HP EliteBook 840 G7 not powering on

[bert-bert]

Hello,

I have a HP EliteBook 840 G7 that does not power on.

I connect a power supply, for these tests a monitor via USB-C that powers the mainboard. I had RAM, speaker connected, but this does not really make a difference. Most of the time just the mainboard with heatsink and the USB-C power.

The charging light lights up in white for about a second and turns off afterwards. I can see that the 32MB and the 16MB BIOS chip are both read with my scope. There is very little activity on the 1MB chip. After the charging light goes out there is no activity on the serial EE memory chips. This all seems ok.

The main CPU voltage stays at 0. I have a second mainboard of the same type. There the CPU voltage I just measured was 830mV. So I connected about 0.65V to the CPU rail of the defective board some time ago and something drew about 1/4W, but no part got hot on a thermal camera. All the other voltage rails that I measured were just the same on the good and the bad board.

On the good board without RAM the speaker makes noise. This does not happen on the bad board.

Both boards have some 10th gen i5 CPU.
The bad board has the serial number CT:PJSQ00WBE95J7

I know, I give you very little specification of the board, but I believe, everything relevant is here. In case you need more, I'll be glad to tell you the missing information. Shall I shoot a photo of the voltage regulator section next to the CPU?

I know nothing about the history of that board. I took it out of a defective laptop that I got as defective.

Thank you in advance for any help!

[mon2]

What voltage and current draw do you see on the usb c meter?

Any available schematics?

[bert-bert]

I have no USB C meter, but I just powered through a 230V wattmeter. For one moment I measure 1.1W and it drops to 0.7W constant consumption. This power supply measures 0 idle power without the mainboard. And there are 20V on the mainboard somewhere.

I do not have schematics and do not know whether they are available.

I'll search for schematics after I get some sleep. It is late in the night here.

[mcplslg123]

Measure resistance to GND on all big grey coils(just put black probe to GND and red on target coils)

Measure voltage,after switching, on all coils. Lets see what is missing.

Since no schematic is linked, take a clear photo of the motherboard and just write resistance to GND values in one pic and voltage on another same pic.

That will give us some idea for further steps.

[bert-bert]

1: 65R, 0V
2: 8R, 0V
3: 33k, 1.8V
4: 6R, 0V
5: 5R, fluctuating up to 3.3V
6: 5R, fluctuating up to 2.8V
7: 19R, fluctuating up to 4V
8: 172R, 1V
9: 177R, 1.4V
10: 69k, 5V
11: 102k, 4.3V
12: 10k, 3.4V
13: 66k, 5.1V

Upon measuring 7, I short circuit the coil with some 13V capacitor. There was a small spark.

Shorting the measurement leads measures 3R.

[mon2]

Upon measuring 7, I short circuit the coil with some 13V capacitor. There was a small spark.

This was extremely dangerous as the nearby part under the heatsink is the CPU. The CPU is not able to handle such a high voltage. Remove all power. No battery. Meter in resistance mode.

What is the resistance to ground on either side of inductor #7? Better to test the side that is towards the CPU / black heatsink.

[bert-bert]

This was extremely dangerous as the nearby part under the heatsink is the CPU. The CPU is not able to handle such a high voltage. Remove all power. No battery. Meter in resistance mode.

What is the resistance to ground on either side of inductor #7? Better to test the side that is towards the CPU / black heatsink.

I did not intend to short that, this was inadvertently by mistake. Somehow my measurement leads make so bad contact with the coils that I hardly measure anything. I see you measure the same voltage on both sides of the coil, but beforhand I measured rubbish and had to measure both sides.
I now measure 21R on both sides of 7. This means here nothing has changed, as my DMM is not so precise, but nevertheless it may have killed the CPU. I shall repair that DMM with contact cleaner spray some day.

I again risked measuring with voltage and used a different DMM and get 0V for 4 to 7 inclusive. With the better DMM I measure 16R from both sides of 7 to GND.

I do not like that there is no voltage on 4 to 7 and believe 4.3V for 11 is much to little. On my other board there are roughly 13V. But there is a capacitor next to 11 that measures 13V on the defective board. How can that be?

What shall I do next?

[bert-bert]

For me, the CPU does not seem shorted, but seems to get no power. Do you also think that way?

Can it be a defective BIOS? Do you know how the mainboard acts like when the BIOS is defective?

Upon fixing a password-locked and then defective BIOS on an other model, I know that when the BIOS is defective, mostly it gets read over and over again, but I do not see that here. All the EEPROMs get read and get silent once the charging LED goes off. Very similar to my working 840 G7.

Do I need to reprogram both the 32MB and the 16MB chip when I want to try a different BIOS. i.e. will it boot and show an error message when 32MB and 16MB chip do not fit together or will it just stay silent?

Thank you for your help and thank you in advance for further help!

[bert-bert]

Hello,

I did some more tests. I tried a different BIOS, that changed the charging LED, but only temporarily, so this seems not to be the issue. Everything that follows is measured with the "good" DMM on the coils with the respective number in the picture above.

And I did even risk measuring my good board with voltage again.

4 to 6 measure 0.8V on the good board, but 0 on the bad.
I guess 4 to 6 will be the main CPU power supplies.
4 to 6 measure 3.5R on the bad board to ground and 5R on the good board.
There is 1V at 7 on the good board whereas 0 on the bad.
And at 7 we have 30R to ground on the good board in contrast to 16R on the bad one.
As mentioned above I already did inject some (below 1) voltage in the main CPU power rail at the cap and there was little current, but nothing got hot.

So it can only be the CPU, can't it?
What do you think?
Can I test some more before I give up this board?

But there is something more I noticed, when I look at pin 2 of the main 32MB BIOS chip, the data out pin, with a scope, I seem to see little more activity on the good board after connecting to power than on the bad board, but I may be wrong. None of the boards have any battery connected, so they immediately (try to) power on when USB-C power is available.

Thank you in advance for any further help!