A1707 no pictures/backlight - Symptoms for a bad Mux chip?

[overshoot]

Hi,

We have a test A1707 mainboard that we use for test and we did plug our customer screen to it for testing, after a small liquid damage on his screen.
Backlight wasn't shorted but anyhow, no picture/backlight on his screen and then on our test screen as well after that.

I saw Paul lately replacing a Mux chip on the same board model after getting no picture and I was wondering what I should check before going that route.

Thanks!

 

[2informaticos]

Mainboard should get protected against bad LCD assembly.
Unless you've plugged in some corroded connector; check if is the case.
Inspect all the connectors from the client LCD assembly.

Do you get image, missing only backlight?
Also test USB activity.

 

[overshoot]

I am not getting a picture from what I can tell.
I am getting external display working using HDMI adapter.

LCD assembly had liquid damage. The cable going from the mainboard to the LCD board has green stuff.

 

[2informaticos]

What board do you have there?
820-xxxx code please...

 

[overshoot]

It's a 820-00281

 

[2informaticos]

Change U9850, it protects CPU DP outputs.

 

[overshoot]

Well I just did and same thing.
I've tried 3 different screens to rule out the screen issue but no difference.
I am still able to get a picture over HDMI though (USB-C adapter)

 

[2informaticos]

Check 3V3/5V_S0SW_LCD presence at LCD connector.
Also post DP_INT_HPD_R and voltage for pins 2, 12, 3, 6, 10 of U8510.
LCD assembly MUST be connected!

 

[overshoot]

While battery is being not plugged, I am getting the sign on the screen that battery is low.
So assuming that...

Once I plug the charger:

3V3_SoSW_LCD will spike at 3.3V for a second
5V_SoSW_LCD will spike at 3.3V for a second

DP_INT_HPD_R will spike at 3.3V for a second

EDP_PANEL_PWR_EN will spike at 3.3V for a second
PANEL_FET__EN_DLY will spike at 3.3V for a second
DFR_DISP_PWR_EN will stay at 3.3V
DFR_DISP_SMC_RST_L will stay at 3.3V
SMC_RESET_L will stay at 3.3V

EDIT: backlight is created and measures 40V for a fraction of a second.

 

[2informaticos]

Do you get the low battery sign on internal LCD?
That means the video circuit is OK.
Do you still get external video?

Check what happens disconnecting touchbar.
Also monitor SMC_LID and SMC_LID_LEFT/RIGHT signals.

 

[overshoot]

Sorry my explanation was unclear. I am getting the "low battery" sign on the external display only, not the internal display.

While testing, I am using only the logicboard and our test screen. (test screen has been tested working fine with another 2017 board)

As I know I am supposed to the the "low battery" sign, I am also supposed to see the screen turning on but it won't.
It seems the screen is being told to turn on and the voltage for the backlight is sent but for a fraction of a second.
There might be an issue between the mainboard and the screen to communicate somehow.

I will check SMC_LID tomorrow but there was no issue in the past on this board, that was our test board which was working until I've plugged a customer bad screen and the internal display never turned on again since.

 

[2informaticos]

If the backlight is created, even for short time, means LCD is recognized by PCH.
Did you try full LCD assembly?
Is it compatible for sure?
Try another one, just in case.

External video is stable?

 

[overshoot]

I have tried that same screen with a 2017 board model and it works just fine. It is a full assembly.
2016 and 2017 screens are supposed to be the same and compatible with both boards. Anyway, that was the original screen which worked fine previously with that board.
External video is stable enough to show me the "low battery" for 2-3 seconds as usual.

 

[overshoot]

As I've said earlier I've tried 3 different assembly which worked with another board so the assemblies are fine.

I can see the screen lighting up quickly (backlight turning on the screen) for a fraction of a second and off.

 

[2informaticos]

Reflow/resolder U9850, maybe need to try another...
Maybe U9800 was affected too, but is not usual.
Monitor all SMC_LID signals, just in case.

 

[overshoot]

You got it... That was a bad donor U9850 or my bad soldering ;-)
Fixed with installing another U9850.

Thanks!