(347) 552-2258 [email protected]

820-00840 with backlight problem and low PP5V_S0 resistance to ground

S

sekidata

Member
This 2017 MacBook Pro board came with a dark screen and liquid damage in the backlight circuity area. I found a shorted cap (C4811) which I replaced. The backlight came on for a few days, then died again. I've replaced U8400 and checked all resistors and traces in that area. Machine boots and backlight enable signal is present. Still no backlight.

Then I found that PP5V_S0, which supplies U8400, has a low resistance to ground (95 ohms; 0.09 V in diode mode); a good board shows 5+ kohms on this rail. I'm having a hard time figuring out where this partial short is located, since injecting 5 V only gets 0.1 A flowing and nothing heats up. Any way to pinpoint the culprit?
 
2informaticos

2informaticos

Administrator
Staff member
If machine boots and have external video and USB activity, then must be a problem with internal video.
If gets image on internal LCD, then go straight to backlight circuit.
Check anything backlight related connected to 5V_S0 rail; remove one by one, until find the culprit.
 
Last edited:
S

sekidata

Member
Ok, I resolved the short on PP5V_S0 by replacing U8400 a second time (it now measures 0.43 V in diode mode. However, the backlight is still nonfunctional.

I verified that the machine boots and gets an image on internal and external display. BLKT_EN is 3.3 V but there is no output from U8400 at all. The only other observation is that the SCL and SDA data line voltages drop from 5.1 to 4.5 V as soon as BKLT_EN turns on.
 
2informaticos

2informaticos

Administrator
Staff member
That may be sign of I2C bus activity; better to check with o-scope.
Check all resistors and traces connected to U8400.
Post Q8400 voltages.

Did you try known good LCD assembly?
 
S

sekidata

Member
Yes, I'm using a known good display. There were burnt vias around U8400 pins 9 and 10 and I had to run wires from those pins to R8400.
Below are the Q8400 and U8400 voltages.

Q8400:
pin 3: 12.6V
pin 4: 13.1V
pin 1,2,5,6: <2V

U8400:
pin 1,2: 0.3V
pin 4: 0.0V
pin 5: 5.1V
pin 9,10: 13.1V
pin 11: 12.4V
pin 12: 0.0V
pin 15,16: 4.5V
pin 17: 3.3V
pin 18: 5.1V
pin 19: 0.2V
pin 21: 0.0V
 
2informaticos

2informaticos

Administrator
Staff member
Backlight adjust is done through I2C bus (BKLT_SCL/SDA lines); is not sufficient to get enable signal.
The information sends by PCH (PWM) passes through LCD T-CON before arriving to U8400 (after conversion to I2C).
You may have other traces, or components damaged around U8400.

Check if EDP_BKLT_PWM is present at least.
 
S

sekidata

Member
EDP_BKLT_PWM is 2.5V.

My understanding of the PWM for these displays is limited. Can you explain the PCH - LCD - U8400 connectivity some more? Is the LCD T-CON the board on the display itself? Does the PCH send info to the LCD via EDP_BKLT_PWM and the LCD then sends info to U8400 on the I2C_BKLT data lines?
 
S

sekidata

Member
I'm at a loss as to where else to look. Here's what I have:

I verified that BKLT_SCL and BKLT_SDA are both connected to the T-CON board.
Both data lines are pulsing (although at a higher frequency than on a known good board).
I verified that EDP_BKLT_PWM is between 2.5 and 3.3 V and connected to J8500.
I have tried 3 displays and 2 different display cables.
 
2informaticos

2informaticos

Administrator
Staff member
Then must be some traces, or other components affected by liquid on the backlight circuitry.
Leakage between backlight output to data lines is very dangerous.
 
S

sekidata

Member
How important is the sense resistor R8400? That's where I had burnt traces. Under what R8400 conditions would U8400 refuse to activate? How can I test it?
 
S

sekidata

Member
Yup, it is. I also replaced R8400 to no effect.

I'm giving up on this board - sometime you just have to accept that a deep mystery remains.
 
You must log in or register to reply here.
Top