Thinkpad E14 Gen 2 NM-C771 - No display no fan no beep no boot

[ongobongo]

Hi there,
I got this laptop which was not working at all. When attached to an amp-meter, the meter wasn't working at all also. I quickly found out that the PD-Power switch was gone and fully shorting to ground. After replacing that (NX20P5090) I got live back into the laptop.
Now the battery is charging, charge indicator is working and the power led is blinking 3 times when i attach a charger. but, unfortunately, this is it. When i push the power-button, the power led turns on, but that's it. No display, no beep, no fan spin, nothing (but cpu gets warm).
I tried flashing a new bios with a CH314a, but that didn't work, so i re-flashed the odl bios that i dumped.
After some reading I'd guess there is a problem with the EC, but I'm totally new to that stuff and don't know how to deal/diagnose that.
Any suggestions?

Thanks a lot
ongobongo

[ongobongo]

I dug a littler deeper and found that VCC on the EC is missing, which is +3VS according to the schematics.
In fact all the VS lines are missing: +3VS, +5VS, +1,8VS and +0.75VDDP are not present while the ALW equivalents are there
As i see it, the mosfets controling the gate for these lines (Q30 and Q24) work, SUSP is created (5,1V) and V9B+ is also there.
Can it really be that both fets are gone?

[mon2]

See attached. Measure the voltage to ground of Q31. Post each measurement.

[ongobongo]

Pins 1-3 = 0V
Pin 4 = 1mV
Pins 5-8 = 3,3V

[ongobongo]

hmm i just realized that the board draws 0,6A once it is attached to the charger and the CPU is getting hot quickly.
I don't think this is a good sign. Boards dead maybe or could this be another sign of a dead EC?

[mon2]

Without a heatsink, all CPUs will get hot quickly but remove all power. Meter in resistance mode. Measure the resistance to ground of the CPU power rail which should be a multiphase circuit. If the resistance is too low then the CPU is truly dead.

[ongobongo]

What i meant to say is that the CPU gets hot as soon as the charger is connected. no need to press power button, it gets hot right away (no battery is attached, only running on charger)
With regards to the power rails:
VDDCR_SOC_PH1: 0,9 ohm
VDDCR_CPU_PH1: 3,6 ohm
VDDCR_CPU_PH2: 3,6 ohm
VDDCR_CPU_PH3: 3,6 ohm

[mon2]

Yes, it may be dead. What you can do here is remove all power. Meter in resistance mode. Place one meter probe on the main power rail at the charger circuit. Other meter probe at the voltage rails you just posted (ie. VDDCR_SOC_PH1). If there is a low resistance between these rails then the PCH / CPU is dead. This is from a high side mosfet leak which will have forced the high voltage rail to be injected onto the very low voltage PCH / CPU power rail.

[ongobongo]

Tested the rails above against 20V, 5V and 3,3v and they look good, no shorts or low ohms.
Against 20V I measured between 20k-100k ohms, depending on the rail.

[ongobongo]

I think I found the current drawing problem. PU4 aka lv5075b gets pretty hot as soon as the charger is attached. I'll exchange that one and report if I finally get a boot or if more problems occur down the line.
Thanks so far
ongobong

[trepachka]

CPU getting hot immediately in this machine is normal according to my experience. It is also normal for the BQ chip as well as QB6 mosfet to get really hot, sometimes going to 90c+ within seconds. Do you have an IR camera?

PU4 getting hot does not mean it's faulty. There may be other components influencing it to get hot. Make sure at least you get 20v, 5v, 3.3v and 1.8V rails available.

Measure the capacitors nears the EC chip. I've had cases I was getting all these rails but the computer would just refuse to start. Then, I'd find several shorted capacitors near the EC chip. Injecting voltage would light up the EC chip like a x-mas tree.

I also noticed that some bare boards do not necessarily kick start when power is applied. You need to attach the daughter board thru the ribbon cable and then push the power button to make it run.

[ongobongo]

Yes, I have an IR camera to spot hotspots.
When I attach the charger (without battery) I get 20V, 9V, 5V, 3,3V and 0,9V (which should be 0,75V VDDP_P if I'm not wrong). I do not get 1,2V.
I measure at PL7 and get 0,9V while the CPU and PU4 are heating up quickly. If i disconnect the charger, PL7 is shorted to ground with 0,6ohm.
I desoldered PJ8 to check which side if shorted to ground and you are right, PU4 seems to be fine, part of PJ8 going to PU4 is not shorted it's about 100 Ohm.
The other side of PJ8 leading to Q22 is short. I injected 0,75V to that side and voltage drops to 0,4/0,5 and current caps at limit.
Unfortunately, I don't see any hotspot, even with 2,5A current limit. According to the schematics, Q22 more or less only goes to CPU, so I fear this is it with the repair.
But if you have any other ideas, I'm happy for everything.
Thanks
Ongo

[rumpumpel1]

The basic question is: how can I tell the difference between a running CPU and a shorted CPU? I asked this question on this forum some time ago, but didn't get an answer. My method is to probe the CLK pin of the BIOS chip with an oscilloscope. If there is a lot of activity then the CPU is running, otherwise the CPU is dead.

[trepachka]

I think PL7 is supposed to measure 1.8V.

With multimeter in resistance mode, place ground probe anywhere on ground and the other probe on coils PL13 and PL8 and report their resistance.

[ongobongo]

Now that I have desoldered PJ8, current draw after attaching charger goes down to 0,16A and nothing is heating up.
With this, I have 0,7ohm resistance to groound on PL8 and 3,5ohm on PL13.

[ongobongo]

I tried to inject some volateg into PL8 to find the short but couldn't find any component heating up, upped the current to 2,7A and I was able to figure out the area that's getting hot.
Unfortunately, it is around the lower left corner of the die :-(
I attach a picture of my IR cam.
I have no experience with dead CPUs but from what I saw in a couple of videos, this is what it looks like, although I am open to other opinions

[trepachka]

PL13 should measure around 20 ohms and PL8 should measure around 100 ohms. Based on what you showed here and the heat signature, you have a shorted CPU, I am afraid.