Connect As" button, that says "Screen Sharing", and click on that and go from there..

^_^

Note, if you are running OSX 10.6, check out this link for getting screen sharing enabled:

http://hints.macworld.com/article.ph...80318190503111

also, this walkthrough is what worked for me, but I put it together from the info at the above site, and also:

http://discussions.apple.com/message...sageID=8424321

http://macs.about.com/od/tipstricks/qt/hiddenfolder.htm

-------

Once you're all done, you can hide the hidden system files in OSX again, by going back into Terminal, and typing:

defaults write com.apple.finder AppleShowAllFiles FALSE

killall Finder

-------

Just as a quick question though, have you tried taking the GPU out of the imac7,1 and seeing if it boots up and chimes?,

then you could see if you can remotely log-in (once you've got screen sharing working, you can go through and enable SSH and remote logins on the iMac directly, without having to go through terminal.

I really appreciate having your help to work out this problem! ^_^

**sylphhawkins** · 04-18-2011, 06:59 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Hi everyone,

Just a quick update:

The refurbished imac PSU arrived today, and installing it has made NO difference at all to the problem + symptoms, which currently are:

iMac will NOT chime or boot, (+ black display) with the current ATI Radeon 2600HD video card installed,

but the iMac will boot and chime fine, and can be remotely accessed and controlled (SSH and screen sharing) when I remove the video card.

So, I'm going to give the Radeon 2600HD the old home-over "reflow" trick, =p, but I have sourced a replacement 2600HD video card Apple OEM, for $130, so if re-heating it doesn't work, I'll purchase the replacement and hope for the best. ^_^

- Will keep you all updated
- Sylph

**sylphhawkins** · 05-02-2011, 04:55 AM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Hi Everyone! - Sorry for the long quiet delay on the posting front, been busy busy. ^_^

I thought that this post might have been the "all working again" final post, as after receiving the replacement 2600 HD video card, the iMac went back to exhibiting it's usual behaviours, sometimes it would chime, sometimes not, and it would always freeze by a few mins in at the best of times, but the normal freezing time was before boot, during the spinning gear.

Then, I decided to tighten the CPU heatsink clamps as tight as they would go, and at the same time, change the RAM that I was using.

Being a 2007 iMac 7,1 it is spec'd to take PC-5200 SODIMM's, but I also had a few slower PC-4200 sodimm's lying around, so I thought, "what the hell", and threw them in.

As usual, it would start up and chime and boot after a few tried turning it on and off after it would get stuck in a "fan spin up", then turn off, then turn on and spin up again" loop.

Then, low and behold, it seemed to be running somewhat stable?

iStat menus showed the CPU core voltage to be down around 0.90v on average, and strangely, it showed up the GPU's voltage (the Radeon 2600 HD Pro) as 0.01v ?

CPU amp draw would up and down between 1.1amp and up to 14amp ! "phew", im sure that the traces on the board aren't rated for that, surely??

anyway, it also again showed the GPU current as a steady 0.00amp?

After a benchmark test to run it hot for a while, it froze again, a few mins in, and the temps looked alright (I had the menu open watching them just before it froze), PSU temp was up around 76*C, and CPU around 35*C, GPU around 40*C and amazingly the AirPort was reading over 72*C!!!!,

--- Anywho, the interesting part is when I rebooted it, and was concerned about the low CPU core voltage, as I remembered seeing odd CPU issues on older macs that I'd overclocked, but where the voltages weren't keeping it stable, so I decided to turn off one of the cores to see if it would stabilise the voltage.

Using terminal to do this, I typed:

sudo nvram boot-args="cpus=1"

and checked that it had be correctly set, by typing:

nvram boot-args

after that, I restarted and it chimed and booted into OSX 10.6 (I installed a 1Tb HDD for testing with 10.6 installed).

Amazingly, it ran smoothly!, istat showed the CPU usage to be only 1 core, and the core voltage was an average 1.17v

The GPU still showed 0.01v and 0.00amp thought, which worried me.

The iMac ran, booted and shut off time and time again, played movies, got moved around, and put all back together, and I thought that perhaps either limiting the CPU to use only 1 core, or dropping the speed of the RAM that it was using might have gotten around whatever the problem was!

---

Alas, after about 1 week of great use, the symptoms arrived again one night, out of no-where, strangely after a long stint of iPhone unlocking.

I tried changing the RAM again, using both 667Mhz and 533Mhz modules to no effect,
and I was able to get it to boot a few times and checked that it was still only loading 1 cpu core, but then it would freeze, barely 2 mins into osX or winXP.

So, Hopefully this info might trigger some-one else to join the thread whose been seeing similar things on their iMac.

One main thing that I am interested in is whether anyone else can confirm that istat menus shows the GPU as having 0.01v and 0.00amp on a known working iMac 7,1??

Can anyone confirm this?

- "Sigh", back to the drawing board for this one, =p

**keebler64** · 06-01-2011, 10:01 PM

Re: Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Registered just for this thread, although I see great things coming from this site, as I do component level repair in my spare time.

Anyway, great work in troubleshooting your consistent power/stability issues. I have a couple of late 2006 24" iMacs. both with the 7300GT MXM GPU. As you're probably aware, they both suffered from thermal breakdown of the BGA balls on the GPU due to the poor ventilation and inadequate fan speed, further exacerbated by the lack of an OEM method of fan speed control (thank you iStat!)

All of that said, I can attest to Numnut's reflow method, or at least to his hypothesis/tests. However, I used a mix of alcohol, and a propane torch in a pinch. I do not have access to a real hot air or infrared SMD repair station.

To reflow the GPU, I removed all thermal paste, gave it a very thorough cleaning, and then laid out alum. foil over the whole surface, cutting out just the GPU. Leaving the white border on, I poured a very small amount of rubbing alcohol onto the GPU. Took my torch, and lit it on fire...The alcohol will not burn for very long, and eventually burn itself out. Afterwards, I put my torch on the lowest setting, and for about 5 seconds went over the GPU.

Yes, you're probably balking, screaming, or calling me any number of insults, but I only paid $20 for a dead card on ebay, and now, I have it, plus my other two GPUs in perfect working order. Numnutz's solution is of the same type of DIY, except I'm only baking the GPU (for the most part), saving the other components from greater thermal abuse.

A caveat to the torch. A hot air gun would be much more beneficial, if you stop in one spot for even a second, can cause horrible damage. Duh.

That said, as to your problem. Have you tried swapping CPUs? (I can't remember if the 2007 has a user replaceable CPU.) Failing the CPU swap, perhaps attempt the same reflow method on the CPU (if it's embedded?) Or, perhaps invest in a BGA setup, sell your services on Craigslist for RROD, and the kit pays for itself.

Good luck, I'll be watching this for further progress.

Announcement

Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Aluminium intel iMac (2007 model), wont POST, no chime, bad PSU?

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment