SMC Issue on MBP 13 - Hardware Troubleshooting Guide

[cmdrdata]

I've been trying to review the schematic posted by pappabear60, and this is my take on how it is supposed to work. I think all the FETs (p or n-channel, i.e., the NTUD3127 complementary FETs) are turned ON when V gate is > 1.5 volts.

LMK if I made a mistake someplace. Normally a regular transistor NPN or PNP, are turned on with +V or GND respectively, but the spec on 3127 says otherwise, unless I misread it.

 

[Papabear60]

I've been trying to review the schematic posted by pappabear60, and this is my take on how it is supposed to work. I think all the FETs (p or n-channel, i.e., the NTUD3127 complementary FETs) are turned ON when V gate is > 1.5 volts.

LMK if I made a mistake someplace. Normally a regular transistor NPN or PNP, are turned on with +V or GND respectively, but the spec on 3127 says otherwise, unless I misread it.

This makes sense so in this case Q6910 is turned off and SMC cannot talk to the magsafe which is why it doesn't give green light and why it only turns on with SMC debug or bypass mode. The million dollar question is why - I'll dig around some more and recheck resistors, caps but I thought they were all ok and I have replaced Q6910 (which lit up like a christmas tree -initial problem) and U6915 which shorted out.

 

[cmdrdata]

Personally from ibooks and xbox 360/ps3's i am a fan of the heat gun. I have baked a macbook in the oven at 375 and saw no adverse effects on any of the plastic.

Yesterday I took apart my 15" MBP (early 2008), and baked the logic board in my pre-heated to 400°F oven for 8 minutes. Then I turned off the oven and let it cool undisturbed. A small test solder that I placed in a tin foil make shift cup next to it melted to a ball, so it must have gotten hot enough to melt solder.
Today, I put it back together, applying a new, 99% pure heat sink compound ($10 at Radio Shack). The system came back to life, and so far everything worked perfectly. The battery is charging, however, it does have a message "replace soon". I wonder if there is a software tool that would reset the battery count?

 

[jugger18]

Yesterday I took apart my 15" MBP (early 2008), and baked the logic board in my pre-heated to 400°F oven for 8 minutes. Then I turned off the oven and let it cool undisturbed. A small test solder that I placed in a tin foil make shift cup next to it melted to a ball, so it must have gotten hot enough to melt solder.
Today, I put it back together, applying a new, 99% pure heat sink compound ($10 at Radio Shack). The system came back to life, and so far everything worked perfectly. The battery is charging, however, it does have a message "replace soon". I wonder if there is a software tool that would reset the battery count?

Good to hear!

 

[cmdrdata]

This makes sense so in this case Q6910 is turned off and SMC cannot talk to the magsafe which is why it doesn't give green light and why it only turns on with SMC debug or bypass mode. The million dollar question is why - I'll dig around some more and recheck resistors, caps but I thought they were all ok and I have replaced Q6910 (which lit up like a christmas tree -initial problem) and U6915 which shorted out.

Here's what I think: You got 17v on the DCIN rail, then Q6910 (n-channel) Drain is at 0v (meaning it is turned on), thus if R6911 and R6912 are respectively 180kOhm and 470kOhm, the voltage at the junction should be:
Vjunction = 17v x 470k/(470k+180k) = 12.29v, but you showed 1v reading. So either the resistor(s) are bad, or Q6910 (p-channel) is bad (again), perhaps blown/shorted to ground.

 

[Papabear60]

This makes sense so in this case Q6910 is turned off and SMC cannot talk to the magsafe which is why it doesn't give green light and why it only turns on with SMC debug or bypass mode. The million dollar question is why - I'll dig around some more and recheck resistors, caps but I thought they were all ok and I have replaced Q6910 (which lit up like a christmas tree -initial problem) and U6915 which shorted out.

Ok I just replaced Q6910 again as all the other components checked out ok and the voltages now look like you suggested so it looks like I replaced it last time with a dud
Still SMC is not happy as no green light on ac adapter so still only powers on under SMC bypass / debug mode so there must be something else that is upsetting the SMC but at least progress

I am now at the same point as others on this thread where everything appears good once SMC is bypassed but just wont power on normally.
SMC has been reset and checked it is the latest version
ac-adapter and magsafe and battery works with another macbook
Initial issue was dim screen repaired with Dadioh's other thread and not liquid spill so it should work and did before the fat fingers short out the overvoltage protection!
Doesn't respond to the grounding of the keyboard connector trick suggested by peislander but the fact that he was successfull sometimes and then works normally afterwards suggests that the SMC has just got itself in a dead or locked state it couldn't get out of ?
I tried re-flashing the firmware from apple but it stops it from downloading with an error message something like "you are already at the latest version"
There is always the possibility that the SMC chip itself is faulty but I would think that would be rare and after reading earlier threads it suggested the SMC is still reporting temps and battery info to other applications so maybe if the macbook powers up in SMC bypass mode it actually is a sign that the SMC is good?
Can any other diagnosis be done on the SMC? It is a BGA chip so you cant get to the pins to measure states but it is upset over something which is preventing power on so my gut feel is I need to sniff around there to try to detect what it is upset about?
Any other suggestions welcome

 

[cmdrdata]

Ok I just replaced Q6910 again as all the other components checked out ok and the voltages now look like you suggested so it looks like I replaced it last time with a dud
Still SMC is not happy as no green light on ac adapter so still only powers on under SMC bypass / debug mode so there must be something else that is upsetting the SMC but at least progress .....

I need more info. All the voltages that are red need the green version after you replace Q6910 again, and also the voltage at the left side of Q6920 as well as the one between the two FETs.
I am also thinking that Q6915 is blown, resulting in the junction between R6917 and R6918 pulled down to 0.5v It should have floated to about 8.5v if the comparator open collector internal transistor is off.

 

[Papabear60]

I need more info. All the voltages that are red need the green version after you replace Q6910 again, and also the voltage at the left side of Q6920 as well as the one between the two FETs.
I am also thinking that Q6915 is blown, resulting in the junction between R6917 and R6918 pulled down to 0.5v It should have floated to about 8.5v if the comparator open collector internal transistor is off.

I think the penny has just dropped for me ! (sorry British saying that may not translate over here?) The voltages I gave you show that Q6920 is off which is why SMC is not seeing magsafe.
Left of Q6920 pin 3 shows a varing voltage around 0.4v which I am guessing is the magsafe sense signal which I think Dadioh called i2c? Pins 2 & 5 of Q6920 are held at close to 0v.
So if I understand your logic Q6915 should be off and its pin 3 should be floating at 8.5v approx which would turn on Q6920 allowing the magsafe sense signal through.
I'll check Q6915 and Q6920 and report back - Thanks

 

[fznuk]

Hi All,

Someone pointed me to this forum and thread when i to have a water damaged 17" brand new Macbook Pro. Salvaged from a Mac shop.

At first it didnt boot and the corrosion looked pretty bad on the logic board.

After a cleanup i got it to boot when the Keyboard and battery wasnt plugged in, the Magsafe would only go green a certain way around...and the Apple powered up instantly (no keyboard so no power button)

Stripped it, cleaned it again, replaced keyboard and bought another battery (the old original one came up as NOT CHARGING)

Now it boots, works fantastically but for the life of me will not charge any battery....the magsafe doesnt go orange.... the battery test button, when pressed, flashes one green LED a few times then goes off.

I originally thought i got another dead battery so i planned a trip to the apple store (not close by) to test the batteries...but after reading this thread i might be able to test components to at least rule out a dead battery or charge circuit...even a fuse?!

where is it best to start in the test procedure, if you dont mind me asking and not needing to read through all the posted threads - I have test equipment...soldering etc...but no SMD soldering equipment.....but do have a magnifier and tweezers if i need to desolder/replace?! - i hope i dont need to go that far and its a further deep clean?!

Thanks guys in advance!!

I have since tried out the charger and battery in another 17" unibody macbook and found the battery is ok and charges with the mag connector led going orange, so it is definitely the logic board charging circuit.

I have an A1297 17" unibody mac. works perfectly apart from no battery charge

when you press the battery check indicator on the side it flashes one green LED

So, does this narrow the problem to a certain component or components to check?

I thought i'd ask before stripping it down again and possibly ruining it further.

cheers guys!

 

[cmdrdata]

I have since tried out the charger and battery in another 17" unibody macbook and found the battery is ok and charges with the mag connector led going orange, so it is definitely the logic board charging circuit.

I have an A1297 17" unibody mac. works perfectly apart from no battery charge when you press the battery check indicator on the side it flashes one green LED

It seemed to me that you may have a defective component, most likely the power FET that provide charging current to the battery. Apple uses power FET such as RJK0305 or FDS6681 to control the current flow. Google it so you know what it looked like physically. I don't have a schematic for A1297 (would be appreciative if someone has it and can send me one) but with power off and you check continuity of the many power FETs, one of these FET DRAIN or SOURCE pin should be connected to the BATT positive pin.

The diagram shown is an example for papabear60's system. As mentioned I don't access to your system's schematic.

I think the battery communication paths called SCL and SDA signaling are working as the battery test button uses these signals to indicate battery status to the SMC, and the single green LED just says that your batt charge is depleted.

 

[fznuk]

Cheers, Cmdrdata, I really appreciate your detailed help - This week im going to try and check the circuits and components you mentioned...looks like i may have to invest in a SMD rework station but £50 isnt a large sum in comparison to the macbook.

I could have got on ok with the laptop not charging as its mainly plugged in, but now the battery has drained it will not keep time, wireless keys or other passwords on bootup unless connected to a wired lan. So its more annoyance than anything.....now determination has kicked in to finally look into it.

I will get back to you if i find and fix the problem.

Cheers again!

 

[longcoo817]

hey cmdrdata whats your email? i tried to shoot you a pm but i cant seem to find the link on your profile

 

[cmdrdata]

hey cmdrdata whats your email? i tried to shoot you a pm but i cant seem to find the link on your profile

If you click on my userid you should see a "Send a private message to cmdrdata" option.

 

[Papabear60]

I need more info. All the voltages that are red need the green version after you replace Q6910 again, and also the voltage at the left side of Q6920 as well as the one between the two FETs.
I am also thinking that Q6915 is blown, resulting in the junction between R6917 and R6918 pulled down to 0.5v It should have floated to about 8.5v if the comparator open collector internal transistor is off.

After replacing U6915 and Q6915 the ac-adapter turned green and I was able to boot normally several times then disaster struck!
When I installed the battery (it had 10.8V on it) and booted the macbook normally it showed the battery as 0% but not charging and then powered down and now power is totally down on the board!
With the battery back out and just using the ac-adapter both fuses have little voltage present (0.3v) (but both are not blown) and I have tested and re-tested both the ac-adapter and magsafe connector in another macbook which works fine.
G3hot 3.4v is down also.
I suspect the SMC informs the magsafe to turn on the 16-18v from the ac-adapter once it is sensed but with everything down I am lost as to where to start looking - Any suggestions???

Boy! talk about one step forward and three steps back !!!

 

[cmdrdata]

After replacing U6915 and Q6915 the ac-adapter turned green and I was able to boot normally several times then disaster struck!
When I installed the battery (it had 10.8V on it) and booted the macbook normally it showed the battery as 0% but not charging and then powered down and now power is totally down on the board!
With the battery back out and just using the ac-adapter both fuses have little voltage present (0.3v) (but both are not blown) and I have tested and re-tested both the ac-adapter and magsafe connector in another macbook which works fine.
G3hot 3.4v is down also.
I suspect the SMC informs the magsafe to turn on the 16-18v from the ac-adapter once it is sensed but with everything down I am lost as to where to start looking - Any suggestions???

Boy! talk about one step forward and three steps back !!!

Wow, a few steps backward all right. Here are some other ideas:
- The 0.3v measured on both fuses (magsafe only, without the battery) indicates power is not being supplied by the magsafe. Magsafe is designed to "see" a load before output is enabled. A load means that the SMC/CHARGER IC know the magsafe is present and enable the power FETs below. However, since G3HOT is off, the SMC and Charger IC has no Vcc (chip power), thus it is a catch-22. I THINK the normal sequence is this: magsafe connected, 18.5v goes through the system (briefly) and G3HOT active thus SMC is powered up, SMC checks system state and if all parameters OK then enables the power FETs to allow system power to come through. If SMC fails to do this, after a few seconds the magsafe drops the 18.5v and thus G3HOT also goes away. I'd love to see if someone (dadioh?) can verify this using their spare/play logicboard that has the same symptom by hooking up an external load (low value resistor) to the input fuse and then try connecting the magsafe. I suspect that the load will keep the magsafe up and thus the G3HOT. It is also possible that the downstream component may burn up if 18.5v is not automatically stopped. :-( so there is risk in trying this test.

- Check Q7000, Q7001, Q7020, Q7021 where the signal PP18V5_G3H_CHGR goes through to provide system power. This is the path in series with the input fuse. If all these are OK, I'd think that the system power should be OK. Next check the charging path.

- The charging path is R7008, Q7050, Q7052. These could be damaged causing the entire system to not working when the battery is connected. If the problem is U7000, you're in trouble. That is an Apple proprietary part.

- Alternately, I'd be curious on battery only powering up. Again through the 2 diodes OR'ed together, battery power can also generate G3HOT so the SMC can get Vcc and do the same thing as above. BTW, the battery pack full charge voltage should be 4.22v x 3 = 12.66v. 10.8 is very near the depleted state.

- Have you tried the SMC bypass method after this episode without the battery installed and see what happened? also what happened when you try just battery power WITHOUT MAGSAFE?

 

[Papabear60]

Wow, a few steps backward all right. Here are some other ideas:
- The 0.3v measured on both fuses (magsafe only, without the battery) indicates power is not being supplied by the magsafe. Magsafe is designed to "see" a load before output is enabled. A load means that the SMC/CHARGER IC know the magsafe is present and enable the power FETs below. However, since G3HOT is off, the SMC and Charger IC has no Vcc (chip power), thus it is a catch-22. I THINK the normal sequence is this: magsafe connected, 18.5v goes through the system (briefly) and G3HOT active thus SMC is powered up, SMC checks system state and if all parameters OK then enables the power FETs to allow system power to come through. If SMC fails to do this, after a few seconds the magsafe drops the 18.5v and thus G3HOT also goes away. I'd love to see if someone (dadioh?) can verify this using their spare/play logicboard that has the same symptom by hooking up an external load (low value resistor) to the input fuse and then try connecting the magsafe. I suspect that the load will keep the magsafe up and thus the G3HOT. It is also possible that the downstream component may burn up if 18.5v is not automatically stopped. :-( so there is risk in trying this test.

- Check Q7000, Q7001, Q7020, Q7021 where the signal PP18V5_G3H_CHGR goes through to provide system power. This is the path in series with the input fuse. If all these are OK, I'd think that the system power should be OK. Next check the charging path.

- The charging path is R7008, Q7050, Q7052. These could be damaged causing the entire system to not working when the battery is connected. If the problem is U7000, you're in trouble. That is an Apple proprietary part.

- Alternately, I'd be curious on battery only powering up. Again through the 2 diodes OR'ed together, battery power can also generate G3HOT so the SMC can get Vcc and do the same thing as above. BTW, the battery pack full charge voltage should be 4.22v x 3 = 12.66v. 10.8 is very near the depleted state.

- Have you tried the SMC bypass method after this episode without the battery installed and see what happened? also what happened when you try just battery power WITHOUT MAGSAFE?

I tried another battery last night and g3hot does come up and the battery fuse shows 11.8V (battery voltage ) so I'll look at the components suggested but you are correct in this "catch 22" because I am puzzled how anything can sense and instruct the adapter to go to full voltage as without the 17-18.5 volts the sense circuit is dead anyway??? Maybe you are correct in that the full voltage is briefly present to check and then drops when it senses trouble.
If that is true then it makes trouble shooting tough as the condition causing the problem may not be detectable with the DC-IN turned off?

 

[cmdrdata]

I tried another battery last night and g3hot does come up and the battery fuse shows 11.8V (battery voltage ) so I'll look at the components suggested but you are correct in this "catch 22" because I am puzzled how anything can sense and instruct the adapter to go to full voltage as without the 17-18.5 volts the sense circuit is dead anyway??? Maybe you are correct in that the full voltage is briefly present to check and then drops when it senses trouble.
If that is true then it makes trouble shooting tough as the condition causing the problem may not be detectable with the DC-IN turned off?

Actually, what you just described is a somewhat encouraging. Seeing 11.8v on the second fuse (by the battery) connector tells me that R7008, Q7050 and QA7052 is allowing battery power to go to the system. The power rail sourced by battery is intact). Perhaps the "gate signals" to these FETs are bad, OR the FET(s) DRAIN and SOURCE are shorted allowing battery power to go through. Can you verify?

What about the bypass mode? Were you able to try it again? Is the magsafe LED still showing green when plugged in?

 

[Papabear60]

Actually, what you just described is a somewhat encouraging. Seeing 11.8v on the second fuse (by the battery) connector tells me that R7008, Q7050 and QA7052 is allowing battery power to go to the system. The power rail sourced by battery is intact). Perhaps the "gate signals" to these FETs are bad, OR the FET(s) DRAIN and SOURCE are shorted allowing battery power to go through. Can you verify?

What about the bypass mode? Were you able to try it again? Is the magsafe LED still showing green when plugged in?

Bypass mode doesn't work and magsafe was not green as the 17v rail in the overvoltage circuit was missing. I think I have done something stupid again (fat fingers) while poking around trying to solve the battery charging issue and blew up Q6910 again! That sucker is so tiny but it seems like the first place to look as at least it does its job and blows whenever a short condition is presented.
I'll replace it again on Monday and see if it gets me back to where we were as I checked the gate diodes and FETs which seem all good but all of that is after this overvoltage protection circuit so I will report back

 

[Dadioh]

Bypass mode doesn't work and magsafe was not green as the 17v rail in the overvoltage circuit was missing. I think I have done something stupid again (fat fingers) while poking around trying to solve the battery charging issue and blew up Q6910 again! That sucker is so tiny but it seems like the first place to look as at least it does its job and blows whenever a short condition is presented.
I'll replace it again on Monday and see if it gets me back to where we were as I checked the gate diodes and FETs which seem all good but all of that is after this overvoltage protection circuit so I will report back

Is the issue with the probes being too large? Regular multimeter probes are far too big for the size of devices we are working with here.

Here are the probes that I use. They are very very fine and work really well.

http://www.newark.com/pomona/6342/test-probe-set/dp/91B0558?Ntt=91B0558

 

[jeydax]

Forgive me if this is the wrong spot to post this but I need some help.

My friend's MacBook Pro 15" (aluminum case, 2006, ATI Graphics) was having issues booting up, and I narrowed it down to the harddrive crapping out on him. He ordered a new harddrive, last night I went to take it apart and throw the new drive in. I've fixed thousands of PC desktops and laptops (tech support at a Big 10 school for 4 years... never. again.) but this was my first foray into taking apart and tinkering with a Mac's insides.

I followed a guide from ifixit, which was simple enough, I didn't want to go all Rambo on a machine I've never played with before. I installed the new drive, but did not put all the screws/keyboard/etc back into place, I wanted to make sure everything worked before securing everything in place.

I tried to fire up the computer and the screen will not turn on. I checked the cables to the monitor, they seemed to be fine. When I push the power button I hear the DVD drive start up, the fans run, the light on the front of the computer turns on. I do not get a chime, the apple logo on the back does not start up, but the light on the front of the computer does turn on. I've tried resetting nvram -- I never get a chime no matter how long I hold the buttons, I've tried the trick at the start of this thread, I've tried pulling the battery and holding the power down for 5 seconds/resetting smc.

I have read a few threads saying take it in to Mac, but I'd like to see if there is anything else I can try before doing that. I've broken plenty of my own computers by tinkering with them, but never a friends/clients... I'd like to keep that record.

TLDR: Replaced harddrive in MBP-15" (2006) and now the monitor won't turn on.

 

[Dadioh]

Forgive me if this is the wrong spot to post this but I need some help.

My friend's MacBook Pro 15" (aluminum case, 2006, ATI Graphics) was having issues booting up, and I narrowed it down to the harddrive crapping out on him. He ordered a new harddrive, last night I went to take it apart and throw the new drive in. I've fixed thousands of PC desktops and laptops (tech support at a Big 10 school for 4 years... never. again.) but this was my first foray into taking apart and tinkering with a Mac's insides.

I followed a guide from ifixit, which was simple enough, I didn't want to go all Rambo on a machine I've never played with before. I installed the new drive, but did not put all the screws/keyboard/etc back into place, I wanted to make sure everything worked before securing everything in place.

I tried to fire up the computer and the screen will not turn on. I checked the cables to the monitor, they seemed to be fine. When I push the power button I hear the DVD drive start up, the fans run, the light on the front of the computer turns on. I do not get a chime, the apple logo on the back does not start up, but the light on the front of the computer does turn on. I've tried resetting nvram -- I never get a chime no matter how long I hold the buttons, I've tried the trick at the start of this thread, I've tried pulling the battery and holding the power down for 5 seconds/resetting smc.

I have read a few threads saying take it in to Mac, but I'd like to see if there is anything else I can try before doing that. I've broken plenty of my own computers by tinkering with them, but never a friends/clients... I'd like to keep that record.

TLDR: Replaced harddrive in MBP-15" (2006) and now the monitor won't turn on.

Not getting a chime means it is not passing self diagnostics. The monitor and backlight are probably fine but the Mac just doesn't get far enough to turn them on. I assume it was chiming alright before and the issue was just related to not booting from the hard drive properly? If so, then something may have been jostled loose. Check all the cable connections and reseat the memory. Although, if it were bad memory you should get beeps (not chime) that tell you so.

The light on the front is the sleep led and normally it comes on for a few seconds and then fades when the Mac chimes and boots.

Try the smc reset (holding power button down with both battery and charger disconnected). Pram reset only happens after chime so that shouldn't be the issue.

 

[jeydax]

Not getting a chime means it is not passing self diagnostics. The monitor and backlight are probably fine but the Mac just doesn't get far enough to turn them on. I assume it was chiming alright before and the issue was just related to not booting from the hard drive properly? If so, then something may have been jostled loose. Check all the cable connections and reseat the memory. Although, if it were bad memory you should get beeps (not chime) that tell you so.

The light on the front is the sleep led and normally it comes on for a few seconds and then fades when the Mac chimes and boots.

Try the smc reset (holding power button down with both battery and charger disconnected). Pram reset only happens after chime so that shouldn't be the issue.

I'm doing a major facepalm right now. Literally fixed thousands of computers and I forget the RAM.

That fixed it, you are awesome, thank you. <3<3<3<3

 

[cmdrdata]

I'm doing a major facepalm right now. Literally fixed thousands of computers and I forget the RAM.
That fixed it, you are awesome, thank you. <3<3<3<3

Taking things apart, putting it back together, then realizing you left out this one little part is aggravating and one that I do quite often. In modern day electronics, the parts are so tiny and often times I break something else along the way of repeating the process. One that comes to mind is "Connector #3" in an iPhone 3GS that I worked on a while back. That thing is so tiny and fragile I broke the flap that clamps the ribbon cable.

 

[Papabear60]

Is the issue with the probes being too large? Regular multimeter probes are far too big for the size of devices we are working with here.

Here are the probes that I use. They are very very fine and work really well.

http://www.newark.com/pomona/6342/test-probe-set/dp/91B0558?Ntt=91B0558

Thanks for the tip (No pun intended )
Well it looks like Q6910 does its job as I have blown this twice now but got completely different start up symptoms so this overvoltage protection circuit is the first place to look I think.
The good news is once I replaced it the ac adapter went green and I can start the Macbook normally once again. The only reaining issue is that it does not charge the battery.
I charged the battery up in another Macbook (to over 12v) but still my Macbook died instantly the ac-adapter is pulled out.
I have tried another known good battery and the battery fuse does show 12v and g3hot is 3.4v but it wont power on under battery so I'm thinking the battery diodes are good otherwise g3hot wouldn't be there so I'll check the battery FETs with those newly ordered probes Dadioh suggested so I only move forward from now on .

 

[cmdrdata]

Thanks for the tip (No pun intended )
The good news is once I replaced it the ac adapter went green and I can start the Macbook normally once again. The only reaining issue is that it does not charge the battery.
I charged the battery up in another Macbook (to over 12v) but still my Macbook died instantly the ac-adapter is pulled out.
I have tried another known good battery and the battery fuse does show 12v and g3hot is 3.4v but it wont power on under battery so I'm thinking the battery diodes are good otherwise g3hot wouldn't be there so I'll check the battery FETs with those newly ordered probes Dadioh suggested so I only move forward from now on .

The fact that you get 12v on the "output" power bus and G3HOT is also good meant that now your power lines are functional. So next is signals related info that the CPU and SMC needs to have to keep system power from battery is correct. There are two things to look for: BMON CURRENT SENSE and PWRGOOD when the battery is the only one supplying power to the system. I'd track this down and see how these are generated. I suspect one of these is not generating the expected signal. Does the battery charge button work, i.e., gives you battery charge status LEDs? If so, we can assume that the SMC is talking to the battery OK (SCL/SDA signaling to the battery is good.). There is a battery sense logic on p44 (BMON). I think this may be the one that is causing the CPU to shut down immediately when trying to power the system on battery.