Mac Air A1466 2012 cpu 300%

[smic717394]

Hi,

I bought a mac air 2012 very cheap, because the customer didn't want to fix. After reflowing the ram chips the laptop works fine, I tested the ram and no error. But The fan goes at max speed after few seconds. Running the Apple diagnostic finds an error 4SNS/1/C0000008: Th1H--124. and the cpu usage is like 300%, After disconnecting the webcam cable that on the 2012 model is on the IO board, the laptop, the laptop runs fine, and running the apple diagnostic finds no errors. So the problem is related with the webcam, but I suspect the problem is no the webcam but the light sensor. Connecting the webcam cable runs very slow and if I test the webcam work but the light sensor no. Removing the problem cures the problem. I´m not going to chenge the screen in this 2012, I can either leave it without the webcam or maybe if there is anyway to disable the light sensor, maybe there is a resistor or something next to the camera connector. And yes I already replace the io board and the cable, same problem.

Thank you.

[mon2]

Check the voltage to ground of the smbus lines (scl and also sda) for the ALS (ambient light sensor). Do so very carefully.

I2C_CAM_SCK
I2C_CAM_SDA

SMBUS_SMC_1_S0_SDA
SMBUS_SMC_1_S0_SCL

[smic717394]

What connector is this? is where the ribbon from the io to main? because the camera conector only has 4 pins i think.

[mon2]

Post your 820-**** board number. The schematic I referenced may be a different version but is the same model.

[smic717394]

Ok, I got it. I´m attaching the schematic and boardview in case someone needs it for the main board 820-3209-A and the schematics for the IO board. I see from the schematics that on the CAMERA/ALS connecter that pin 5 is scl and pin 6 is sda, I´ll check them tomorrow under the microscope they too small. Just to confirm pin 5 and 6 of the connector are the las 2 from the left wright? Maybe I can desolder pin 5 and 6 just to try.

[MOD EDIT] Schematics -> https://www.badcaps.net/forum/troubl...-and-boardview

[mon2]

Suggest to measure the voltage to ground on each pin of this connector. We can figure out the rest. The expectation is that pin # 1 will be the one with zero volts. You can also confirm this by removing all power, meter in diode mode, short your meter probes and you will hear a tone beep. Then connect one meter probe to ground on the logic board (any metal shield) and other meter probe onto pin # 1. You should hear a tone beep again. Continue to test the rest of the lines in DIODE mode.

In diode mode, connect the red meter probe to ground; black meter probe to the point to test. Post each measurement.

Next, move to th DC volts mode on your meter (5 volts or higher scale is ok).

Red meter probe to point to test; black meter probe to ground (shield is ok).

Measure each pin again. Need to confirm that there is a voltage present on pins # 5 & # 6 unless these lines are stuck or defective. On I2C lines, there is always a pull-up resistor which goes to a pull-up voltage. It is that pull-up voltage we expect to see on a known good I2C SCL or SDA line. Anything too low will halt the I2C communication traffic which will cause all other devices / sensors on that communication network to fail and hence will result in a crazy high fan spin.

[smic717394]

Thank you for taking the time to reply, good idea, pin 1 is GND so the opposite is i2c bus. I will measure the voltages tomorrow with the connector plugged and removed but I think desoldering the 2 i2c pins should fix the problem, I´ll update tomorrow. Thank you.

[smic717394]

ok. Its working now. Pin 1 is the ground and its the first pin towards the speaker connector. I measured the voltages not on the pins so I dont short any but on the test points. Second and third pin was showing like 5V then droping same for pin 4. I should have measure the resistance with the camera cable connected on pin 5 and 6. Pin 5 or 6 i cant remember was showing 3,3v and the other 0,5v. Anyway what I did is desolder pin 5 and 6 that are the last 2 pins so they don´t make contact with the board, and now runs fine.

The fan speed its normal, the webcam works fine, the cpu usage is normal. Thank you for the guidance.

[mon2]

Good to hear of your progress. You will probably lose out on some I2C bus device function by removing the respective lines. At this time, not sure what exactly. Suggest to seek out a used but known good LIO board as a replacement. Suspecting that there will be an ESD TVS diode on these I2C lines on the LIO board somewhere. Perhaps SC-70 or SOT-23 package. Some of the parts used by Foxconn / Apple can be tiny and the black solder mask does not help. Either way, it is probably that ESD diode that is shunting the line from an ESD event. Just an educated guess as we wasted (invested?) 3 days on a similar repair where multiple local shops said was a no fix. Came down to a < $ 0.50 USD TVS dual diode that is sacrificial by design. This part, after an ESD event, will shunt the lines they protect to ground which will halt all I2C traffic on the network.

[smic717394]

I did replace the IO board and the cable before, with the same result, so the problem must be on the screen itself. I checked the track from the scl/sda pins and they pass to a test point and from there i see it gows directly to the big connecter, nothing in between, so its either the main board or the screen, either way since the only problem is the auto brightness and for its age, its good as it is for now. Just to education purpose I will check where this lines goes from the logic board,

[mon2]

Well noted. Good investigative work. This is one of the screens where we actually carved open to fix a defective LVDS cable and then sealed it back up again. There are YT videos on this exact display for the LCD replacement. Not impossible but took 2 of us to perform this surgery. As you have noted, leave it alone if it works for you.

[xloomx]

Ok, I got it. I´m attaching the schematic and boardview in case someone needs it for the main board 820-3209-A and the schematics for the IO board. I see from the schematics that on the CAMERA/ALS connecter that pin 5 is scl and pin 6 is sda, I´ll check them tomorrow under the microscope they too small. Just to confirm pin 5 and 6 of the connector are the las 2 from the left wright? Maybe I can desolder pin 5 and 6 just to try.

Thanks for the files!