I have a minicomputer with an Intel N5095 processor, and BIOS support isn’t officially available. I recently used BGA Workbench to replace it with an Intel N6005. Before doing so, I verified that both CPUs had the same CPUID (06C0) and socket (FCBGA1338), and confirmed that the motherboard BIOS contained the necessary microcode. However, after the replacement, the computer’s boot screen remains dark. The CPU fan runs normally, and the CPU gets hot,the mouse didn’t light up either…
I suspected a BIOS issue, so I tried adding microcode, replacing GOP-related modules, and updating ME using a CH341A programmer, but none of these solutions worked. Since the motherboard lacks expansion slots, I can’t use a debug card to trace the issue. Given that the CPU seems to be functioning and ME settings are being recognized, I assume the hardware is likely fine.
Now I can’t think of any other possible part that could be causing this problem, can someone help me?Any advice or assistance would be greatly appreciated.
CPU getting hotter doesn’t mean its fully connect, you know for sure that a tiny BGA ball not entirely soldered can lead to this, VRM area not affected during re-balling?!?! No schematic/brdview for circuit diag?!?!
USB ports not getting power that’s pré-post processing of bios… so no CPU-PCH die/EC initialization, that needs cpu.
I don’t see mcode, ME or GOP as issue, Jasper Lake gen is equal to both
My POV only, wait for other users.
Thank you for your reply. the engineer who helped me with BGA do the rework has many years of experience and has successfully used this method to replace the CPU of many notebooks, I think there’s a good chance I can trust him.After replacing the CPU, we measured the voltage in various parts of the motherboard, and it was very strange that it worked normally from the voltage point of view.
I’m sorry I don’t have a schematic/brdview for circuit diag, because the manufacturer of my computer doesn’t intend to support it.So I can only guess to locate the problem.
I have learned that N6005 has Integrated Wireless, but N5095 does not have this part. Do you think it may affect the startup of CPU?
Thanks again for your reply and hope you have a nice day!
In that case, i’m forced to trust it also…being an experienced reworked.
Out of ideias then…im not seeing a black_listed specific cpu model in the OEM bios code.
When modding standard Skylake and Kaby Lake motherboards for accepting Coffee Lake CPUs (particularly Xeons), you should disable the Intel ME by setting the HAP bit. I have no experience with the CPUs you mention, but I’d still try to flash an image with ME disabled.
Thank you for your reply. For the situation you mentioned, I tried to disable the Intel ME not long ago, but at that time I used another method on the network:The FIT setting Intel ME Kernel-reserved-yes handles ME, which it claims has the same effect as setting HAP bits.
Seriously, I feel that I may have been misled in this regard, thanks to your reminder, I will try another way to setting HAP bits disable the Intel ME .But since I didn’t find any information about the HAP bit offset for intel ME 13.5, I figured it might take a long time.
Thank you again!
I’ve just disabled ME on my Coffee Lake laptop (Lenovo Ideapad Gaming 3i, i5-10300H, ME 14), using a SPI programmer, though. I reported my experience here. Check that and the linked post, it might be useful.
Your experience helped me a lot. I was looking for a me_cleaner that could handle jasper lake, but I didn’t find any that could perform.I used the steps mentioned in your experience to deal with my bios, but unfortunately, after I flashed it, the computer did not react after pressing the power button except the LED light of the motherboard, the fan did not turn, and it would automatically shut down after a short time.
The changes to the bios are in effect, I think the reason for this may be that my motherboard must have ME in order to boot properly.Or is it because the HAP bits corresponding to jasper lake are different?
Anyway, this is a very meaningful attempt. Thank you again for your experience, which provides me with new ideas.
Oh, though the system is behaving in a different way now, it seems.
At least on my machine (chipset HM470, ME 14), the HAP bit is set in the flash descriptor area. When looking at the stock and the modded flash descriptors side by side, these are the changes:
The first four squares represent the access level for each regions (BIOS, ME, GbE and EC), which I modded to “FF” to allow read/write access to all of them (as described here).
Then, the fifth square is the HAP bit.
It seems the format of the flash descriptor varies depending on ME version. From ME 11 onwards, it’s like mine. Older versions are different. But you say you’re using ME 13.5, so it should be like mine.
I checked and compared it to my backup bios and it looks like the HAP bit is fine, maybe I should modify the Intel Flash descriptor read/write access as you suggested, but they don’t seem to need any additional modifications…I will try and hope it works.Thank you for your advice!
