820-4924 dead after short on J4802

Fizzl92

Member
Hello,
this 820-494 is dead after the customer swapped a faulty trackpad cable. He did not removed the battery and as he want to plug it in he saw a spark on J4802 (the corrosponding connector at the trackpad - not at the board). So i guess he shorted PPVIN_S4_Tpad to GND. Now its dead and missing PM_SLP_S5_L and i dont really know where to start. The first missing rail is PP5V_S4 not enabled by PM_SLP_S5_L...
 

2informaticos

Administrator
Staff member
No S4 voltage present on board with only battery.
Unless machine was in suspend state, closing the lid.
Nor even S5 voltage is present when machine is really turned off and charger is disconnected.

3V42_G3H is present on battery only, but it doesn't go to J4802, but J4813.
Do you get PPBUS_G3H and 3V42_G3H now?
 

2informaticos

Administrator
Staff member
Do you get Magsafe light?
Original 60/85W Magsafe 2 charger used?

Post SMC_ADAPTER_EN, PM_BATLOW_L, 3v3_S5, 3V3_SUS.
 

Fizzl92

Member
Magsafe working normally. Yes original 85W charger.
SMC_ADAPTER_EN = 3.39 V
PM_BATLOW_L = 3.35 V
3v3_S5 = 3.34 V
3V3_SUS = 3.33 V
 

2informaticos

Administrator
Staff member
PM_RSMRST_L steady high?
PM_PWRBTN_L reacts triggering SMC_ONOFF_L signal?
If yes, any pulse at one of PM_SLP_Sx_L signals?
 

Fizzl92

Member
PM_RSMRST_L is steady 3.3 V
PM_PWRBTN_L goes down from 3.4 V to 0.35 V if i trigger SMC_ONOFF_L ald stays low as long as i short SMC_ONOFF_L
On my scope i can see a pulse on SLP_S4_L a few seconds after plugging power in. But ist in mV scale. SLP_S5_L jummps up a few mV after plugging power in and stays there.
 

2informaticos

Administrator
Staff member
Test the board alone, without RIO board and TP/KB connected; also no battery.
It should start automatically when charger is connected.
Trigger SMC_ONOFF_L only to check signals reaction.

Check basic signals for U0500 on page 13 (SYM 8 OF 19).
 

Fizzl92

Member
All measurements are i took were with the board outside the case.
I checked all the signals around U0500 and here are the Signals that go in U0500 (SYM 8 of 19):

PCH_SUSACK_L = 3.3 V
PM_SYSRST_L = 0 V
PM_PCH_SYS_PWROK = 0 V
PM_PCH_PWROK = 0 V
PM_PCH_APWROK = 0 V
PM_RSMRST_L = 3.3 V
PM_PWRBTN_L = 3.42 V
SMC_ADAPTER_EN = 3.38 V
PM_BATLOW_L = 3.33 V
PCH_DSWVRMEN = 3.24 V
PM_DSW_PWRGD = 3.42 V
PCIE_WAKE_L = 3.33 V
PM_CLKRUN_L = 0 V

There are a few missing that are critical...
 

2informaticos

Administrator
Staff member
If PM_PWRBTN_L reacts triggering SMC_ONOFF_L, then PM_SLP_Sx_L signals should react too.
If not, this looks like PCH issue.

Do you get 1V05_SUS?
 

Fizzl92

Member
If i hold SMC_ONOFF_L shortet PM_PWRBTN_L seays at 0.35 V. Should the sleep signals stay up while i hold SMC_ONOFF_L shorted? Or just pulse? I cant see a reaction of the sleep signals...
PP1V05_SUS is present.
 

2informaticos

Administrator
Staff member
PM_SLP_Sx_l signals should all go high (steady), after pressing power button; same as triggering SMC_ONOFF_L.
You don't need to keep SMC_ONOFF_L low; just a temporary contact.
An o-scope can catch a fast 3V pulse on these signals.
That will mean PCH starts the power sequence.
If not, likely bad PCH.

Did you try to start in SMC bypass mode?
 
Last edited:

Fizzl92

Member
I cant see a pulse on my scope on the SLP_Sx_L signals. Its also not turning on in SMC bypass mode.
I also talked to the customer and he said that the charger was not attached but maybe it was just sleeping not turned off as he changed the cable. That would explain the spark on J4802.
I checked the resistances of all rails as i could fix a few boards by comparing resistance values with other good boards and i think i found some shorts there. I did not checked these rails before as PP5V_S4 was the first missing rail and it is not shorted... The following resistance values seem to low to me:

PP3V3_SUS: 380 Ohm (i would expect kOhms here)
PP3V3_S3: 1.1 Ohm
PP3V3_S0: 1.2 Ohm
PP3V3_S0_FET: 0.8 Ohm
PP1V05_S0: 0.9 Ohm

I am not sure how they all can go bad (maybe the short was on PP3V3_S4_TPAD killing other 3V3 rails?!) and where they all are connected to each other but eg. PP1V05_S0 shorted would mean that the PCH is not turning on and explain why SLP_Sx_L signals are not pulsing.
 

2informaticos

Administrator
Staff member
All these power rails have a common point, which is PCH.
Some of them are low enough to inject voltage.
Remove CPU heatsink and inject on 3V3_S0/SUS; start with 1V and go up if needed.
You should be able to detect if PCH die gets hot.
 
Top