820-3437 no backlight boost

sbergin

New member
I have an 820-3437 that had some liquid damage. I have replaced a few components and now have normal behavior (fan-spin, image, proper rails, etc). Only issue is that backlight starts out at about 25V and slowly drops down to the non-boosted voltage as I measure it. I've tried multiple screens, and I replaced LP8550. I confirmed that the feedback line has continuity from Pin 5 on LP8550 to the other side of the board near D7701. Any thoughts? Thanks!
 

2informaticos

Administrator
Staff member
"Pin 5 on LP8550 to the other side of the board near D7701"
Not near; do you have contact with p?n K of D7701?

What happen with BKL_EN/PWM?
Did you try PRAM reset?
Did you try known good LCD assembly?

Check all traces, resistors and resolder all them (around U7701); pay atention to R7704 and the 6 return resistors.
 

sbergin

New member
Sorry, yes. I have continuity between pin 5 on D7701 and the boost side of D7701. BKL_EN is 2.85, BKL_PWN is 0.00 (both values are when screen is attached). All resistors measure within spec and traces look good
 

2informaticos

Administrator
Staff member
"BKL_PWN is 0.00 "
Not normal at all, check R7704.
If good, remove it and run a wire from R7715 to C7704.
Test if backlight appears.
 

sbergin

New member
yes, that gives us backlight with the jumper in place. R7704 tests fine... OK so what does this tell us? BKL_PWN is the source of the problem? Which means I have a CPU issue? I'm assuming I can't leave this jumper in place full time...

FWIW with R7704 removed EDP_BKLT_PWM measures 2.1M ohms to ground (no short), and I get about 2.4v on that line for a split second during POST. But then it quickly drops to 1.4v and then 0.0v.
 

sbergin

New member
I just read the sticky on this topic, sorry for not finding it sooner. I see now why you suggested a PRAM zap...to reset what might have been a fully lowered brightness setting. So I guess I could leave the jumper in place if the owner approves the loss of brightness control. I am curious where the issue with BKLT_PWM actually lies, at least now I can dig a bit more on my own. Thanks SO MUCH for your guidance!
 
Top