Hp 840 G5 charging LED white and orange blinking

[questore]

Hi Everyone, I need some help on charity project
My friend gave me 3 of these faulty hp 840 G5 that I intend to repair and give it to an orphanage. (If the project succeeds then i have 3 more of hp 840 g6 with the same intentions)
HISTORY:
All came as no power no previous info
+ I have changed the main intersil chip and shorted mosfets and board shows sign of life now. No short on the board, have ALW 3.3v ,5v, 1.0v and 1.8V.
No reaction to power button. 20V and 0.03Amps drawn from USB-C. Power button white LED flashes for a split second and this repeats about every 3-4 seconds.
+ Main 32Mb BIOS chip is completely missing but I extracted the 16Mb chip from all and they have serial number and model number.
+ I tryed to flash a 16+32Mb from a working laptop (with cleaned ME) but still gave me the same LED error. (Maybe Nuvoton EC has a different version from these chips and that is why the error?)
+ tryed to flash 16Mb chip with a clean empty BIOS extracted from the HP bios installer but Then I got only orange blinking charger LED. Wifi LED on the keyboard if constantly on.

What are your suggestions?
Should i try to create a 32Mb new BIOS from scratch based on the 16Mb chip? How?

[Boardview Bios and Schematic Bot]

Hi Everyone, I need some help on charity project
My friend gave me 3 of these faulty hp 840 G5 that I intend to repair and give it to an orphanage. (If the project succeeds then i have 3 more of hp 840 g6 with the same intentions)
HISTORY:
All came as no power no previous info
+ I have changed the main intersil chip and shorted mosfets and board shows sign of life now. No short on the board, have ALW 3.3v ,5v, 1.0v and 1.8V.
No reaction to power button. 20V and 0.03Amps drawn from USB-C. Power button white LED flashes for a split second and this repeats about every 3-4 seconds.
+ Main 32Mb BIOS chip is completely missing but I extracted the 16Mb chip from all and they have serial number and model number.
+ I tryed to flash a 16+32Mb from a working laptop (with cleaned ME) but still gave me the same LED error. (Maybe Nuvoton EC has a different version from these chips and that is why the error?)
+ tryed to flash 16Mb chip with a clean empty BIOS extracted from the HP bios installer but Then I got only orange blinking charger LED. Wifi LED on the keyboard if constantly on.

What are your suggestions?
Should i try to create a 32Mb new BIOS from scratch based on the 16Mb chip? How?

The main issue here boils down to the HP Sure Start security mechanism. Your approach would be completely fine for standard consumer laptops and would normally solve the issue. However, these enterprise-grade machines are a different beast. Their underlying logic and encryption security are much stricter, so there are a lot more factors to consider when troubleshooting a dead board.

Based on what you've described, here are my conclusions and advice:

1. About the extracted HP update files:
Never extract a .bin file directly from the official HP .exe installer and flash it using a programmer. Those files are merely update payloads, not complete factory firmware dumps (they are missing the ME region and the correct file structure). Doing this essentially bricks the chip. This explains exactly why your board went into a hard halt, showing a solid orange charging LED and a constantly on WiFi LED.

2. About the Nuvoton EC mismatch:
HP tightly binds the security data inside that 16MB chip directly to the Nuvoton EC. If you take a 16MB + 32MB dump extracted straight from another working board and flash it, your current EC will immediately detect a signature mismatch. When this happens, the EC intentionally blocks the normal power sequence, resulting in that blinking white and orange LED error. Taking someone else's already-encrypted BIOS and forcing it onto your board is useless.

What you need to do:
I feel you need to find the actual, exact Motherboard Part Number printed on the PCB, because different part numbers represent different board architectures.

Then, you need a natively matched "clean" 16MB + 32MB firmware dump specifically for your exact board model.

What does "clean" mean? It means the firmware state must be equivalent to a brand-new laptop straight from the factory. Only this virgin state can match your board, allowing the new BIOS to sync with your EC and automatically generate new encryption keys.

You either need to find a pre-cleaned dump for your board number, or ask one of the pros here to "clean" your donor dump by completely resetting the ME region and wiping the Sure Start security block back to a blank, factory-new state.

I hope this explanation makes sense to you. Let me know if you have tried going this route yet.

[questore]

Thanks, I have already found an exact modell number matching working 16Mb+32Mb BIOS.
So here is how i think will proceed: (Please correct me if anyone thinks its not correct)
1. My 16Mb BIOS should be integrated to the modell number matching 32Mb to match serial and etc.
(Replace "BIOS region" block with my 16Mb original Block with software UEFITool 0.21.5)
2. Then I will clean the ME region from 32Mb
3. Reset Sure Start security block (dont know how but i will gather some info about it)

[questore]

I could not clean the Sure Start, don't know how to do that.
I did everything else No luck yet. Maybe i found a wrong board BIOS?
error: CAPSLOCK light is solid on and charging LED is flashing white/orange
Attahed my 16MB BIOS:

[questore]

Meantime I tryed to read the EC from NUVOTON NPCX797HA0BX.
This config should work according to others but no connect:
6050a2945601 KB_pins=42 LEFT DOWN 05MM FLASH=128 JENK=17 TCK=2 TMS=20 TDI=3 TDO=14 CHIP=NPCX797HA0BX
Maybe that means that Nuvoton needs to be replaced?

[mon2]

If your current configuration for the NPCX797HA0BX on the 6050A2945601 (HP EliteBook 840 G5/850 G5) isn't connecting, the issue usually stems from the power state of the board or a mismatch in the JTAG pin mapping.

Troubleshooting Checklist

• Power State: The motherboard must be powered to read the EC. Connect the original AC adapter. If it still doesn't connect, try holding the power button for a few seconds during the "Check" or "Read" process to wake the EC.
• Grounding: Ensure you have a common ground between your programmer (Vertyanov/SVOD) and the motherboard. Using the FPC/FFC cable usually handles this, but a dedicated GND wire to a screw hole can help stability.
• Pin Alignment: Verify that "LEFT" means Pin 1 of your FFC cable aligns with Pin 1 of the keyboard connector. On this specific HP board, "LEFT" or "RIGHT" can be ambiguous depending on board orientation; double-check the small "1" silkscreened near the connector.

Alternative Configuration

Some technicians have reported success with a modified JTAG mapping for this specific HP series. If your current pins fail, try:

• JENK (Reset): 17
• TCK: 2
• TMS: 20
• TDI: 3
• TDO: 14
• Note: If connection persists in failing, try swapping TDI and TDO (some revisions of this board swap these traces).

Hardware Specifics

The 6050A2945601 motherboard is the core of the HP EliteBook 840 G5. Because this EC (Nuvoton NPCX797) is often paired with a sophisticated power management circuit, ensure the CMOS battery is connected or, conversely, try disconnecting it to force a "first-boot" EC state when AC is applied.

Are you using a Vertyanov (JIG) or SVOD programmer, and what error message (e.g., "No connection" or "Unknown ID") is appearing?

[questore]

Thanks for the tips.
- I have checked pin config and ground as always.
- Also after connecting I always check the first, the last and a random pin connection between Verytanov and board.
- Swapping TDI and TDO unfortunately not helped.
- Also connected RTC battery 3v
- Tryed pressing power button.

Here is my EC reading log when trying to connect:

2026. 04. 27. 19:38:09
Vertyanov JIG FW_v51
[AK08TJ77][414B3038544A3737]
[W10326] [573130333236FF151222]
EC RD|WR counters:
ITE:53|31 ENE:4|2 MEC:0|0 NUV:0|0 EPF:0|0
## JIG registered ##
Nuvoton JTAG ID 0xFFFFFFFF

Wrong JTAG ID.
Check the settings and ground!
Motherboard is connected and powered?

[questore]

If Nobody have more idea then I will order a new Nuvoton chip and see if a new one solves the problem.

[mon2]

I would hit up the programmer hardware company to confirm the wiring for this EC. You are not receiving the proper ID through a simple JTAG code extracting. This is either due to:

a) improper wiring of the target device
b) target device is not being powered
c) JTAG is not working properly
d) EC is defective

(in the order of concern). I recall when we built our own JTAG tool which was time consuming but in the end, was able to Read out the vendor specific ID out of a low cost FPGA device.

With JTAG - the programmer supplies the data; clock; reset states. As long as the wiring is ok and the part is not dead, the target should reply back.

Any chance you have a logic analyzer? They are not expensive these days and ~$70 USD can buy a real nice one (we paid $4k USD in Taiwan many years ago and this < $100 unit is better). Sipeed has the unit if you want more details and is based on USB3 front side bus. With the tool you can validate what is really happening over the programmer interface using the JTAG plug-in.

[reformatt]

I did see something on the SVOD ******** group that the JTAG interface doesn't work on a Nuvoton EC if there is a SPI flash connected. Removal of the IC was required apparently.

[questore]

I did see something on the SVOD ******** group that the JTAG interface doesn't work on a Nuvoton EC if there is a SPI flash connected. Removal of the IC was required apparently.

Now that you mention it I remember that some hp BIOS was only readable when moved to an soc8 adapter because with POGO pin reader I could only read a 64kb chip of some kind. I will try it.

[questore]

I would hit up the programmer hardware company to confirm the wiring for this EC. You are not receiving the proper ID through a simple JTAG code extracting. This is either due to:

a) improper wiring of the target device
b) target device is not being powered
c) JTAG is not working properly
d) EC is defective

(in the order of concern). I recall when we built our own JTAG tool which was time consuming but in the end, was able to Read out the vendor specific ID out of a low cost FPGA device.

With JTAG - the programmer supplies the data; clock; reset states. As long as the wiring is ok and the part is not dead, the target should reply back.

Any chance you have a logic analyzer? They are not expensive these days and ~$70 USD can buy a real nice one (we paid $4k USD in Taiwan many years ago and this < $100 unit is better). Sipeed has the unit if you want more details and is based on USB3 front side bus. With the tool you can validate what is really happening over the programmer interface using the JTAG plug-in.

Sorry but dont have a logic analyzer.

[questore]

- tryed removing 32MB BIOS IC - no change
- after that tryed removing 16MB+32MB BIOS IC - no change
Any other ideas?

[mon2]

What was the feedback from Vertyanov factory?

[questore]

What was the feedback from Vertyanov factory?

No, but I have found a technical pdf posted in this forum also. I originaly find this also at vertyanov forums.

More info about my setup: My vertyanov JIG is a clone, but i did not have any problem with it yet.

[questore]

I took the chance to translate Vertyanov pdf to english for others to read easyer.

[questore]

Some forums reported that replacing the EC+3 BIOS chips from an other CPU dead motherboard works, so First I will try replacing only the EC to an empty new one, maybe it will autoprogram itself from 16MB BIOS and try to rebuild 32MB from scratch.
Next thing I will try is to transfer only the EC+16MB BIOS from a CPU dead board and try to rebuild 32MB from scratch.
I will get back with the result.

[questore]

Replced the EC to a new original one did not helped. (Same behaviour)
To.be.continue...

[questore]

This is continues from after replaced the Nuvoton chip to a new one...
replaced tha 16MB BIOS to another from a CPU dead donor. (from now i call this "Donor 16MB BIOS")
Now charging LED blinking continously with orange light only. (no light on the Keyboard, not reacting to power button. )
Also replaced the 131K EC BIOS CHIP from this CPU dead board, still oragange blinking.
Replaced my 32Mb BIOS region to the "Donor 16MB BIOS" and ME cleaned it. Still the same orange blinking.
Any other idea?

[kenkit]

The main issue here boils down to the HP Sure Start security mechanism. Your approach would be completely fine for standard consumer laptops and would normally solve the issue. However, these enterprise-grade machines are a different beast. Their underlying logic and encryption security are much stricter, so there are a lot more factors to consider when troubleshooting a dead board.

Based on what you've described, here are my conclusions and advice:

1. About the extracted HP update files:
Never extract a .bin file directly from the official HP .exe installer and flash it using a programmer. Those files are merely update payloads, not complete factory firmware dumps (they are missing the ME region and the correct file structure). Doing this essentially bricks the chip. This explains exactly why your board went into a hard halt, showing a solid orange charging LED and a constantly on WiFi LED.

2. About the Nuvoton EC mismatch:
HP tightly binds the security data inside that 16MB chip directly to the Nuvoton EC. If you take a 16MB + 32MB dump extracted straight from another working board and flash it, your current EC will immediately detect a signature mismatch. When this happens, the EC intentionally blocks the normal power sequence, resulting in that blinking white and orange LED error. Taking someone else's already-encrypted BIOS and forcing it onto your board is useless.

What you need to do:
I feel you need to find the actual, exact Motherboard Part Number printed on the PCB, because different part numbers represent different board architectures.

Then, you need a natively matched "clean" 16MB + 32MB firmware dump specifically for your exact board model.

What does "clean" mean? It means the firmware state must be equivalent to a brand-new laptop straight from the factory. Only this virgin state can match your board, allowing the new BIOS to sync with your EC and automatically generate new encryption keys.

You either need to find a pre-cleaned dump for your board number, or ask one of the pros here to "clean" your donor dump by completely resetting the ME region and wiping the Sure Start security block back to a blank, factory-new state.

I hope this explanation makes sense to you. Let me know if you have tried going this route yet.

I also have this corporate laptop, I managed to Clean the EC this time I see it trying to boot but it doesn't start I suspect the board has a power related issue, been trying to look for the schematic but only the boardview is available. If anyone has an idea how the power rails on the Coorporate behave I would be happy.
Also I did get a dump file for this machine, and it was even harder to find than than looking for a schematic, I even made an imhex pattern matcher to assist and it's how I managed to find the dump, except it blinks fast as if the bios chip is missing with that dump highlighting the issue of unmatched dumps, I do wonder if reprogramming the 1mb ec bios chip would work ? However finding the third chips and matching all three dumps is impossible. I don't know if wiping that chip will make the ec reprogram it ?