820-2936 - Green / Orange Light, Fan spin, No Chime & No display.

smiba

New member
I don't even know where to start at this moment anymore, I've been working on this device most of my day and It just doesn't want to work.

Story:
Customer handed me his macbook. Told me he got the motherboard replaced at company *X* for $700 a year ago, but now it keeps rebooting, is unstable or doesn't want to power up at times. He checked and said there was lots of corrosion but company *X* told him it was his fault and not theirs.

After opening up the macbook and inspecting the motherboard it looked like absolutely trash underneath the microscope. Flux everywhere and it just looks like all the parts were resolderd atleast once. On every single part of the board where the pads would usually be nostuff there is solder now, indicating someone went past it with his iron.
They also used this annoying thick flux which is just surrounding most of the chips, its not solid but its definitely hard to remove at room temperature, it also makes it hard for me to probe points.
After some good inspection the most screwed up part was around the SMC, quite the bit of corrosion. I cleaned most of the corrosion up with 99.9% 2-Propanol and checked but it didn't solve the issues (When will it ever be simple? haha), the system was still acting rather oddly and most of the time wouldn't even power on..
I expected corrosion underneath the SMC, since its the logical explanation why power management is acting weird. I de-soldered the SMC and placed on a new one. Well this is where the bigger issues started, now the system always powers on.. But there is no Chime and No display.

Often the system stays on for about 8 seconds before it shuts down completely, if you quickly reconnect the charger it will power on again and stay on (But still no Chime and display).

VCore is 1.15V and the CPU gets warm, if I power it on without any ram modules installed the system would start beeping, indicating an issue with ram.

I then replaced the SMC chip with another new one (You never really know what kind of crap china sends you), however the issue stayed exactly the same.

I also checked the data going in to the SPI Chip (U6100), the CLK is there and it seems like data is both going IN&OUT. The clock and data looks pretty odd compared to a working 820-3115, in the start there is about 1-2 seconds of 0.5 second long high/low pulses before actual data is being sent. Also sometimes these long pulses happen while also data is going over it.
It might not be relatable to this issue, though.

Short Version:

Macbook had issues with booting at times, very unstable. Corrosion on the board and just a lot of flux everywhere from someone else. Cleaned corrosion and decided to replace SMC chip due lots of corrosion around that area. After new SMC the light on the charger shows, the fan spins but there is no Chime and no Display or backlight. Replaced with another SMC chip (from the same batch), same issues.
CPU VCore is 1.15V, CPU gets warm and when booted up without RAM memory it will beep.
System often shuts down after about 8 seconds, but when booted up for the second time in a very short period it will stay on.
Checked U6100 (SPI Chip's clock and data, looked a bit off compared to a 820-3115. I however don't have another 820-2936 to compare.

All major power lines are up as far as I checked

I got a digital oscilloscope, its not state of the art technology but good enough to see clocks and data over non-high speed chips. If you want me to check some part with the oscilloscope I can post screenshots of its output.

Questions:

- Where to look? What could be causing this issue?
- Might the second SMC also be bad? Is it worth a fair try to replace it a 3rd time but with a totally different seller's stock?
- Is it possible I toasted the chipset (U1800) while soldering the SMC back on? (I doubt it, I didn't use /insane/ temperatures. I can't exclude it either though..)

Thanks in advance, let me know if there are any questions.
 
Last edited:

dukefawks

Administrator
You need to use SMC from an identical donor board because they have firmware. Not a new one or some China random reballed one.
 

smiba

New member
Oh my, what a noob mistake.
Only if I'd just asked it earlier I could've been hanging on the couch instead haha

Just to be sure, if I get the SMC from a 820-2926 donor board it should work? If not I'll just get a BGA stencil for the SMC and I'll reball the original SMC I de-soldered.
 

smiba

New member
Oh yeah obviously, 3AM is really getting to me.

Thanks for your help again, sorry for having to listen to my dumb questions!
 
Last edited:

smiba

New member
Finally had time to solder on a SMC from a 820-2936 donor board but it does not POST.

Fan stays on and gets louder the longer it stays on, the CPU gets warm.

I'll check tomorrow why it doesn't complete its POST, this behaviour seems to match the issue the customer described. Although I can not be sure if this is actually an issue that was there when it got to me or if the replacement SMC is still screwing up. (I have no idea if the SMC can prevent the system from completing POST but still getting into an S0 state)
 

smiba

New member
Actually I've been mistaken, the system does POST! I had the assumption the chime would sound from both speakers, however it only does from the speaker next to the wireless module. My bad
(And since it posts the USB mouse indeed did light up)

It runs unbearably slow though. I ran AHT and it came back with (4SNS/1/40000000: VD0R-19.314)

Which if I'm correctly indicates a sensor issue (4SNS), with the DC voltage sensor. I think the 19.314 number might be a voltage, but even without the charger plugged in this number stays above 19V so its not likely PPDCIN_G3H. When it runs off the battery its slightly lower though (3 tests, ~19.01-19.19 on battery, 2 tests >19.3 on charger). Maybe its noting like a voltage though and I'm just speculating too much

Since the SMC (If I recall correctly) processes the sensor information, is it still perfectly possible I didn't solder the SMC back on properly or the replacement SMC is damaged?

Thanks again for the support
 

dukefawks

Administrator
As been discussed many times before Q5300 and resistors around it. Also run the proper EFI ASD test, not this built in garbage.
 

smiba

New member
PBUS_S0_VSENSE is 12.55V and SMC_PBUS_VSENSE is 2.07V, R5301-R5304 all within their 1%. Nothing wrong with the Q5300 circuit as far as I'm aware

I'll run ASD in a moment (First time I heard of that tool, thanks), if it doesn't give me any more clarity I'll reball the SMC in case I may have messed that up.

Will report back later today
 

smiba

New member
EFI ASD came back with all the sensors passed except VD0R (DC in Rail 0)

"Sensor is reading above the high limit"

Which sounds very odd to me, its not shorted with another power line or sensor because I would've seen that when I probed SMC_PBUS_VSENSE.
Can you confirm 2.07V is a normal value for SMC_PBUS_VSENSE?
 

dukefawks

Administrator
Are you using a 60W charger? If you use a 85W one it will report this error. You can also check the hardware profile in ASD and all sensor values are listed there.
 

smiba

New member
I was using a 85W charger, I had no idea this would cause the test to fail.
The test comes back with failed on the DC in voltage too with a 60W charger and without charger at all

The value measured is between 19.1-19.4V

I do have another 820-2936 laying around but I'd have to take it apart from the laptop, I'll see if I can check what SMC_PBUS_VSENSE measures from there
 

smiba

New member
C5304 powered on 2.07V
C5314 powered on 2.79V

C5304 powered off 0.00V
C5314 powered off 2.82V
 

smiba

New member
Some additional measurements

PPDCIN_G3H:16.85V
DCIN_S5_VSENSE: 16.84V

R5311: 99.8k Ohm [Should be 100kOhm]
R5312: 100k Ohm [Should be 100kOhm]
R5313: 27.2kOhm [Should be 27.4kOhm]
R5314: 5.46kOhm [Should be 5.49kOhm]
 

dukefawks

Administrator
If you continue the ASD after the error is this the only sensor that fails? 2.8V after the resistor divider is correct for about 16V of DCIN voltage. What is ASD reporting for VD0r in the hardware profile?
 

smiba

New member
Yes, this is the only sensor thats failing, the other sensors come back without any errors and with appropriate values

It currently reports that VD0R is 19.105V with the 60W adapter

Is it worth a try to remove R5313? The low-limit is set to 0V so it should report back fine with that resistor removed. If the SMC still reads a high value but there is no voltage on SMC_DCIN_VSENSE its to be sure the SMC is malfuncioning

EDIT:
One thing in the Hardware Profile what stood out to me is that on "ACAdapterPort" it reports the adapter-voltage as 19.485V, its probably because of the faulty VD0R readout?
 
Last edited:
Top