Macbook 820-00840 no image, no backlight

[kbless7]

I have an 820-00840 that has water damage. The corrosion was around the backlight driver, WiFI chip, inside the screen connector, and the BGA chip (PI3USB32) next to the USB port that doesn't show up on schematics. The board was initially stuck at 5V, but there were no shorts or obviously blown chips. After brushing the corrosion away from the PI3USB32 chip (no reflow yet), the board started taking 20V with peak 0.6A that settled to 0.3A. That seems normal from what I've found online.

I brushed the screen connector and checked the resistance between PPVOUT_S0_LCDBKLT and EDP_INT_AUX_N. It was ~1MOhm, so I figured the screen was still good. I plugged in the screen connector and powered it on. Lo and behold the screen turned, and the apple logo appeared with the loading bar. I didn't have any connectors plugged in at that point, so disconnected power and put the board back in the computer and hooked everything up. I powered it back on to check functionality before reflowing the corroded areas, but now the screen doesn't turn on

I could not see an image using a flashlight, so it appears that I'm at no image, no backlight. I get a single chime when booting, and an external monitor, trackpad, keyboard, etc. works, so the CPU isn't completely dead. I don't know if the above screen pins shorted when I had the board upside down checking the backlight circuit and killed that part CPU or what. I don't have another screen to test, so I'm hoping someone has an idea of how to determine if it's the CPU, screen cable, screen, or something else.

Some things I checked that may help. I don't know what signal is necessary for the CPU to know the screen is present and should turn it on.
EDP_BKLT_EN - 0V
PP3V3_S0SW_LCD - 0V
PP5V_S0SW_LCD - 0V
DP_INT_HPD - 0V

[piernov]

Re: Macbook 820-00840 no image, no backlight

Make sure you tried PRAM reset. Confirm both SMC_LID signals are high.

The LCD power rails should come up for a short while when turning on the machine. They'll fall back down if the LCD cannot be recognized.
Was there any sign of damage between the backlight rail and the eDP AUX line in the connector (on board/LCD cable)?

[kbless7]

Re: Macbook 820-00840 no image, no backlight

Make sure you tried PRAM reset. Confirm both SMC_LID signals are high.

The LCD power rails should come up for a short while when turning on the machine. They'll fall back down if the LCD cannot be recognized.
Was there any sign of damage between the backlight rail and the eDP AUX line in the connector (on board/LCD cable)?

I did a PRAM reset. It did not help, but I do have another problem that it won't post with J6951 connected; that may be due to connecting and disconnecting the battery though.

I can't measure the SMC_LID signals since they are on the bottom of board when I have to mount the board in the case. I am able to measure LID_OPEN, which is the ORed output of both SMC_LID signals, and it is high when the screen is open. Does that mean it doesn't recognize the screen?

I was able to measure EDP_PANEL_PWR_EN, PANEL_P3V3_EN_DLY, PP5V_S0SW_LCD, PP3V3_S0SW_LCD, and BUF_EDP_PANEL_PWR_EN with my scope. All signals went high for 500-600ms once the current draw decreased to 0.3A. DP_INT_HPD was 3.3V for 350ms at about the same time as the other other signals. I was not able to see anything on EDP_INT_AUX_N or EDP_INT_AUX_P.

Yes, there was charring in both connectors shown in the attached images. I brushed away the junk and measured the resistance between EDP_INT_AUX_N and PPVOUT_S0_LCDBKLT to be 0.5MOhms. That's a tad lower than the 2-3MOhms I saw in a Rossmann video. The screen worked initially, so I'm wondering if the 2 pins shorted when the cable moved around.

[piernov]

Re: Macbook 820-00840 no image, no backlight

Once HPD goes high there should be communication on the eDP AUX lines.
Pretty sure that killed the eDP AUX lines. No way to get back internal display.
I'm surprised it somehow worked for a moment, it shouldn't have. Note that carbonized material is conductive so if connector is not replaced (or at least burnt stuff digged out entirely) then it can still arc.
You could check in diode mode instead, I don't know what the value should be but it might be a good indication if the lines are indeed dead. Also diode mode measurement should be the same between _P and _N (couple mV difference at most).

[kbless7]

Re: Macbook 820-00840 no image, no backlight

I did clean the carbonized material with an ESD brush before first connecting the display, so perhaps that's why it worked at first.

What exactly should I be measuring between diode mode? Between both EDP_AUX and PPVOUT it is open circuit with 1mA that my DMM using. It's also open circuit between EDP_AUX and ground. I tried the probes the in both direction to make sure.

[kbless7]

Re: Macbook 820-00840 no image, no backlight

Well this board is just messing with me now. I hooked up all of the connectors including battery and the backlight came on with 44V at PPVOUT. Of course there was no image, and I couldn't get the backlight to come on again.

[piernov]

Re: Macbook 820-00840 no image, no backlight

Sorry I meant diode mode to ground. Must be checked before the capacitor, on the CPU side.
It is very weird that you had it working once and now backlight came on. Maybe the eDP AUX line isn't dead (and that would be a first) and it's just the bad connectors messing with you.

[kbless7]

Re: Macbook 820-00840 no image, no backlight

Thanks for the tip. I measured diode and resistance on the CPU side of C8529 (EDP_AUXCH_C_N) and C8528 (EDP_AUXCH_C_P). Here are the results in a pseudo table.

                  AUX_N    AUX_P
Diode mode (red to ground)     0.256V   0.259V
Diode mode (black to ground)    0.289V   0.292V
Resistance (red to ground)     1.56k    1.56k
Resistance (black to ground)    10.1k     10.k

So the AUX channels don't appear to be shorted to ground.

[kbless7]

Re: Macbook 820-00840 no image, no backlight

I fixed the computer! I replaced the screen connector, and everything is good again. The resistance between PPVOUT_S0_LCDBKLT and EDP_INT_AUX_N was measuring 500kOhm before replacing the connector. After replacing the connector it was infinite resistance. So the decreased resistance was enough to interfere with communication or there was physical contact between the AUX pins due to the connector being burnt.