820-3115 must short power pads

Bouba148

New member
Hi Team,

I have a 3115 board that missing S4 state, after following the schematic I've seen that PM_SLP_S5_L was missing, and this one is provided by the PCH.
At this point we now that it could be many things:
I've mesured
PM_BATLOW_L: 3.307V OK
SMC_ADAPTER_EN: 3.446
PPVRTC_G3H: 3.31
Then I checked SMC_ONOFF_L to be sure that it is NOT always low and keeps the SMC in reset mode: solid 3.483V , never pulsing down.
I decided to short the power pads at R5015 or R5016 and suddenly it starts.

WHY?
could some one please explain me the sequence/ conditions for the SMC to start: does it only needs SMC_RESET_L to start or something else? (of course if all the PWR and SLP rails present).
Should I try to put back the board in the MBP and try an SMC + PRAM Reset?
 

Bouba148

New member
After a full night rest , now the board doesn't boot when i short the power pads but only a quarter fan spin.

PPBUS_G3H: 12.60V
PP3V42_G3H: 3.485V
PPVRTC_G3H: 3.313V
PP5V_S5 : 5.020V
PP3V3_S5 : 3.313V
PP5V_SUS : 5.020V
PP3V3_SUS : 3.311V
PP3V3_S4 : 0.40V
PP5V_S3 : 0V
PP3V3_S3 : 0.011V

P3V3S4_EN: 0V
PP3V3_S4_P3V3S4FET : 3.313V

PP3V3_S5_PWRCTL: 3.313V
PM_SLP_S5_L: 0V but pulsing when short power pads to gnd.
SMC_S4_WAKESRC_EN: 0V but pulsing when short power pads to gnd.

ALL_SYS_PWRGD: pulsing

PPVCORE_S0_CPU: 0V resistance to gnd: 1.5 Ohm

then I removed R7510 and mesured the resistance to gnd on both sides:
L7510 side: infinite
C1657 side: 0.8 Ohm
Without R7510 it quarter fan spins in loop.....
So my short to ground is for sure on the CPU side, any clue where to start by as there's a bunch of critical condensators in parallel at he CPU VCORE DECOUPLING level.....
I've removed the 4 easiest and checked : C1680, C1681, C1682, C1683 .....not OK that would be too easy....
Maybe the CPU ?

I've also replaced the bios chip: NOK
Dumped the original one cleaned the ME Region and re writed in: NOK ( Thanks for Duke for his quick tutorial here: https://www.rossmanngroup.com/boards/forum/tools/26816-clean-me-region )

Sometimes it do 2 quarter fan spin, sometimes 1 and sometimes none.

Waiting for your feebacks......thanks in advance.....
 
Last edited:

2informaticos

Administrator
Staff member
There is very low resistance to ground as normal value for i7-3xxx CPU...

L7510 pulses too?
Check R7402/03.
 

Bouba148

New member
OK
at L7510 with R7510 the DVM can barely see the pulse but without R7510 it's 0.6V pulsing to 1.3V
Checked R7402/03: OK
 

Bouba148

New member
Hi, I'm back with that board after several weeks......

So I took again the main voltages:
PPBUS_G3H: 12.30V
PP3V42_G3H: 3.479V
PPVRTC_G3H: 3.313V
PP5V_S5 : 5.018V
PP3V3_S5: 3.313V
PP5V_SUS: 5.015V
PP3V3_SUS: 3.312V

I've found out that R5123 was missing ( Directly attached to Pin 2 of of the U6100) so I've put a new one now the board still have a dimm green light but with PPBUS_G3h and PP3V42 OK

And sometimes it starts with fan at full speed but delayed from the dimm green light from 5 sec, chime, load the OS at almost 75% then stops. The board is NOT heating.
When I mesured SMC_ONOFF_L this one is always 0.108 meaning for me ( correct me please ), that MAYBE the board is always in SMC Bypass mode......
Also the BIL lights stays always off with charged battery, that's why all my suspicious is going to the SMC, please confirm.
Thanks in advance
 

2informaticos

Administrator
Staff member
"PPBUS_G3H: 12.60V
-------------
PPBUS_G3H: 12.30V"
You had correct PPBUS_G3H.
Now you have problem with SM Battery Bus.
SMC works, if board starts (even get image), but that bus can be damaged, or affected by external component.

"SMC_ONOFF_L to be sure that it is NOT always low and keeps the SMC in reset mode: solid 3.483V , never pulsing down
-----------------------------------
SMC_ONOFF_L this one is always 0.108"
Your values change drastically...
Is the board liquid spilled?
Board cannot stay on more than 4s if SMC_ONOFF_L is low.

You alse have SYS_ONEWIRE circuit affected, possibly crap under SMC.
Reball it...
 

Bouba148

New member
SMC_ONOFF_L was shorted to gnd because of the SMC, I replaced it now the short is gone.

PPBUS_G3H: 12.60V green light present

Sometimes it's pulsing once up to CPU_Vcore and when it's not pulsing I have to short the power pads to make it pulse once in both cases after that pulse I'm loosing S5 state, SMC_ADAPTER_EN

It turned ON once with only the battery to Fall full speed then stopped like V or I Sense issue, since then no way to turn it again from battery.

At the pulse SMC_CPU_HI_ISENSE pulses to 1.2V,
Then I've removed R5400, it's rising to 3.3V

Is the CPU dead?
 

Bouba148

New member
Yes of course I've already tried, I didn't post it because it's acting exactly the same : fan pulses at full speed once then stops.
Regarding page 72, sorry but I didn't got you, because everything is pulsing there, and when it's not, it's because PM_SLP_S5_L is missing.
U6200???? are you sure? you didn't meant U7200?

thanks
 

Bouba148

New member
On Pin 1 of U6200 it's also pulsing and there's no short there as in diode mode I'm mesuring 0.497V on pin 1 of C6203.
 

Bouba148

New member
could you please confirm that those 4 pads are not connected any where.......if not that's the reason of my issue....

3115 SMC.jpg
 

Bouba148

New member
reballed SMC same behaviour:
sometimes quater fan spin and everything pulses or no quater fan spin but I have to short the pads to have the pulse and quarter fan spin after what I'm loosing S5 state because of No SMC_PM_G2_EN.
In that "mode" the battery is not charging because CHGR_BGATE is high.
If not it is charging and I have activity on SDA and SCL lines
Re-writed EFI U6100 SMC with a fresh clean dump in which I've injected the Serial into it and did a checksum CRC with the medusa: same issue.

I have to admit that I'm out of idea ......any clue?
 

2informaticos

Administrator
Staff member
SMC_PM_G2_EN signal lost after reballed SMC?
Change SMC now and inspect traces and pads under SMC, ones removed.
Also check for knocked off components.
 

Bouba148

New member
Hi 2 Informaticos,

SMC_PM_G2_EN is present but it's disapearing after quarter fan spin or power pads being pressed....so I follwed the signal up to SMC and SMC_PM_G2_EN.
Question: Is there any signal/rail upstream before SMC that triggers the SMC_PM_G2_EN? in other words what signal, the SMC needs to generate SMC_PM_G2_EN?
 

2informaticos

Administrator
Staff member
Looks like SMC detects a problem and pull down the signal, to block the 3V3_S5 power supply.
Try starting in SMC bypass mode first.
 
Top