SMC Issue on MBP 13 - Hardware Troubleshooting Guide

[papaille44]

Hi everyones (sorry for my poor english, i'm french )

Post is very impressive, losts of infos in it!

In my way, i brought a 15" mbp unibody mid 2010 (I5 proc and 4go RAM) with coffe damage.

In the past i've backed to life a 13" mid 2010 (wisky coca damage), 90° alcohol and friends make it alive.

But today, couple don't make my 15" alive.

First, if I plug magsafe directly to the mac, nothing append. I have to put it on another mac to get the green light. Then when i plug it to the faultly 15", it turn green.

When i mesure voltage on magsafe board, i got 16,7V on power line, and 3,87V on sense (when light green on).
I've tried to shortcut powerpads, nothing. (with an without battery).
Powerpads with magsafe on reads 3,37V on one an 0V on the other.

Batterie is out (2,37V mesured)

The way i dont understand, on the magsafe bloc, i read 16,5-18,5V power 85W. But i only have 16,5V to the board, never 18,5V. Does the logic board can start up with 16,5V? Faultly power magsafe?

Thank a lot, and sorry again for the headhake trying to read me

 

[hadbav]

Hello,

First - I have searched for solutions for hours now, but nothing works, or I am lacking the ability to code the script properly

Issue - My fan sensor is broken and reads as 0 RPM, with i stat nano, SMC control and HD fan control. The fan runs at maximum speed no matter what the laptop is doing. I have tried the SMC / PRAM reset but it does not fix anything.
Temperature is fine, usually a cool 40*

Is there an easy solution script or program in laymen's terms that I can run to set a maximum fan speed to bypass the faulty sensor? or is my only solution a new logic board?

 

[Dadioh]

Hello,

First - I have searched for solutions for hours now, but nothing works, or I am lacking the ability to code the script properly

Issue - My fan sensor is broken and reads as 0 RPM, with i stat nano, SMC control and HD fan control. The fan runs at maximum speed no matter what the laptop is doing. I have tried the SMC / PRAM reset but it does not fix anything.
Temperature is fine, usually a cool 40*

Is there an easy solution script or program in laymen's terms that I can run to set a maximum fan speed to bypass the faulty sensor? or is my only solution a new logic board?

I have used smc fan control to set a max speed in the case of a MacBook that had a sensor issue and wanted to run the fan fast. I made some scripts that could be used to set t to three different speeds since it reset to fast fans after waking from sleep.

However, this relies on the fan control reading the fan tach signal. It would no be able to work if fan tach is always reading 0.

Have you had a close look at the connector on both the fan and on the logic board to make sure all 4 pins are making contact?

 

[movino4me]

Macbook Pro 13" unibody smc problem

I have a fluid damaged unit. Replaced the logic board and the charge light comes on and the battery is charging but the unit will not start up. G3Hot has 3.4 volts and I tried to short the circuit to hot jump the smc but stil will not start. Have done the smc reset with no response also. Any thoughts on what to check? I have also tried another magsafe to be sure.

 

[turbzz]

I've had another repair success. It is an I-series A1286, 820-2936-B. Fault symptoms were no battery charge (charge led stayed green) and no power on (pressing power button did nothing). I checked for all the main voltages and found that I had lost PP3V3_S5. With just the magsafe connected it should read 3.3v on the multimeter but i got 0v. Looking at the schematic, U7201 provides the 3.3v and 5v. The 5v is only present when the logic board powers on. The supply voltage to U7201 was ok so I knew it had to be a faulty IC. I replaced it with one from a scrapped board and when i checked at L7260 i was reading 3.3v. I connected the keyboard, added the RAM, fan and speaker, pressed the power button and the board powered up and I heard the boot chimes. Powered off the board, connected the LCD, powered on again and the LCD came to life with the flashing icon..success

regards
Rog

Thanks this helped with one of the boards I had lying around

----------

Hi All

Have managed to solve a couple of the boards I had lying around

I would like some advise please... the lp8550 chip has been previously removed from this board by the former owner

It just does not look right what left on the pads and lines

also there is a resistor missing... whats the value on the ones around the lp8550?

the other problem is that it will not boot... I get a flicker on the screen.. but nothing... no fan... no chime.... it will not even start via a force smc start via the magsafe and power button

Any pointer where to look? check?

Thanks guys for the awesome threads

 

[Joris01]

Before doing anything drastic, disconnect battery and magsafe. Clean magsafe PLUG and CONNECTOR with isopropyl alcohol. Ensure that contact pins on the plug is not stuck. It should be somewhat springy. If it is a bad DCIN board, be careful when testing replacing it as many people in this forum accidentally short it to the logic board causing more serious damage while testing out of the chassis.

Hello there again, it's time for an update!

I received the new Magsafe-board and installed it... Plugged in the Magsafe-charged and voila! A bright green light on the magsafe!

It turns on like normal (that means without the SMC bypass!). System runs like a charm, nothing weird, except for one thing.

It says "battery not charging" whilst having a bright green light on the magsafe. Tried an SMC reset, but did not help.
Before my Magsafe-board replacement it showed an "X" when booted with SMC bypass (meaning battery was not being recognized). This time it does recognize my battery I think.. It says "not being charged" See image below for batterystats.

So what should I do now? Is it possible that both my Magsafe-board and Battery got fried at the same time (i'm suspecting a power shortage when my old charger's cable got ripped in two pieces). The problems started after this happened..

Thanks again!

----------

Edit: Also see Coconut battery

 

[Joris01]

Eureka!

I tried another SMC reset, this time with replugging the magsafe charger.
It now charges again and everything fixed .

This forum helped me a lot! And i'm very thankful for this community.
By reading what other people experienced and doing "try and error" testing (based on the symptoms I read from other people) I managed to fix it as a non-technician. I bet the Apple Reseller would have charged me for a new logic board! I advise people to try this aswell but do it with precaution, you do not want to damage the macbook because of clumsyness! And be careful when replacing the magsafe-board, I accidently (and I even was warned for this!) put 1 screw through one of the magsafe-board cables, meaning I had to buy a new magsafe-board again. So this fix costed me about 60 euro's and a pair of screwdrivers (10 euro)...

 

[papaille44]

Hi everyones (sorry for my poor english, i'm french )

Post is very impressive, losts of infos in it!

In my way, i brought a 15" mbp unibody mid 2010 (I5 proc and 4go RAM) with coffe damage.

In the past i've backed to life a 13" mid 2010 (wisky coca damage), 90° alcohol and friends make it alive.

But today, couple don't make my 15" alive.

First, if I plug magsafe directly to the mac, nothing append. I have to put it on another mac to get the green light. Then when i plug it to the faultly 15", it turn green.

When i mesure voltage on magsafe board, i got 16,7V on power line, and 3,87V on sense (when light green on).
I've tried to shortcut powerpads, nothing. (with an without battery).
Powerpads with magsafe on reads 3,37V on one an 0V on the other.

Batterie is out (2,37V mesured)

The way i dont understand, on the magsafe bloc, i read 16,5-18,5V power 85W. But i only have 16,5V to the board, never 18,5V. Does the logic board can start up with 16,5V? Faultly power magsafe?

Thank a lot, and sorry again for the headhake trying to read me

UP? thanks a lot

For remember, i got G3Hot ok (3,37V), but nothing when trying to start. Another strange thing is when i directly plug magsafe nothing happend, when i plug it after plug to an other mac (making light be green), light goes on and power is able to the board.
Another precision, battery is totally out. But problem still the same if i disconnect it or if it's connected. A way to charge the battery without the mac, and trying to boot with a charged battery?

Thanks a lots, i really want to make him back to life, i'm sure it's only a little thing, but what...

 

[laptech]

Another success - 820-2936-B K90i

Hi,

Have just had another repair success. This time it is a 820-2936-B K90i logic board. Fault symptoms were it would power on, the fan would spin but it would not post, no chimes and no display. After checking all the main voltages, i started to look deeper into the CPU side of things because it was cold to the touch. Experience has taught me that if a BGA chip is working, be it CPU, Northbridge, or GPU, it gets very warm to the touch, almost to the point where the heat of the IC starts to hurt your fingers when you touch it. In my case, the Northbridge was getting very hot so I knew to some degree that it was working but the CPU chip was only luke warm. When this chip is functioning properly it gets extremely hot but it was not doing it in this case so that basically told me to start focusing on the CPU side of things.

I checked around a few capacitors (C1683, C1681, C1680 & C1682) to see if i had one of the main CPU voltages. On the faulty board i got 0v but when i checked against a working board, the voltages on those capacitors was around 0.8v. This told me I was missing one of the main CPU voltages. I checked on the schematic and found i had lost PPVCORE_S0_CPU. Tracing this signal back to it's source, Q7520 i found this component was not functioning properly. Upon further checking I found that the input signals to Q7520 where missing. Therefore tracing further the source of those signals I found that they originate from U7400 (main CPU control IC).

This now meant I had to figure out why U7400 was not working. I checked to make sure i had 5v (PP5V_S0) on pins 15,24 and 40, which I did. Please note you will only see the 5v on those pins when the board is powered on due to the power on state signal S0. I went back to the working board to make sure what voltage was present on pin#2 CPUIMVP_TON which was 12.5v. I subsquently found out from the working board that this voltage is present when the board is powered off.

I checked the faulty board and guess what, this voltage was not present so i checked the input of the this voltage which passes through a resistor R7402 and low and behold, i had 12.5v on one side (PPBUS_S5_HS_COMPUTING_ISNS) of the resistor but not the other. This told me the resistor had to be faulty. I therefore replaced it with one from a scrapped board. Just had the magsafe plugged in but did not power on (same as i did with the working board), measured pin#2 of U7400 and now i got 12.5v. I therefore connected everything back onto the logic board, RAM, FAN, LCD. Powered on and it displayed with the flashing icon...another success.

regards
Rog

 

[cmdrdata]

Hi,

Have just had another repair success. This time it is a 820-2936-B K90i logic board. Fault symptoms were it would power on, the fan would spin but it would not post, no chimes and no display. After checking all the main voltages, i started to look deeper into the CPU side of things because it was cold to the touch. Experience has taught me that if a BGA chip is working, be it CPU, Northbridge, or GPU, it gets very warm to the touch, almost to the point where the heat of the IC starts to hurt your fingers when you touch it. In my case, the Northbridge was getting very hot so I knew to some degree that it was working but the CPU chip was only luke warm. When this chip is functioning properly it gets extremely hot but it was not doing it in this case so that basically told me to start focusing on the CPU side of things.

I checked around a few capacitors (C1683, C1681, C1680 & C1682) to see if i had one of the main CPU voltages. On the faulty board i got 0v but when i checked against a working board, the voltages on those capacitors was around 0.8v. This told me I was missing one of the main CPU voltages. I checked on the schematic and found i had lost PPVCORE_S0_CPU. Tracing this signal back to it's source, Q7520 i found this component was not functioning properly. Upon further checking I found that the input signals to Q7520 where missing. Therefore tracing further the source of those signals I found that they originate from U7400 (main CPU control IC).

This now meant I had to figure out why U7400 was not working. I checked to make sure i had 5v (PP5V_S0) on pins 15,24 and 40, which I did. Please note you will only see the 5v on those pins when the board is powered on due to the power on state signal S0. I went back to the working board to make sure what voltage was present on pin#2 CPUIMVP_TON which was 12.5v. I subsquently found out from the working board that this voltage is present when the board is powered off.

I checked the faulty board and guess what, this voltage was not present so i checked the input of the this voltage which passes through a resistor R7402 and low and behold, i had 12.5v on one side (PPBUS_S5_HS_COMPUTING_ISNS) of the resistor but not the other. This told me the resistor had to be faulty. I therefore replaced it with one from a scrapped board. Just had the magsafe plugged in but did not power on (same as i did with the working board), measured pin#2 of U7400 and now i got 12.5v. I therefore connected everything back onto the logic board, RAM, FAN, LCD. Powered on and it displayed with the flashing icon...another success.

regards
Rog

Very nice and methodical detective work.

 

[chigwelldave]

Hi,

I have a 15" 2915 i7 logic board with some strange issues after liquid damage (looks like coke). It only powers from battery although the mag safe lights green. I've measured around the charger chip U7000 and see it is not switching on the charger S gate FET so no power or battery charging is possible. I can't find a boardview file so troubleshooting is a little tricky but I found R7080 and R7081 and they read fine.
When switching on the board will start to boot, HDD spins, ODD starts to reset but then it switches off before the chime. Powering on again completes the ODD reset then off again. I have had it boot fully then ASD reported everything fine, except no charger attached when it was connected, but after I shut it down to look at the External DC path it's back to the boot-off sequence. Strange thing was the screen brightened when the charger was connected. Probably 2 separate faults, but any ideas anyone?

 

[papaille44]

Hi everybodies (again)

so i continue my investigations, and i want to charge my macbook pro 15" mid 2010 batterie, outside of the case (motherboard no charging), to try boot on it when little bit charged.

Precedent post showed an extern alimentation directly pinned to the batterie, and charging it a little. This is what i wanna do.

I have found the DC+ (15,6V) and - on my external power (90W, 15,6V to 19V, 4,2A). But on my macbook batterie, because there is no way to determinate wich connector is the wright or wrong, and because a mistake can be really dangerous on li-ion polymère... I want to know if someone can tell me who is who.

My laptop:



My interresment:



I just want to plug DC+ from external power to dc+ batterie, same thing with - (ground). A control it with multimeter DC 20V calibrate.

Batterie pins show 0V at every borns today. So nothing can be verificated..

Thank a lot!

 

[cmdrdata]

Hi everybodies (again)

so i continue my investigations, and i want to charge my macbook pro 15" mid 2010 batterie, outside of the case (motherboard no charging), to try boot on it when little bit charged.

Precedent post showed an extern alimentation directly pinned to the batterie, and charging it a little. This is what i wanna do.

I have found the DC+ (15,6V) and - on my external power (90W, 15,6V to 19V, 4,2A). But on my macbook batterie, because there is no way to determinate wich connector is the wright or wrong, and because a mistake can be really dangerous on li-ion polymère... I want to know if someone can tell me who is who.

My laptop:

[url=http://imageshack.us/a/img11/8972/6iqw.th.jpg]Image[/URL]

My interresment:

[url=http://imageshack.us/a/img809/1382/akwf.th.jpg]Image[/URL]

I just want to plug DC+ from external power to dc+ batterie, same thing with - (ground). A control it with multimeter DC 20V calibrate.

Batterie pins show 0V at every borns today. So nothing can be verificated..

Thank a lot!

Internal MBP battery typically has 9 pins: 3 pins for - or GND, 3 pins for + or POS, 2 pins for serial I2L bus and 1 pin for sense/enable line.
The I2L is a bus, with the SMC being the bus master and one or more devices connected to this bus as the bus slave. This line should normally be pulled up to 3.3v by pull-up resistors connected to G3HOT. The devices driving these lines toggle the lines to ground for short period of time, thus creating pulse Only one device can drive these lines, one at time signals. The pulses on the data line (SDA) are coded so that it sends an address code and then data code. They are also accompanied by a clocking signal on the SCL line.
The sense/enable line is pulled low to 0v when the battery is plugged into the logic board by a 10KOhm resistor to ground. If the battery is still good, it will enable the battery to output the battery output voltage on the +/- lines. If the battery is in "self-protect mode", it will completely disable itself from outputting any voltage, and will not even respond to SDA/SCL communication. On one battery I have, I can measure battery voltage with a voltmeter by grounding the sense/enable line. Charging it offline is DANGEROUS and you have to know a lot about Li-Ion battery charging method.

 

[papaille44]

Internal MBP battery typically has 9 pins: 3 pins for - or GND, 3 pins for + or POS, 2 pins for serial I2L bus and 1 pin for sense/enable line.
The I2L is a bus, with the SMC being the bus master and one or more devices connected to this bus as the bus slave. This line should normally be pulled up to 3.3v by pull-up resistors connected to G3HOT. The devices driving these lines toggle the lines to ground for short period of time, thus creating pulse Only one device can drive these lines, one at time signals. The pulses on the data line (SDA) are coded so that it sends an address code and then data code. They are also accompanied by a clocking signal on the SCL line.
The sense/enable line is pulled low to 0v when the battery is plugged into the logic board by a 10KOhm resistor to ground. If the battery is still good, it will enable the battery to output the battery output voltage on the +/- lines. If the battery is in "self-protect mode", it will completely disable itself from outputting any voltage, and will not even respond to SDA/SCL communication. On one battery I have, I can measure battery voltage with a voltmeter by grounding the sense/enable line. Charging it offline is DANGEROUS and you have to know a lot about Li-Ion battery charging method.

Thanks a lot cmdrdata!

Recent batteries does have a protection to prevent inversion/overvalues/down charging. I know external charging is critical, because only protection i have is the batterie protect. MBP and is logicboard/security system not present.

If i understand right, putting a 10Komh between center pin (5 => SDA/SCL) and ground or 12+ is only to discharge it.

My montage is this one, tell me if i'm correct?

Is there a way to know if my batterie is in self protection mode?

thanks much

 

[annk]

A wiki would be really helpful, so that one can do this per model in a
sorted manner and to gather facts / quirks / solutions.

Yes, great idea! This way everyone could contribute. This thread is great but we need a way to organize this information.

I would be willing to set up and host a wiki site if we can get enough knowledgeable people interested. Let me know what you guys think.

Sorry. Not yet. Very busy with other things at the moment. But still in my plans at some point.

I would certainly help carrying information from this thread to a wiki.
However, I don't think a privately hosted wiki is a good idea (costs, the
content may get lost if the owner is not interested anymore, software
gets outdated, maintenance, ...).

How about a hosted wiki from, say, wikia.com?

Either way would work, I was just going to look into adding a MediaWiki to my current website, set it up and host it on there. What ever I can do to help I'm in .

Sounds good, I and many others will be looking forward to it.

Reposting info from MacRumors on another site or using threads here to drive traffic to another site isn't allowed under our rules, but we encourage members to create guides on MacRumors. It might be just what you're looking for. Check them out here.

 

[cmdrdata]

Thanks a lot cmdrdata!

Recent batteries does have a protection to prevent inversion/overvalues/down charging. I know external charging is critical, because only protection i have is the batterie protect. MBP and is logicboard/security system not present.

If i understand right, putting a 10Komh between center pin (5 => SDA/SCL) and ground or 12+ is only to discharge it.

My montage is this one, tell me if i'm correct?

Image

Is there a way to know if my batterie is in self protection mode?

thanks much

As shown in the examples below, MBPs manufactured at different time/year have slightly different pin signals. See the example below between M98 (early 2009) and K6/K24/K90i (after 2009) models. So in your case I don't know if Pin-5 is the enable. Look at the schematic for your model. I believe if you are reading 10v between the -/+ pins, your battery is not in shutdown mode. It has low charge and in need of complete charging. Internally the pack has 6 connected in 3 pairs, thus a fully charge batterie
s should be 3 x ~4.2v = ~12.6v. The pairs doubles the power capacity (pairing is often done in Li-Ion batteries because of Li-Ion discharge profile).

 

[Leehs]

Hi all, my first post here.

I have a 15" MacBook pro 2008-2009 I believe. It has no evidence of fluid damage. I have tried the bypass, G3HOT, and keyboard pin 5 to ground and it doesn't respond to any. I have done the basic tests like switching out ram, disconnected HDD and DVD etc.

Magsafe charger and battery are both brand new working items. I have tested the DC jack, white fuse to the left of the memory ports. G3HOT has 3.31V with charger or battery alone - I have included the results in the attached image.

The charger light will turn orange when the battery is in and seems to be charging. When I press the button on the battery all of the LED's light up and the last one flashes 3 times.

Sometimes when I connect the charger to the dc jack, the green light will not come on, but I still get voltage to the board with similar reading that I have included in this post below, albeit 0.2v higher on the center signal pin on the dc in board.

Is there supposed to be voltage either side of the green component in the image?

Info doesn't show up too well in the image:

DC in: 0, 16.71V, 3.32V, 16.71V, 0
Fuse to the top (left of ram) 12.21V
G3HOT: 0V on the top pad, bottom pad has 3.31V

Where should I look next? suggestions would be highly appreciated.

 

[Snake-Doc]

Awesome thread, hoping someone here can help me out.

I came to possess a late 2011 2.2ghz MBP 15" that can no longer recognize the built in keyboard or trackpad. External USB devices work just fine. Macbook boots no problems otherwise.

I have read up and see the MBP uses the CY8C24794 PSOC chip for USB emulation and communicating with these two devices. Would I be correct in assuming this going bad would result in no detection of these two devices? Is there a fuse/s that could be causing this issue as well?

How can I confirm if this is functioning correctly? Is this something that can be swapped out with an off-the-shelf replacement chip or will I need to salvage one due to some programming on the chip?

When I ran AHT I am getting a TsOP error which is related to the trackpad temp sensor I do believe - so I know I will like need to replace trackpad regardless.

Also, if anyone has the schematic or would like to tell me if there is any testing I can do on the board to troubleshoot, that would be great - thanks!

 

[cmdrdata]

Awesome thread, hoping someone here can help me out.

I came to possess a late 2011 2.2ghz MBP 15" that can no longer recognize the built in keyboard or trackpad. External USB devices work just fine. Macbook boots no problems otherwise.

I have read up and see the MBP uses the CY8C24794 PSOC chip for USB emulation and communicating with these two devices. Would I be correct in assuming this going bad would result in no detection of these two devices? Is there a fuse/s that could be causing this issue as well?

How can I confirm if this is functioning correctly? Is this something that can be swapped out with an off-the-shelf replacement chip or will I need to salvage one due to some programming on the chip?

When I ran AHT I am getting a TsOP error which is related to the trackpad temp sensor I do believe - so I know I will like need to replace trackpad regardless.

Also, if anyone has the schematic or would like to tell me if there is any testing I can do on the board to troubleshoot, that would be great - thanks!

If it was mine, I'd replace the keyboard trackpad as a first attempt to fix it before replacing any chips on the logic board. That would be much riskier repair attempt. Since your USB port works, there is a likelihood that your keyboard/trackpad is bad.

 

[Snake-Doc]

If it was mine, I'd replace the keyboard trackpad as a first attempt to fix it before replacing any chips on the logic board. That would be much riskier repair attempt. Since your USB port works, there is a likelihood that your keyboard/trackpad is bad.

Is there any way with System Profiler to confirm the USB hub they use is present? I would much prefer a simple hardware swap over a qfn repair for sure.

Thank you!

 

[varungargi]

Hi

i know its a forum for MBP but my query is with MacbookAir(MBA).

but i have a MBA 2012 which had suffered liquid damage . .
clean it with isopropyl alcohol . .

The problem is the battery comes up with a cross when it boots.
The MBA boots up fine when connected to macsafe connector.

I tried the battery in a another MBA the battery works fine . . .
I also checked the white color fuse near the battery seems to be okay . . .

Need your help here guys to look for the faulty component on the logicboard . . .

I know its a forum for just MBP , if any one can redirect me to forum or thread like this for MBA would be appreciated or where i can start the debuging from.

thanks

 

[Leehs]

I'm still wading through this fantastic thread to check for anyone with the same problem as me but I'm only up to page 28.

From my request 5 posts up from here. I'm beginning to wonder if it has a GPU issue causing it to not power up. I know generally you still get the boot chime etc with most gpu issues, but I do know in some cases the machine will not respond to powering up at all. This one has me flummoxed. I would love to breathe life back into this thing. Apple quoted me £380 for a new logic board

Could it even be the dc in board, even if its showing relatively normal voltages? It's just sometimes when I connect the MagSafe adaptor it'll not show the green light but still show voltages on the cd in board. The light will come on if I move the MagSafe plug slightly. This is odd because it is a brand new apple psu that works fine in my older MacBook Pro. The pins on both the psu and dc in are clean and free.

Any help to reduce the grey hair this is causing, would be highly appreciated

 

[cmdrdata]

Is there any way with System Profiler to confirm the USB hub they use is present? I would much prefer a simple hardware swap over a qfn repair for sure.

Thank you!

If you can view the system profiler you'll see the USB info and if your keyboard is recognizable, it will be listed under the USB panel, along with the trackpad info. If it shows no keyboard, chances are the keyboard or trackpad is bad. All my hunches are based on the assumption that this is not a water damaged MBP.

 

[Leehs]

Looking through the schematics for the power side of things, it states that G3Hot should be provided with a minimum of 3.31v. That is the exact reading I get on the G3Hot pad, while others seems to be getting 3.42v - I'm wondering if it is possible that my cheap multimeter could be reading 0.1v or 0.2v high thus it seeing less than 3.31v and not allowing the computer to boot at all?

I've just bought a new dc in board to try first off eBay. Then I'll try to figure things out from there with the 820-2330-A board view / schematics that someone kindly put a link to earlier in this thread. I think it was Dadioh, so many thanks to them.

 

[Dadioh]

I'm still wading through this fantastic thread to check for anyone with the same problem as me but I'm only up to page 28.

From my request 5 posts up from here. I'm beginning to wonder if it has a GPU issue causing it to not power up. I know generally you still get the boot chime etc with most gpu issues, but I do know in some cases the machine will not respond to powering up at all. This one has me flummoxed. I would love to breathe life back into this thing. Apple quoted me £380 for a new logic board

Could it even be the dc in board, even if its showing relatively normal voltages? It's just sometimes when I connect the MagSafe adaptor it'll not show the green light but still show voltages on the cd in board. The light will come on if I move the MagSafe plug slightly. This is odd because it is a brand new apple psu that works fine in my older MacBook Pro. The pins on both the psu and dc in are clean and free.

Any help to reduce the grey hair this is causing, would be highly appreciated

I doubt it is the DCin board if you are getting a green led. The behavior with wiggling the connector a bit is not too unusual. Th spring loaded pins may not be making proper contact so a little movement can suddenly engage them.

Re: G3Hot, it usually measures bang on 3.4V for me but it is possible that your meter is in error. Usually G3Hot either works or it doesn't. I can't remember seeing a case where it was putting out a slightly lower voltage. I think that the components that rely on it are probably forgiving of a slightly lower voltage.