SMC Issue on MBP 13 - Hardware Troubleshooting Guide

Dadioh · Jun 10, 2011

cmdrdata said:
I picked up an MBP (mid 2009 unibody) recently and it is totally dead. I tried the SMC bypass trick, but no response at all. No chime, battery state on the side of the MBP shows no charge remains and no charging is going on. The magsafe adaptter is good (worked on my lat 2008 white MacBook), so the problem is definitely on the MBP. When the magsafe is connected, the orange/green LED on the magsafe plug doesn't even glow. The person that asked me to look into this swore that it has never been exposed to any liquid. Taking the bottom cover apart showed that the inside looked pristine and has never been disassembled. ANy suggestions on what to do next is welcome.

Previously I have had experience with a a liquid damaged MB and cleaning it up and letting it dry thoroughly brought it back to life with th eexception of a bad superdrive. I was lucky then. TIA>

Here are a few quick checks.

Down near the battery connection there is a white bodied component that is the fuse that protects the main 12V supply to the board. If you (carefully) measure the voltage on both ends of the fuse then you should see a little over 12V. If not, check that fuse.

If you do not have power at the input to that fuse then the other fuse is on the underside of the board near the magsafe. That is the main fuse on the 18.5V input from the magsafe.

If that fuse doesn't have 18.5V at its input then you probably have a bad magsafe board. Give it a visual inspection for damage.

cmdrdata · Jun 10, 2011

Dadioh said:
Here are a few quick checks.

Down near the battery connection there is a white bodied component that is the fuse that protects the main 12V supply to the board. If you (carefully) measure the voltage on both ends of the fuse then you should see a little over 12V. If not, check that fuse.

If you do not have power at the input to that fuse then the other fuse is on the underside of the board near the magsafe. That is the main fuse on the 18.5V input from the magsafe.

If that fuse doesn't have 18.5V at its input then you probably have a bad magsafe board. Give it a visual inspection for damage.

Fuse near the battery connector measured 0 ohm. So that fuse is OK, and there is no 12vdc power measured in either side. To get to the other one, I will have to tear down the logic board, so that will probably have to wait till later this evening. Measuring the 5 small solder spots (they are in a row on the magsafe board visible from the top (well actually the bottom of the unit with the cover off), I saw this:

starting from the spot nearest the back of the MBP:

0v 0.34v 0v 0.34v 0v

On the magsafe itself, disconected from the MBP (open ended plug, I measured this (with respect to -dc):
-dc +dc sense +dc -
0v 6.63v 0v 6.63v 0v

@

cmdrdata said:
Fuse near the battery connector measured 0 ohm. So that fuse is OK, and there is no 12vdc power measured in either side. To get to the other one, I will have to tear down the logic board, so that will probably have to wait till later this evening. Measuring the 5 small solder spots (they are in a row on the magsafe board visible from the top (well actually the bottom of the unit with the cover off), I saw this:

starting from the spot nearest the back of the MBP:

0v 0.34v 0v 0.34v 0v

On the magsafe itself, disconected from the MBP (open ended plug, I measured this (with respect to -dc):
-dc +dc sense +dc -
0v 6.63v 0v 6.63v 0v

@Dadioh: Since my battery is totally discharged, I wonder if I jury-rigged a 16-17 vdc power to the -dc and +dc pin of the pins I shown above (with magsafe plug disconnected, would it charge the battery. I am thinking that if I can get some charge into the battery and the MBP boots up, then I would feel a little better that the magsafe board is the bad guy. what do you think?

Dadioh · Jun 10, 2011

cmdrdata said:
@

@Dadioh: Since my battery is totally discharged, I wonder if I jury-rigged a 16-17 vdc power to the -dc and +dc pin of the pins I shown above (with magsafe plug disconnected, would it charge the battery. I am thinking that if I can get some charge into the battery and the MBP boots up, then I would feel a little better that the magsafe board is the bad guy. what do you think?

I would be very nervous about bypassing things like that. The Magsafe is an intelligent charger and gets information from the SMC on whether to deliver current or not. The battery "is" removeable with a tri spoke screwdriver if you had a buddy who would let you swap it into their macbook to do a charge

The fact that you do not have 12V at that fuse with the magsafe on could mean it is the magsafe board. They are not too expensive so that may be a good scenario for you.

I am going to have a look for those solder pads you mentioned. I never noticed them Be right back....

Dadioh · Jun 10, 2011

cmdrdata said:
Fuse near the battery connector measured 0 ohm. So that fuse is OK, and there is no 12vdc power measured in either side. To get to the other one, I will have to tear down the logic board, so that will probably have to wait till later this evening. Measuring the 5 small solder spots (they are in a row on the magsafe board visible from the top (well actually the bottom of the unit with the cover off), I saw this:

starting from the spot nearest the back of the MBP:

0v 0.34v 0v 0.34v 0v

On the magsafe itself, disconected from the MBP (open ended plug, I measured this (with respect to -dc):
-dc +dc sense +dc -
0v 6.63v 0v 6.63v 0v

I am not sure what solder spots you are measuring. However, look at my attached image. You should be able to read 18.5V at the point noted. If you do not have 18.5V there then I think your magsafe board needs to be replaced (or cleaned if it is liquid damaged).

cmdrdata · Jun 11, 2011

Dadioh said:
I am not sure what solder spots you are measuring. However, look at my attached image. You should be able to read 18.5V at the point noted. If you do not have 18.5V there then I think your magsafe board needs to be replaced (or cleaned if it is liquid damaged).

I didn't get to remove the logic board last night, so with everything still on the MBP except the bottom cover, and the magsafe adapter connceted:
The five spots in a row I am refering to are the visible solder spots on the small magsafe DC board itself. Since the battery is totally depleted, and pin 2 and 4 of the DC board are only showing 0.34 vdc with respect to 1 and 5, I am not getting the 18.5 vdc you mentioned at the point indicated. I think that the magsafe center pin is not "sensing" the correct value, whatever that is to turn on the 18.5v you mentioned. BTW, with the battery unplugged, the system still will not turn on the the magsafe adapter orange/green LED.

I do have a spare magsafe/DC in board that I board from eBay in 2009 for a MB (not MBP), and may try that later on the MBP.

Is it safe to power up the logic board via the DC-in-board and magsafe after
you remove it from the chassis with nothing else attached?

Dadioh · Jun 11, 2011

cmdrdata said:
I didn't get to remove the logic board last night, so with everything still on the MBP except the bottom cover, and the magsafe adapter connceted:
The five spots in a row I am refering to are the visible solder spots on the small magsafe DC board itself. Since the battery is totally depleted, and pin 2 and 4 of the DC board are only showing 0.34 vdc with respect to 1 and 5, I am not getting the 18.5 vdc you mentioned at the point indicated. I think that the magsafe center pin is not "sensing" the correct value, whatever that is to turn on the 18.5v you mentioned. BTW, with the battery unplugged, the system still will not turn on the the magsafe adapter orange/green LED.

I think you are correct. That center pin is a bi-directional serial communications channel to the SMC. If the SMC is unhappy about anything it will not tell the charger to enable full charging current. I think the default from the charger is a slow trickle charge that takes days to fully charge a battery.

cmdrdata said:
I do have a spare magsafe/DC in board that I board from eBay in 2009 for a MB (not MBP), and may try that later on the MBP.

It has the same pinout so you can use it to troubleshoot if the issue is the magsafe circuit. However, it will not physically fit in the Macbook so you can not replace your existing magsafe. I have a spare one that I use in this manner for testing logic boards out of their cases. BTW, a lesson that I learned the hard way... Use some Kapton tape (or electrical tape) to tape up the exposed magsafe board or you can accidently touch it to the card and fry the card (speaking from experience

Now I just keep my spare magsafe wrapped in Kapton and don't need to worry about it.

cmdrdata said:
Is it safe to power up the logic board via the DC-in-board and magsafe after
you remove it from the chassis with nothing else attached?

I do this all the time when working on logic boards. Make sure you have a non-conductive, clean surface to lay it on. Also, if you think the logic board will actual power up then make sure the heatsink is attached. Additionally, if you may be running it for mote than a minute or 2 place the fan in the heatsink and mount it with the 1 screw that goes into the logic board. Depending on which side of the board you are working on you can prop it up a bit to get airflow into the fan intake.

To power up the board without the keyboard there are some pins you can short out. I think I documented those earlier in this thread.

cmdrdata · Jun 12, 2011

cmdrdata said:
Fuse near the battery connector measured 0 ohm. So that fuse is OK, and there is no 12vdc power measured in either side. To get to the other one, I will have to tear down the logic board, so that will probably have to wait till later this evening. Measuring the 5 small solder spots (they are in a row on the magsafe board visible from the top (well actually the bottom of the unit with the cover off), I saw this:

starting from the spot nearest the back of the MBP:

0v 0.34v 0v 0.34v 0v

On the magsafe itself, disconected from the MBP (open ended plug, I measured this (with respect to -dc):
-dc +dc sense +dc -
0v 6.63v 0v 6.63v 0v

@Dadioh: can you post a picture of the underside of the logic board of the FUSE you mentioned? The one visible under the bottom cover is ok, so before taking my logic boardoff, I'd like to know that mine do have a second fuse on the inner side. Thanks.

Dadioh · Jun 12, 2011

cmdrdata said:
@Dadioh: can you post a picture of the underside of the logic board of the FUSE you mentioned? The one visible under the bottom cover is ok, so before taking my logic boardoff, I'd like to know that mine do have a second fuse on the inner side. Thanks.

I am not near my work bench now. The fuse is easy to spot. It has a white body and it is near the MagSafe.

cmdrdata · Jun 12, 2011

Dadioh said:
I am not near my work bench now. The fuse is easy to spot. It has a white body and it is near the MagSafe.

NO hurry, just whenever you get a chance. Currently I have the internal battery taken off the MBP and am trying to charge it off line using a 16.5 vdc to the left and right hand pins of the battery plug. This plug has about 9 wires/pins connected to it, and it looked like the left most 3 pins (looking from the top of the plug)is GND, and the 3 right most pins are the +10.9 vdc. I suspect the middle pins are sense pins and perhaps temperature status.

Dadioh · Jun 12, 2011

cmdrdata said:
NO hurry, just whenever you get a chance. Currently I have the internal battery taken off the MBP and am trying to charge it off line using a 16.5 vdc to the left and right hand pins of the battery plug. This plug has about 9 wires/pins connected to it, and it looked like the left most 3 pins (looking from the top of the plug)is GND, and the 3 right most pins are the +10.9 vdc. I suspect the middle pins are sense pins and perhaps temperature status.

Here is a picture of the main input fuse on the back. It is a 6 Amp 24V fuse.

The battery connector pinout is:

Pin Signal
1 +12V
2 +12V
3 +12V
4 SMBUS_SMC_BSA_SCL
5 SYS_DETECT_L
6 SMBUS_SMC_BSA_SDA
7 GND
8 GND
9 GND

Be careful to current limit the source you are charging the battery with. If you give it too much current you will overheat and damage the cells in the battery. Maybe google searching on battery charging for some guidelines.

cmdrdata · Jun 12, 2011

Dadioh said:
Here is a picture of the main input fuse on the back. It is a 6 Amp 24V fuse.

The battery connector pinout is:

Pin Signal
1 +12V
2 +12V
3 +12V
4 SMBUS_SMC_BSA_SCL
5 SYS_DETECT_L
6 SMBUS_SMC_BSA_SDA
7 GND
8 GND
9 GND

Be careful to current limit the source you are charging the battery with. If you give it too much current you will overheat and damage the cells in the battery. Maybe google searching on battery charging for some guidelines.

Thanks for the warning and battery terminal info. Yes, I'm familiar with the Li-Ion charging requirements. My plan is to just get it off from the 0v reading and up to some nominal pack voltage (10-11v or so), but it seemed that after hooking it up for about 1 hour with a voltemeter connected (meter reading 16.4v which seemed to indicate to me the cells are not loading/charging). Measuring the +/- pins after disconnecting the charger, I only got 0.26v (previously 0.0 v). So I am wondering if the battery went bad, took the SMC out or "underneath fuse" and thus the no charge/magsafe issue.

Do you know what signal the magsafe adapter center pin 3 expect to see (ground, dc-voltage level, or pulse signal?) to make the magsafe adapter think it is connected to a good system?

Dadioh · Jun 12, 2011

cmdrdata said:
Do you know what signal the magsafe adapter center pin 3 expect to see (ground, dc-voltage level, or pulse signal?) to make the magsafe adapter think it is connected to a good system?

It is not a simple dc signal. It is an i2c bus that communicates with the smc.

http://en.m.wikipedia.org/wiki/I²C

Sounds like your battery is dead. :-(

cmdrdata · Jun 12, 2011

Dadioh said:
It is not a simple dc signal. It is an i2c bus that communicates with the smc.

http://en.m.wikipedia.org/wiki/I²C

Sounds like your battery is dead. :-(

Ordered another battery from eBay (type A1322, APN 661-5557 (the newer version of this battery, capacity 63.5W). I'll wait till I get a good known battery before taking apart my logic board from the system. I also replace the tri-wing screw with regular philips head for now. While looking at the logic board, just next to the video cable I saw another white block that looked loked the other fuse by the battery plug. So for girns I measure the resistance accross it, and it was OPEN. Is this another fuse that you haven't mentioned? See the attached picture.

Dadioh said:
Here is a picture of the main input fuse on the back. It is a 6 Amp 24V fuse.

The battery connector pinout is:

Pin Signal
1 +12V
2 +12V
3 +12V
4 SMBUS_SMC_BSA_SCL
5 SYS_DETECT_L
6 SMBUS_SMC_BSA_SDA
7 GND
8 GND
9 GND

Be careful to current limit the source you are charging the battery with. If you give it too much current you will overheat and damage the cells in the battery. Maybe google searching on battery charging for some guidelines.

Off topic: how do you get such a clear focused picture like this one? The one I just attached was taken with my iPhone, and I cannot get close or such a sharp focus.

Dadioh · Jun 12, 2011

cmdrdata said:
Off topic: how do you get such a clear focused picture like this one? The one I just attached was taken with my iPhone, and I cannot get close or such a sharp focus.

I have a stereo microscope that I use for fine soldering. I just hold my iPhone up to one of the eye pieces and snap a pic. Works great

There is also an app for the iPhone called magnifying glass. There is a free version and a paid version. It might help you do clear closeups. I haven't used it that way but might be worth a shot (and a lot cheaper than a microscope

cmdrdata said:
Ordered another battery from eBay (type A1322, APN 661-5557 (the newer version of this battery, capacity 63.5W). I'll wait till I get a good known battery before taking apart my logic board from the system. I also replace the tri-wing screw with regular philips head for now. While looking at the logic board, just next to the video cable I saw another white block that looked loked the other fuse by the battery plug. So for girns I measure the resistance accross it, and it was OPEN. Is this another fuse that you haven't mentioned? See the attached picture.

That white part is a high frequency capacitor for the led backlight boost circuit. It should measure open circuit. I have another thread on here about backlight driver failures that gives a lot of info about repairing backlight failures. Just do a search on my posts.

cmdrdata · Jun 14, 2011

Dadioh said:
Here is a picture of the main input fuse on the back. It is a 6 Amp 24V fuse.

The battery connector pinout is:

Pin Signal
1 +12V
2 +12V
3 +12V
4 SMBUS_SMC_BSA_SCL
5 SYS_DETECT_L
6 SMBUS_SMC_BSA_SDA
7 GND
8 GND
9 GND

Be careful to current limit the source you are charging the battery with. If you give it too much current you will overheat and damage the cells in the battery. Maybe google searching on battery charging for some guidelines.

@Dadioh: In the 15" MBP schematic that was posted in this forum, the input power fuse is the one you mentioned above. In that schematic, U5705/D6905/Q6910/R6905 and C6905 are all connected to right side of the fuse (the left side is to the magsafe +DCIN. Is any of this component on the accessible side of the 13.3" MBP logic board with the bottom cover off? If so, can you tell me which ones they are? I'm itching to check that fuse without taking things apart yet until I received my "new/used" battery pack to arrive soon. Thanks.

cmdrdata · Jun 15, 2011

jpeezy said:
Hi,

I'm having the same exact issue on a 15" 2010 as cmdrdata, except my magesafe LED is green when plugged in and the battery indicators only work when its plugged in. The computer was working fine at 7AM one morning, it went to sleep and an hour later it wouldn't power back on. I tried the SMC reset as described but it didn't work. I can only assume at this point that the logic board needs to be replaced.

Sadly, its 1 month out of warranty

Thanks

@jpeezy: I've been working on my own MBP 13.3" and I am wondering if the no-start in your MBP is due to a bad battery. With the magsafe disconnected, when you press the button for the battery status at the left side of the MBP, does ANY of the LED status indicating battery charge light up? If no light at all, perhaps your battery is totally dead like mine (0 volt at the battery pack terminal). I am thinking that maybe mybe the fact that the battery went to zero volt caused the charger circuitry to pass excessive current to the dead battery and thus zap the input (DCIN fuse). I am about to take apart my MBP to verify this soon. But before I do that, I've just ordered from ebay a known good battery to ensure that the logic board can start from a battery by itself. Stay tuned.

jpeezy · Jun 15, 2011

I was thinking the same thing. Ill keep an eye on this thread.

Thanks

cmdrdata · Jun 15, 2011

jpeezy said:
I was thinking the same thing. Ill keep an eye on this thread.

Thanks

@jpeezy: there might be one simple thing to try. Since you actually had the green light on the magsafe plug (mine is totally dark although I know the magsfee adapter worked in my white MB), Take the bottom cover off, and disconnect the battery from the MBP logic board. Then try to power up the system with just the magsafe.

jpeezy · Jun 15, 2011

I tried that already and it didn't work

With the battery out, the first light flash rapidly when I press the battery indicator.

Also, I noticed that when I try the reset the mag safe LED doesn't come on and the battery indicators do not work. When I pull the power and plug it back in the indicators work fine along with the LED on the mag safe. So I assume the SMC is working fine.

cmdrdata · Jun 15, 2011

cmdrdata said:
@jpeezy: there might be one simple thing to try. Since you actually had the green light on the magsafe plug (mine is totally dark although I know the magsfee adapter worked in my white MB), Take the bottom cover off, and disconnect the battery from the MBP logic board. Then try to power up the system with just the magsafe.

@dadi_oh: If you're reading this, I have another related question. In the unibody MBP with "non-removable" battery, the system has a "wired OR" power source select (D6905) implemented with wo diodes with the cathodes connected together. If the battery is at nominal voltage, the G3HOT supply will indeed be always hot and provide the expected voltage for system startup, including providing power to SMC to communicate with the magsafe via the magsafe sense line. However, if the battery (like mine) is at 0 volt, the G3HOT is no longer hot when the the magsafe is attached to the MBP. It seemed like a deficient design for the magsafe to expect signaling from the SMC when the G3HOT is not on. I wonder if in the older MB/MBP design where people are reporting that they can start from the adapter with the battery removed the "G3HOT" logic 9if it exist at all) is somewhat different? Recall that my magsafe open circuit voltage is 6.8 v, but drops down to nothing when attached to the MBP.

jpeezy · Jun 15, 2011

cmdrdata said:
@dadi_oh: If you're reading this, I have another related question. In the unibody MBP with "non-removable" battery, the system has a "wired OR" power source select (D6905) implemented with wo diodes with the cathodes connected together. If the battery is at nominal voltage, the G3HOT supply will indeed be always hot and provide the expected voltage for system startup, including providing power to SMC to communicate with the magsafe via the magsafe sense line. However, if the battery (like mine) is at 0 volt, the G3HOT is no longer hot when the the magsafe is attached to the MBP. It seemed like a deficient design for the magsafe to expect signaling from the SMC when the G3HOT is not on. I wonder if in the older MB/MBP design where people are reporting that they can start from the adapter with the battery removed the "G3HOT" logic 9if it exist at all) is somewhat different? Recall that my magsafe open circuit voltage is 6.8 v, but drops down to nothing when attached to the MBP.

I had a 2008 that I could use without the battery installed. I was wondering the same thing about the newer models. Hoping they won't turn on without the battery installed.

cmdrdata · Jun 15, 2011

jpeezy said:
I tried that already and it didn't work With the battery out, the first light flash rapidly when I press the battery indicator.

Also, I noticed that when I try the reset the mag safe LED doesn't come on and the battery indicators do not work. When I pull the power and plug it back in the indicators work fine along with the LED on the mag safe. So I assume the SMC is working fine.

The "new/used" battery from eBay arrived today, so I proceeded to install it in my MBP. The battery check indicator did not light up when I press the check button. Bad news. I went ahead and connect the magsafe. Still no green or orange on the magsafe plug. When I press the power on button, nothing happened. So I took the battery out, measured the voltage between pins 1/2/3 (+vdc) and pins 7/8/9 (-vdc) and I got 0 volt. Another dead battery?

cmdrdata said:
The "new/used" battery from eBay arrived today, so I proceeded to install it in my MBP. The battery check indicator did not light up when I press the check button. Bad news. I went ahead and connect the magsafe. Still no green or orange on the magsafe plug. When I press the power on button, nothing happened. So I took the battery out, measured the voltage between pins 1/2/3 (+vdc) and pins 7/8/9 (-vdc) and I got 0 volt. Another dead battery?

BTW, I carefully and painstakingly cut the welded seam on the first "dead" battery. Inside that pack are 6 square batteries connected in series pairs. The first pair measured 1.40v, the second 3.16v, and the third pair 2.98v. Based on this I conclude that the charging/protection circuitry (a small 0.3"x1" board where the battery connector is attached is somehow disabling the battery from enabling the output voltage, or there is a fuse in there some place (it is potted with black stuff).

Kharlitro · Jun 16, 2011

Hi, i managed to get isopropanol (99.8% sigma-aldrich reagent grade) and this evening i'll try to clean first the dc-in board, and then the logic.
Do you have some tips ?

Dadioh · Jun 16, 2011

cmdrdata said:
The "new/used" battery from eBay arrived today, so I proceeded to install it in my MBP. The battery check indicator did not light up when I press the check button. Bad news. I went ahead and connect the magsafe. Still no green or orange on the magsafe plug. When I press the power on button, nothing happened. So I took the battery out, measured the voltage between pins 1/2/3 (+vdc) and pins 7/8/9 (-vdc) and I got 0 volt. Another dead battery?

Dead battery... bummer...

If both the battery and the magsafe are not providing power the G3Hot is off and not sure you can go any further.

Kharlitro said:
Hi, i managed to get isopropanol (99.8% sigma-aldrich reagent grade) and this evening i'll try to clean first the dc-in board, and then the logic.
Do you have some tips ?

Buy a brand new firm bristle toothbrush. Pour some alcohol into a small dish and then dip the toothbrush and scrub liberally in the affected areas. Keep applying new alcohol. Use random back and forth and circular motion to let the bristles get into every nook and cranny. Let any excess alcohol drip off (hopefully taking impurities off the board and not allowing them to re-settle on the board). Let the part dry for at least 30 minutes before applying power.

cmdrdata said:
@dadi_oh: If you're reading this, I have another related question. In the unibody MBP with "non-removable" battery, the system has a "wired OR" power source select (D6905) implemented with wo diodes with the cathodes connected together. If the battery is at nominal voltage, the G3HOT supply will indeed be always hot and provide the expected voltage for system startup, including providing power to SMC to communicate with the magsafe via the magsafe sense line. However, if the battery (like mine) is at 0 volt, the G3HOT is no longer hot when the the magsafe is attached to the MBP. It seemed like a deficient design for the magsafe to expect signaling from the SMC when the G3HOT is not on. I wonder if in the older MB/MBP design where people are reporting that they can start from the adapter with the battery removed the "G3HOT" logic 9if it exist at all) is somewhat different? Recall that my magsafe open circuit voltage is 6.8 v, but drops down to nothing when attached to the MBP.

You are correct. If G3Hot is dead then the SMC is dead. If the magsafe is defective and the battery is dead then it seems there is no way for the SMC to tell the charger to send the current. However, I think the magsafe does apply 18.5V but just will not enable a higher current mode unless it gets the signal. The trickle current should be enough to power the G3Hot I would think.

In any case, we are talking about a defect somewhere (either Magsafe or battery) so I wouldn't think Apple would be designing for that scenario.

cmdrdata · Jun 16, 2011

Dadioh said:
You are correct. If G3Hot is dead then the SMC is dead. If the magsafe is defective and the battery is dead then it seems there is no way for the SMC to tell the charger to send the current. However, I think the magsafe does apply 18.5V but just will not enable a higher current mode unless it gets the signal. The trickle current should be enough to power the G3Hot I would think.

In any case, we are talking about a defect somewhere (either Magsafe or battery) so I wouldn't think Apple would be designing for that scenario.

Last night the eBay seller re-imbursed the money I paid for the dead battery. I immediately placed another order for a brand new "China made" A1322 battery. It should arrive by the last week of June.

In the mean time, I've analyzed the schematic of the G3HOT. It definitely needs a battery power to be alive 24/7. Without G3HOT powered, the SMC (H8S2117 IC chip) will have no PP3V42_G3H (3.425v). It is the Vcc power source that the SMC needed, and thus when the magsafe is connected, the center pin of the magsafe (adapter_sense signal) canot get the sys_onewire signal output from the SMC. This is "catch-22" scenario. The system expect G3HOT always on, but the battery did not keep it alive. The adapter will not turn on the 16.5v because it doesn't get a signal from the SMC. The SMC is not sending the signal because it doesn't have the 3.425v suppy from the G3HOT.

I know for a fact that my magsafe adapters (I have 3 of them) are good, so the only plausible conclusion is that I need a good battery that are at at least have nominal voltage. My dead ones are showing 0 volt, so this is definitely not going to allow the MBP to start. BTW, the G3HOT is implemented using an LT3470, a DC/DC converter that consume very little power (<0.1 uA), thus under normal condition would hardly deplete the MBP battery pack.

On the dissected battery, I mentioned earlier that the series-parallel connected cells produced about 7.5 volts. This is below the low threshold 10.9v of the Li_Ion discharged voltage, so I am guessing that the little protection cirrcuit inside the battery pack turned off the output to the battery connector, thus the zero volt output.

@Dadioh: do you have a schematic of iBook or older generation of MB/MBP that has removable battery? I suspect that the magsafe/battery power-on logic is somewhat different from later generation MB/MBP with internal battery. The older design would allow you to start the system without battery in place. I don't know about y'all but I find this problem very educational and interesting

SMC Issue on MBP 13 - Hardware Troubleshooting Guide

macrumors 65816

macrumors 6502

macrumors 65816

macrumors 65816

Attachments

macrumors 6502

macrumors 65816

macrumors 6502

macrumors 65816

macrumors 6502

macrumors 65816

Attachments

macrumors 6502

macrumors 65816

macrumors 6502

Attachments

macrumors 65816

macrumors 6502

macrumors 6502

macrumors regular

macrumors 6502

macrumors regular

macrumors 6502

macrumors regular

macrumors 6502

macrumors newbie

macrumors 65816

macrumors 6502

Our Staff