HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

[chbalnuweit]

Hi everyone,

i am facing a strange problem with an HP Elitebook 840 G5. Unfortunatley no board schematics are to be found on this one (COMPUTRO-6050A2945601)

The laptop runs at 0.4 GHz and is always in BD PROCHOT protection. Even in the BIOS it shows 400MHz CPU clock... And same issue when booting a Live Linux. So I am pretty sure I can rule out problems with the OS. It must be lower level.
I've been researching this behavior the past 2 days and tried EVERYTHING I could find on that topic.

What I noticed playing around is that when I disable PROCHOT using ThrottleStop the CPU frequency goes up to 0.8 GHz. If I then disconnect the laptop battery, the CPU frequency is acting normal and goes up to spec (1,9 GHz).

Since CPU temps around 50degC are fine and most likely not the issue, it is more likely that PROCHOT is active due to power (current) limitation, in my opinion.

So I went ahead measuring around the charging controller ISL9538H and found a pin called PROCHOT which according to the datasheet (I could only find datahseet from similar chip ISL95338) is pulled low when overcurrent protection is active. And it measures low indeed, so I am wondering if there is actually something off with the power supply circuit. I tested different chargers and bench power supply; all same behavior.

Now I checked the charger id pin in order to see if the overcurrent protection might be a result of an undetected charger. And that pin is reading only 0.58V when plugged in (around 11V when not plugged in).. I checked on an older Gen 4 Elitebook and the charger id voltage is around 2,6V when plugged in (same charger)

Unfortunatley I cannot find any schematics or voltage readings for this pin. All I could find is a schematic of the Gen 6 Elitebook, which appears to be similar but not the same.

I found one picture online with voltage readings of all coils on this particular board. I checked them all end everything is fine except v_core being low with 0.7V instead 0.9V, since it is throttling.

Is there someone with the same board (or personal experience) who could tell me if PROCHOT (Pin 23 on ISL9538H) is supposed to be low on a good system and if the charger ID pin is supposed to read 0.58V when plugged in?

I really hope someone can pinpoint me in the right direction with this problem.

Thank you all in advance!

I will attach schematics of Gen 6 Elitebook and ISL95338

[MOD EDIT]
Link to schematic - https://www.badcaps.net/forum/showthread.php?t=108193
Reminder: rules and organization of this section - https://www.badcaps.net/forum/showthread.php?t=117483
.

[chbalnuweit]

Re: HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

I found a similar Elitebook 850 Gen 6 from a friend and measured more voltages of a good board:

1. id pin reading of about 0.58V is correct for this board.
2. PROCHOT pin of ISL charging IC reads about 1V on a good system.

I followed the PROCHOT pin (OCP_CHG) from ISL chip and found the schematic (see attachement)

I thought maybe the mosfet is shortet to ground and removed it. But signal was still low.

I ordered a new ISL IC and hope that the IC itself is faulty and pulls the signal low.

I will keep you guys posted.

[reformatt]

Re: HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

PROCHOT is open drain, so needs a pull up. It will be pulled low if the input and output current (measured via CSIN/CSIP and CSON/CSOP pins) is considered excessive. Going off the datasheet, you have series resistors to each pin from Rs1/Rs2. Check these values as they need to precisely match.

Without a schematic, it's also possible that this line is also being pulled low elsewhere on the board if it's using some kind of combinational bus. Depends where the pull up is. PROCHOT can also be asserted by the EC or the CPU itself. You also haven't mentioned whether this is happening with/without battery, charger only, battery only. Sometimes bad batteries can also cause this problem.

[chbalnuweit]

Re: HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

Hi, thank you for your response!

I couldn't measure the resistors since the boardview from Gen6 does not quite match the actual Gen5 board.

However I measured the Signals CSIN and CSIP:

CSIP=19.61V
CSIN=19.13V

CSIP-CSIN = 480mV

Unfortunately I dont have the full datasheet for ISL9538H (which is on my board) but according to similar ISL95338 the delta voltage should be between 10-80mV
If the chips have same spec, it would mean ISL chip measures input current of 24Amps...

Unfortunately thats all assumptions since I can not find proper datasheet of ISL9538H...

I will change the ISL9538H when it arrives and keep you updated.

Thanks!

[mon2]

Re: HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

Remove all power. Meter in lowest resistance mode.

What is the path resistance between CSIP&CSIN? Measure at the charger pins.

[reformatt]

Re: HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

Precisely. ~500mv across across a current sense resistor (which should be in the order of milliohms) is not expected. Way too high.

[chbalnuweit]

Re: HP Elitebook G5 (COMPUTRO-6050A2945601) Power Throttling Issue

So I officially gave up on this one... :-(

While trying to take a few more measurements I must have shorted the ISL chip and maybe more since the charger wasn't being detected anymore suddenly after.

I replaced the ISL chip and the charger would be recognized again but charging light would cut in and out about every 500ms with a noticable ticking sound from the motherboard. Sounded like motherboard going into over current protection or so while trying to charge the battery.

I replaced the battery as well but it didn't change anything.

Laptop would turn on with battery only or charger only, but CPU is still limited to 400MHz...

I'm too much in the dark without proper schematics and boardview.

Luckily I found a cheap donor motherboard that I'm gonna swap.

Thank you guys for your support!

[elbostan]

Hello everyone
I have seen such cases happen a lot in HP, Dell devices I want to mention to you an experience with HP 840 g5 6050a2945601 The device was very bad due to the exaggerated slowness. I decided to follow this fault from the beginning of the charging circuit to the VRM control, passing through the processor H_prochot# In the event of this signal The processor performance is pulled to its lowest levels, which is that it works at the lowest hertz and the lowest number of cores This signal is linked to the charging control and the processor on the Mediol Alermal. It is also linked to the VRM. They all have a pull up resistor with a value of 1 volt. By tracking their resistances, it became clear that there was no damage and the resistors were intact. I was only able to modify the 1 volt supply path to Hprochot# and the device returned to its performance again

[buchoni]

Hello everyone
I have seen such cases happen a lot in HP, Dell devices I want to mention to you an experience with HP 840 g5 6050a2945601 The device was very bad due to the exaggerated slowness. I decided to follow this fault from the beginning of the charging circuit to the VRM control, passing through the processor H_prochot# In the event of this signal The processor performance is pulled to its lowest levels, which is that it works at the lowest hertz and the lowest number of cores This signal is linked to the charging control and the processor on the Mediol Alermal. It is also linked to the VRM. They all have a pull up resistor with a value of 1 volt. By tracking their resistances, it became clear that there was no damage and the resistors were intact. I was only able to modify the 1 volt supply path to Hprochot# and the device returned to its performance again

Hi everybody. I´m gettiing the same problem with my Elitebook 840 G5. CPU stuck at 0.4Ghz. I´ve tried everything, even rollback bios to version 1.25. The issue continues.
¿How can I put the 1v to the Hprochot# in the mainboard, to make the processor run as it should? Thanks.

[cavirules]

Hi! Same problem here. I'm also looking for the way to put 1V to Hprochot#...any idea? Thanks

[mon2]

Are you confident that the CPU is properly heatsinked with enough thermal paste that is not dried up?

What it is:
PROCHOT# is a signal pin on Intel processors that signals to the rest of the system that the CPU has reached its maximum safe operating temperatur

[cavirules]

Yes thermal paste is in good condition and properly heatsinked. Core temperature is always below 50

[cavirules]

buchoni , chbalnuweit same problem than you, same laptop (hp Elitebook 840 G5) processor stuck at 0.4 GHz and is always in BD PROCHOT protection. Finally I could fix the problem by changing ISL9538H chip. It works for me. I hope this post will help you. Thanks

[mon2]

Very nice. Thanks for sharing the resolution.