TheBloke
New member
Hi all,
Thanks to Louis for his great videos and this forum. Firstly, let me say that I am not a repair shop. I'm just trying to repair a single MBP 13" late-2011 that I bought cheap with a known fault, hoping I could repair it. I've spent more than a week on this, hoping to fix it on my own from watching lots of Louis' videos. But I think I do need some advice before I try any more drastic repairs, like reflowing the SMC.
Apologies if this post is rather long. I can't be sure what is useful info and what isn't, so I thought I'd share everything I've already done.
Here's the symptoms:
Of the seven failures, five of them are heat-affected and often go OK as the temperature of the board - or just the SMC - rises. VD0R, VN0R, IB0R, IC0R and ID0R will sometimes go OK on their own if I keep ASD running a while. And if I take my heat gun set to 100C and wave it briefly over the SMC area, on several occasions all five of these went OK immediately - though this doesn't work every time. If it's any help, I've noted that IB0R and IC0R tend to go OK at exactly the same time, and usually before the other three do. So I guess they require less extra heat to make work.
VC0C and VP0R have never gone OK. And I note that VP0R is the only sensor 'below the low limit'. So maybe these two, or just VP0R, have a different cause to the rest. If I do the Hardware Log thing in ASD EFI, it shows VP0R reading something like -12.5V.
I've tried rebooting into the OS after some of the sensors have gone OK in ASD, and this does seem to fix the symptoms - I even had a period of about four hours where it continued working perfeclty, including powering on with a single press of the button. Which I find confusing given Louis has shown that just VP0R on its own can cause the slowness, and that sensor never goes OK for me.
Examining the board under microscope does show signs of corrosion in multiple places - those little flecks of green that Louis talks about in some videos. Including around the SMC. I don't believe the board has ever had major water damage, it's just various small areas, especially around the edges.
I have spent a long time measuring resistors and haven't found a single one that consistently measure wrong according to the schematic. A couple of times I've found resistors that did seem to measure low on multiple readings, but then eventually they'd always measure correct, so I assumed this was caused by it being difficult to correctly measure these tiny SMD parts.
I have particularly concentrated my measuring and studying around VP0R both because it's the only 'below low limit' and because I've watched Louis fix it a couple of times. I can't find a single bad resistor anywhere in the VP0R circuit. It could be a bad cap or transistor? But as I can't measure those in-circuit, I'd just have to go replacing them one by one in random hope. I did actually try replacing one resistor, R5303 (27.4k), because it looked a bit dodgy and it was one that often measured low. But replacing it didn't change anything.
Today I put the board in my ultrasonic cleaner. I only have a cheap Chinese 2L device that runs at 42Khz - probably the sort that Louis calls a jewellery cleaner. I bought some PCB flux/corrosion cleaner and 99.99% pure distilled water, mixed it 10:1 according to the instructions on the label, then ran the bottom half of the board (the half with the SMC) through twice at 65C, 2 minutes per side. I did a little longer than the 1m Louis recommended as my cleaner only does a single frequency which is lower than the 60-80Khz a Crest uses. I can only fit half the board in the cleaner at a time so I figured for now I'd only clean the half that has both the SMC and the various components involved in VP0R and most/all of the other sensors. After cleaning I washed it with plenty of 99% IPA and dried with a hand fan.
The good news is that the ultrasonic didn't kill the board or break anything else. But nor did it fix anything. It's possible that I needed to run it for even longer given my cleaner is much more basic than an expensive one, and if need be I can run it again. But I wanted to get some advice before taking any more risks.
Based on my very limited knowledge, one possible next step would be to reflow the SMC? Given that I know heat on the SMC fixes five of the sensors, maybe this suggests dodgy solder joints? I saw Louis do that in one video - put flux all around the SMC, run the heat gun over it for a while until the solder melts enough that the chip will move, make sure it's still aligned then let it cool, helped with a 99% alcohol wash.
Below are some pictures, taken before the ultrasonic. At the end there's two whole-board views taken at 20MP, but as I don't have a macro lens you can't see anything too closely. Before that there are two composite images taken under the microscope: one around the SMC, one around the VP0R components. Apologies that they look like crap. My cheap-ass digital microscope can only see a very small area, so I had to stitch multiple snapshots together to make these images. They're at approx 40x magnification. On the VP0R composite I've labelled some components so you can see which parts of the board I'm showing.
As you can tell I'm working with shit tools on top of my complete lack of experience. But I'm broke and even if I wasn't, I can't justify spending too much more just for one laptop. I know you guys would rather help fellow professionals, but I'm hoping you might take pity and talk me through a few next steps. The one thing I do have going for me is I'm willing to spend a lot of time on this, and attempt anything!
Thanks in advance
Composite of SMC and surrounding area, at 40x magnification:
Composite of components in VP0R circuit, at 40x:
Board front and back:
Thanks to Louis for his great videos and this forum. Firstly, let me say that I am not a repair shop. I'm just trying to repair a single MBP 13" late-2011 that I bought cheap with a known fault, hoping I could repair it. I've spent more than a week on this, hoping to fix it on my own from watching lots of Louis' videos. But I think I do need some advice before I try any more drastic repairs, like reflowing the SMC.
Apologies if this post is rather long. I can't be sure what is useful info and what isn't, so I thought I'd share everything I've already done.
Here's the symptoms:
- Will only turn on by holding power button for about 10 seconds, until the chime is heard. Pressing it just once causes the light to flash briefly, then nothing. Therefore the device can't be woken from sleep, only a full power-on each time.
- Mouse-juddering slow in the OS.
- Voltage: VC0C - CPU V Core - above the high limit
- Voltage: VD0R - DC in Rail 0 - above the high limit
- Voltage: VN0R - MCP VCore - above the high limit
- Voltage: VP0R - PBUS - below the low limit
- Current: IB0R - Battery - above the high limit
- Current: IC0R - Computing Combined Cores - above the high limit
- Current: ID0R - DC In - above the high limit
Of the seven failures, five of them are heat-affected and often go OK as the temperature of the board - or just the SMC - rises. VD0R, VN0R, IB0R, IC0R and ID0R will sometimes go OK on their own if I keep ASD running a while. And if I take my heat gun set to 100C and wave it briefly over the SMC area, on several occasions all five of these went OK immediately - though this doesn't work every time. If it's any help, I've noted that IB0R and IC0R tend to go OK at exactly the same time, and usually before the other three do. So I guess they require less extra heat to make work.
VC0C and VP0R have never gone OK. And I note that VP0R is the only sensor 'below the low limit'. So maybe these two, or just VP0R, have a different cause to the rest. If I do the Hardware Log thing in ASD EFI, it shows VP0R reading something like -12.5V.
I've tried rebooting into the OS after some of the sensors have gone OK in ASD, and this does seem to fix the symptoms - I even had a period of about four hours where it continued working perfeclty, including powering on with a single press of the button. Which I find confusing given Louis has shown that just VP0R on its own can cause the slowness, and that sensor never goes OK for me.
Examining the board under microscope does show signs of corrosion in multiple places - those little flecks of green that Louis talks about in some videos. Including around the SMC. I don't believe the board has ever had major water damage, it's just various small areas, especially around the edges.
I have spent a long time measuring resistors and haven't found a single one that consistently measure wrong according to the schematic. A couple of times I've found resistors that did seem to measure low on multiple readings, but then eventually they'd always measure correct, so I assumed this was caused by it being difficult to correctly measure these tiny SMD parts.
I have particularly concentrated my measuring and studying around VP0R both because it's the only 'below low limit' and because I've watched Louis fix it a couple of times. I can't find a single bad resistor anywhere in the VP0R circuit. It could be a bad cap or transistor? But as I can't measure those in-circuit, I'd just have to go replacing them one by one in random hope. I did actually try replacing one resistor, R5303 (27.4k), because it looked a bit dodgy and it was one that often measured low. But replacing it didn't change anything.
Today I put the board in my ultrasonic cleaner. I only have a cheap Chinese 2L device that runs at 42Khz - probably the sort that Louis calls a jewellery cleaner. I bought some PCB flux/corrosion cleaner and 99.99% pure distilled water, mixed it 10:1 according to the instructions on the label, then ran the bottom half of the board (the half with the SMC) through twice at 65C, 2 minutes per side. I did a little longer than the 1m Louis recommended as my cleaner only does a single frequency which is lower than the 60-80Khz a Crest uses. I can only fit half the board in the cleaner at a time so I figured for now I'd only clean the half that has both the SMC and the various components involved in VP0R and most/all of the other sensors. After cleaning I washed it with plenty of 99% IPA and dried with a hand fan.
The good news is that the ultrasonic didn't kill the board or break anything else. But nor did it fix anything. It's possible that I needed to run it for even longer given my cleaner is much more basic than an expensive one, and if need be I can run it again. But I wanted to get some advice before taking any more risks.
Based on my very limited knowledge, one possible next step would be to reflow the SMC? Given that I know heat on the SMC fixes five of the sensors, maybe this suggests dodgy solder joints? I saw Louis do that in one video - put flux all around the SMC, run the heat gun over it for a while until the solder melts enough that the chip will move, make sure it's still aligned then let it cool, helped with a 99% alcohol wash.
Below are some pictures, taken before the ultrasonic. At the end there's two whole-board views taken at 20MP, but as I don't have a macro lens you can't see anything too closely. Before that there are two composite images taken under the microscope: one around the SMC, one around the VP0R components. Apologies that they look like crap. My cheap-ass digital microscope can only see a very small area, so I had to stitch multiple snapshots together to make these images. They're at approx 40x magnification. On the VP0R composite I've labelled some components so you can see which parts of the board I'm showing.
As you can tell I'm working with shit tools on top of my complete lack of experience. But I'm broke and even if I wasn't, I can't justify spending too much more just for one laptop. I know you guys would rather help fellow professionals, but I'm hoping you might take pity and talk me through a few next steps. The one thing I do have going for me is I'm willing to spend a lot of time on this, and attempt anything!
Thanks in advance
Composite of SMC and surrounding area, at 40x magnification:
Composite of components in VP0R circuit, at 40x:
Board front and back:
Thanks again for your help.