Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
SMC Issue on MBP 13 - Hardware Troubleshooting Guide
- Thread starter Dadioh
- Start date
- Sort by reaction score
Measurements taken (see attached) the values in green are from the working LB, the red are from LB that has start-up issue.
The chip is getting 12V but TG, TGR and BG all have incorrect voltages. So are these the gates that open/close when they have proper voltages?
Also, the CPUIMVP_BOOT1 supposed to have 6.1V but instead has 0.8V
So you think it is the chip that is bad or the INing singals CPUIMVP_BOOT1 and CPUIMVP_LGATE1 are source of the problem?
Attachments
Another M97 fixed 
Hi, not been able to contribute lately because I was waiting on parts to arrive so had to stop working on the mac boards and focus on other work. Parts have arrived now so am back fixing mac boards. I have managed to fix another M97 820-2327-A logic board. It would not power on via magsafe or battery. When just magsafe connected there was no green LED. First i checked to see if i had voltages at the two main fuses, F6905 (16v) and F7000 (12.5v). F6905 was ok but i got no 12.5v at F7000 (PPBUS_G3H). Fuse was OK so knew U7000 was not working, this was the part i was waiting for. I had already made sure I got 16v at Q7000 and Q7001 . I replaced U7000 and now got 12.5v at fuse F7000. I still had no green LED on the magsafe so I checked the usual suspects Q6910 and U6915, and both had gone faulty. Replaced but still no green LED on magsafe. I checked to see if I go 3.42v out of U7000 pin#14 (CHGR_ACOK), which was not present. I knew U7000 was OK because I had replaced it with a new one, same with Q6910 & U6915, both new parts so I started looking elsewhere. I put my multimeter on 'continuity' measurement to see if U7000 pin#14 was being dragged to ground somehow and i was getting a reading of 55ohm which i knew was wrong because taking the same measurement on a working logic board the meter read 'OL'. So i looked for a part which could be dragging the signal SMC_BC_ACOK to ground. I took out R6910 and measure across both solder pads, one side measured 'OL' the other which is connected to Q6910 pin#2 measured 55ohm. I knew Q6910 could not be the cause because I had already replaced it with a new one and the only other component connected to pin2 is C6910. I removed C6910 and the measurement changed from 55ohm to 'OL', i had found the fault. Replaced the faulty capacitor with a good one from another board, connected the magsafe and the LED went green. Great, I had fixed part of the board, now to see if the rest had been fixed. Plugged in RAM, speakers, fan, lcd and keyboard, press the power button, board powered on, heard the chimes and LCD came on..another success
Rog
Hi, not been able to contribute lately because I was waiting on parts to arrive so had to stop working on the mac boards and focus on other work. Parts have arrived now so am back fixing mac boards. I have managed to fix another M97 820-2327-A logic board. It would not power on via magsafe or battery. When just magsafe connected there was no green LED. First i checked to see if i had voltages at the two main fuses, F6905 (16v) and F7000 (12.5v). F6905 was ok but i got no 12.5v at F7000 (PPBUS_G3H). Fuse was OK so knew U7000 was not working, this was the part i was waiting for. I had already made sure I got 16v at Q7000 and Q7001 . I replaced U7000 and now got 12.5v at fuse F7000. I still had no green LED on the magsafe so I checked the usual suspects Q6910 and U6915, and both had gone faulty. Replaced but still no green LED on magsafe. I checked to see if I go 3.42v out of U7000 pin#14 (CHGR_ACOK), which was not present. I knew U7000 was OK because I had replaced it with a new one, same with Q6910 & U6915, both new parts so I started looking elsewhere. I put my multimeter on 'continuity' measurement to see if U7000 pin#14 was being dragged to ground somehow and i was getting a reading of 55ohm which i knew was wrong because taking the same measurement on a working logic board the meter read 'OL'. So i looked for a part which could be dragging the signal SMC_BC_ACOK to ground. I took out R6910 and measure across both solder pads, one side measured 'OL' the other which is connected to Q6910 pin#2 measured 55ohm. I knew Q6910 could not be the cause because I had already replaced it with a new one and the only other component connected to pin2 is C6910. I removed C6910 and the measurement changed from 55ohm to 'OL', i had found the fault. Replaced the faulty capacitor with a good one from another board, connected the magsafe and the LED went green. Great, I had fixed part of the board, now to see if the rest had been fixed. Plugged in RAM, speakers, fan, lcd and keyboard, press the power button, board powered on, heard the chimes and LCD came on..another success
Rog
Hi, not been able to contribute lately because I was waiting on parts to arrive so had to stop working on the mac boards and focus on other work. Parts have arrived now so am back fixing mac boards. I have managed to fix another M97 820-2327-A logic board. It would not power on via magsafe or battery. When just magsafe connected there was no green LED. First i checked to see if i had voltages at the two main fuses, F6905 (16v) and F7000 (12.5v). F6905 was ok but i got no 12.5v at F7000 (PPBUS_G3H). Fuse was OK so knew U7000 was not working, this was the part i was waiting for. I had already made sure I got 16v at Q7000 and Q7001 . I replaced U7000 and now got 12.5v at fuse F7000. I still had no green LED on the magsafe so I checked the usual suspects Q6910 and U6915, and both had gone faulty. Replaced but still no green LED on magsafe. I checked to see if I go 3.42v out of U7000 pin#14 (CHGR_ACOK), which was not present. I knew U7000 was OK because I had replaced it with a new one, same with Q6910 & U6915, both new parts so I started looking elsewhere. I put my multimeter on 'continuity' measurement to see if U7000 pin#14 was being dragged to ground somehow and i was getting a reading of 55ohm which i knew was wrong because taking the same measurement on a working logic board the meter read 'OL'. So i looked for a part which could be dragging the signal SMC_BC_ACOK to ground. I took out R6910 and measure across both solder pads, one side measured 'OL' the other which is connected to Q6910 pin#2 measured 55ohm. I knew Q6910 could not be the cause because I had already replaced it with a new one and the only other component connected to pin2 is C6910. I removed C6910 and the measurement changed from 55ohm to 'OL', i had found the fault. Replaced the faulty capacitor with a good one from another board, connected the magsafe and the LED went green. Great, I had fixed part of the board, now to see if the rest had been fixed. Plugged in RAM, speakers, fan, lcd and keyboard, press the power button, board powered on, heard the chimes and LCD came on..another success
Rog
Nothing like a success story and how you got there. Hopefully this will help me and others in the future. Congratulations!
Q7510 is the output driver for the 1.1v power supply that powers the CPU. This voltage is generated by down regulating the 12v (PPBUS) to 1.1v by turning LGATE and UPGATE signals on/off via alternating square waves (just like the 12v that charge the battery). If you have a logic probe or oscope, you will probably see the square waves at these LGATE signals on the good board. So my next guess is the problem is further back to the component that drives the UGATE/LGATE signals.
Has anyone here used the solder rework station T862 for replacing BGA packages? I have quite a few boards that I believe the SMC is bad and wanted to be able to replace BGA packages. Sample listing below... there are lot's of them for sale on eBay from various sources for a little over $200.
Wanted to get first hand experience from someone that may have used one on their MacBook logic boards. Thanks.
http://www.ebay.ca/itm/PRO-T862-BGA...RDA-WELDER-SOLDERING-MACHINE-g1-/130620054775
Note: I also posted this question on the backlight thread. Not sure if all the same repair people frequent both threads.
Wanted to get first hand experience from someone that may have used one on their MacBook logic boards. Thanks.
http://www.ebay.ca/itm/PRO-T862-BGA...RDA-WELDER-SOLDERING-MACHINE-g1-/130620054775
Note: I also posted this question on the backlight thread. Not sure if all the same repair people frequent both threads.
I've never used IRA BGA machines, only hot air ones from companies such as Weller and Honton. Hot air gives a better flow of heat compared to IRA. IRA does have a tendency to burn the chips and the board if you don't get the temperature and airflow exactly right. Company has a Jovy RE-7500 BGA rework machine but it is useless for doing BGA rework as you cannot control the heat properly, it just keeps rising and rising until you see the solder melt, remove the part then set the machine too cool down. All our BGA work is done with a Honton HT-R 390. I used it to successfully reflow some logic board Intel Northbridge BGA's and the infamous graphics BGA. Have not tried to reflow an SMC bga chip yet because I have not come across one that needs reflowing.
I've been working with BGA chips and machines for over 20yrs now and I can say without any doubt, reflowing and changing BGA chips should be done using a BGA rework machine that uses hot air and not IRA. Yes by all means IRA can do BGA rework but it is not very reliable. I've worked for companies such as NEC, Dell, Packard Bell, Cellestica and not one of these companies has ever used an IRA BGA rework machine so that tells me there is something not right with them if such companies refuse to use them.
Rog
Hey Dadioh
Don't waste your money.This will not be good enough for what you want.You will ruin most boards.
For the SMC (U4900) I just use hot air gun & flux to remove & resolder a different one.Just use tweezes to pick it up so you don't damage the .3mm balls too bad unless your good at reballing.You'll have to either mark the board or have another board for reference to put it back cause there are no lines on the board.
Good Luck
Thanks guys (PEIslander and Rog). That is what I suspected. A proper station would be far too costly for me to consider. I might try the SMC with a manual method. What have I got to lose?
I have a board that I also suspect SMC is bad and want to replace it. Now, if I went your way and use hot air I still have few issues/questions:
--In case If you decide to reball (clean the chip)
1. Where do I get the datasheet for SMC. It is proprietary to Apple and I looked almost everyone to get it but no luck.
2. Where did you get the stencil for it (145 balls)? Do you have at least a Gerber file?
3. Positioning... Let's say I have all the balls placed on the chip and I'm ready to put it in place on the LB. Don't you think that it would be hard to align all the 145 balls by hand even using a good scope. Once you position the chip the problem is you don't see what's beneath it. Of course there are software controlled rework stations for $15-20k to do it precisely but who wants to spend this kind of money.
4. Do you know if the size of the balls for GPU, CPU and SB also .3mm?
If you doing it 'the inexpensive' way what are your tips and best practices if you decide to reball?
Or you always tried to remove the chip and leaving the balls intact?
Last edited:
Please help guys. Urgently needed.
I have a 2010 MacBook pro A1278
And after a short fall off my bed it no longer powers up.
I have tried everything ( SMC reset , PMU reset. Ect ect) but nothing.
The MagSafe light is green and the battery indicator shows green nearly full.
I have tried jumping the power switch ect but no signs of life.
Any ideas if there's a fuse somewhere maybe? And location. Or what else it might be.
Cheers in advance guys I'm so desperate to get this back up and running to Finish my exams.
Regards
Sy
I have a 2010 MacBook pro A1278
And after a short fall off my bed it no longer powers up.
I have tried everything ( SMC reset , PMU reset. Ect ect) but nothing.
The MagSafe light is green and the battery indicator shows green nearly full.
I have tried jumping the power switch ect but no signs of life.
Any ideas if there's a fuse somewhere maybe? And location. Or what else it might be.
Cheers in advance guys I'm so desperate to get this back up and running to Finish my exams.
Regards
Sy
Hi there,
I've got myself a water-damaged A1278 / K90i.
The symptoms match with the ones mentioned in the first post of this thread.
Using the SMC bypass the laptop turns on (didn't boot it yet as I am missing suitable RAM).
From what I've read, I am missing about 0.25V at G3Hot, as I am
measuring ~3.19V there and 3.425V are expected.
I've measured so far:
- F6905: 16.5V (OK)
- R6990: 16.2V (OK, expected voltage drop by resistor)
- D6990: 16.0V (OK, expected voltage drop by diodes)
- C6991: No short (if broken, FB might regulate the output voltage down)
- C6996: No short
- C6999: No short
- C6994: No short
As far as I can see, U6990 supplies the G3Hot voltage and is fed
with 16V, which seems alright to me. However, the resulting voltage
(Pin 9, Pin 6) is too low (3.19V), so I suspect that the U6990 is broken.
Everything else seems to work just fine, except that everything's
off by ~0.2V, for example SMC_BC_ACOK, which seems logically to me.
I'm not entirely sure if I didn't miss something so I'm asking for a
second opinion. Is my conclusion that U6990 is broken correct?
<strike>In case I'm correct I would be happy about every hint on where I can
buy a replacement of what seems to be a ST PM6644.</strike>
Nevermind, the PM6644TR seems to be the compatible successor and is available
for sale.
Thanks!
EDIT #1: I just found out that the display backlight does not work as well.
Can this be connected somehow?
EDIT #2: I did some measurements on the two PPVOUT_SW_LCDBKLT pins
on J9000 (the display cable) and I'm getting on both 12V (11.9V). I don't know if
this is the right voltage (didn't find any reference) but intuitively I would
say it is alright. I guess the inverter is affected somehow, so disassembling
the display unit is next.
EDIT #3: According to this post from Dadioh 12V are OK.
I've got myself a water-damaged A1278 / K90i.
The symptoms match with the ones mentioned in the first post of this thread.
Using the SMC bypass the laptop turns on (didn't boot it yet as I am missing suitable RAM).
From what I've read, I am missing about 0.25V at G3Hot, as I am
measuring ~3.19V there and 3.425V are expected.
I've measured so far:
- F6905: 16.5V (OK)
- R6990: 16.2V (OK, expected voltage drop by resistor)
- D6990: 16.0V (OK, expected voltage drop by diodes)
- C6991: No short (if broken, FB might regulate the output voltage down)
- C6996: No short
- C6999: No short
- C6994: No short
As far as I can see, U6990 supplies the G3Hot voltage and is fed
with 16V, which seems alright to me. However, the resulting voltage
(Pin 9, Pin 6) is too low (3.19V), so I suspect that the U6990 is broken.
Everything else seems to work just fine, except that everything's
off by ~0.2V, for example SMC_BC_ACOK, which seems logically to me.
I'm not entirely sure if I didn't miss something so I'm asking for a
second opinion. Is my conclusion that U6990 is broken correct?
<strike>In case I'm correct I would be happy about every hint on where I can
buy a replacement of what seems to be a ST PM6644.</strike>
Nevermind, the PM6644TR seems to be the compatible successor and is available
for sale.
Thanks!
EDIT #1: I just found out that the display backlight does not work as well.
Can this be connected somehow?
EDIT #2: I did some measurements on the two PPVOUT_SW_LCDBKLT pins
on J9000 (the display cable) and I'm getting on both 12V (11.9V). I don't know if
this is the right voltage (didn't find any reference) but intuitively I would
say it is alright. I guess the inverter is affected somehow, so disassembling
the display unit is next.
EDIT #3: According to this post from Dadioh 12V are OK.
Last edited:
Hey
As far as I know you can buy a new chip but can't get the software to program it.You'll need a parts board for the replacement.
I don't have a stencil for it yet so I try to reuse the solder already on the board & chip.
I'll include a link where at the first of the video buddy just puts flux on the chip or board & applies solder to iron then to board or chip.The amount of solder that sticks to the pads is close to .3mm & seem's to work for me if I mess up any solder pads.
Just take a good visual inspection of where it should go or mark it.It's gonna move on ya when you start heating it anyway.Just hold it close to where it should go & when things get hot it should suck right in place with a very gentle tap or two..Before applying power double check alignment with another board or maybe a hi def pic you took before removing.
I don't know what size the GPU or CPU.I'm guessing .6mm or maybe seen it the same place I seen the SMC was .3mm.
I got lucky on the first one I replaced & its still working about a year later.I just pulled & replaced no soldering iron at all just hot air.
Good Luck
http://www.youtube.com/watch?v=eo6ZH72yvsg
You was extremely lucky in your SMC replacement. Reballing has to be precise, it should never ever be attempted by guess work using a soldering iron and solder to try and get the correct amount of solder on each pad. Even if the solder you replaced is out by 0.0001% in mass then you will suffer a 'dry joint' meaning when the solder heats up, due to the incorrect amount of solder you tried to 'guess' was the right amount, it will fail to make contact with the solder pad. If you press down on the chip gently during the reflow stage, any pad that has had too much solder applied will cause a 'short' because the excess solder will spread out and make contact with the solder pad next too it.
Unless you are highly skilled in BGA rework/reballing I would not even consider contemplating reballing an SMC IC by hand (soldering iron and solder). By all means take one off a faulty board and TRY to re-seat it on the new board but are taking a big risk in that either you will get a 'dry joint' or possibly worse a 'short' which could potentialy damage the SMC IC and depending on the type of 'short', more damage to the board.
It amazes me the amount of mis-information there is out there about BGA reworking/reballing. In every instructional video i have seen, either beit on youtube, dailymotion or another video source, not one tell's you their fail rate precentage, all they show you is the 'one' success. I know from personal experience of working with BGA's over the past 20yrs that the fail rate of BGA rework/reballing, even when done by professionals with all the proper kit is extremely high.
Please think very carefully before you consider trying to reball an SMC chip.
Have you tried to unplug the battery then do the SMC bypass mentioned in the first post of this thread.Even try a different wall outlet or power cord?
Hi Rog
Yes I'm a newbie redneck and don't really have a clue about this stuff.I would love to see a video of someone reballing the SMC chip proper.
As dadioh said "what have I got to lose?"The board is either not working or not charging anyway.I don't have money to get fixed professional.I'm not experienced enough to do it right yet.I have replaced that chip my way between 10 & 20 times & I guess I'm real lucky (not)cause most times it works.I have had a couple where the pad came off the board & chip.They were water damaged though.I think even if I had a BGA rework station I would use my hot air station to mount to the board.I would like to learn to reball it first though.I don't recommend doing it my way unless its your last option.Fix it or **** it.LOL
Last edited:
That very last bit would be good for a repair shop slogan 'Fix it or **** it' haha
The only other easy thing to try is to unplug everything except maybe fan & speakers.Try booting up with the power pads on the board.I had one that the keyboard & DVD drive wouldn't let it power up but it at least tried to start.Other than that I don't know what else to try.You may have to do some more reading of this thread & take some voltage readings.
Good Luck
Hello all. Hopefully you can give me a little guidance too. LOTS of info in this thread.
I have a 13 inch, 2009 macbook pro from a family member that had a slight liquid spill. I've scrubbed the logic board down with an industrial 99.9% denatured ethyl alcohol. Here is the fun part: if I put in a charged battery, it boots up normally, but wont charge. If I do the SMC bypass, it boots from the magsafe, but doesn't charge the battery. Any guidance for this lost sole?
I have a 13 inch, 2009 macbook pro from a family member that had a slight liquid spill. I've scrubbed the logic board down with an industrial 99.9% denatured ethyl alcohol. Here is the fun part: if I put in a charged battery, it boots up normally, but wont charge. If I do the SMC bypass, it boots from the magsafe, but doesn't charge the battery. Any guidance for this lost sole?
The IC responsible for charging might be broken (or one of the peripherals around it). The name is ISL6258 (or ISL6259). Measure the voltage at pin 14. If it is not around 3.3V, there's something broken.
If your MacBook is fully functional from the battery but not from the magsafe then the overvoltage protection circuit has got damaged. There will be no green led or battery charge orange led from the magsafe. The post from nemo points to one possible problem but in my experience this is very rare. When the charge IC ISL6295 goes faulty, it tends to go in a dramatic way which prevents the whole MacBook powering on, even from battery. I have not had one yet that has gone faulty and only pin#14 has been the problem, hence why in my opinion I believe ISL6295 is not the problem, but please do not discount this yet. The overvoltage protection circuit needs two control signals, one from the SMC and the other from the charge IC ISL6295. Usually when electronic components fail, they do so in a manner which damages 99% of the component which in the majority of cases the MacBook would never power on via magsafe or battery. Due to your MacBook fully working via the battery, this would indicate that the SMC and charge IC have to be working 100% BUT it is possible that both are working 99% and the 1% fail is due to both components loosing the control signal to the overvoltage protection circuit. On ALL the boards I have fixed that are fully functional via the battery and not the magsafe it has always been 1 or 2 components in the overvoltage protection circuit that has gone faulty.
To be able to pinpoint what actual parts possibly need replacing and a diagnostic tip on determining which part needs replacing I need to know specifically the logic board type. On the labels attached to the memory socket it will usually have an id like K6 or K24 or K90. If not then imprinted onto the logic board itself will be an 820- number. The reason is there are too many MacBook pro variants and some of the circuitry is different in certain models so we need to be sure as to which model of logic board you have so we can continue onto the next part, the diagnostic tip.
I have tried that but still the same. I have take voltage readings from the power board to the main board and all reads fine. I'm just unsure where else on the main logic board to take readings and location of fuses as my board being late 2010 to early 2011 is different from those mentioned in earlier posts on this thread.
Any help would be great fully appreciated please.
Kindest regards
It would appear as though U6990 is faulty but measure both pin#9 & pin#6 their resistance to ground to make sure that nothing it pulling those pins to ground, your meter should read 'OL'. As for the backlight, the voltages of those pins you measured is wrong, pin#21 & pin#22 ( PPVOUT_SW_LCDBKLT ) should both be 27v. Check to make sure the fuse F9700 supplying voltage to lcd controller IC U9701 is OK