For testing, to see if this causes the memory issue, can you borrow some other memory so you can run matched memory in this system. I know you said previous issue, but now with fixed BIOS test again if you can.
If always this error, there may be missing or damaged tiny 1-2mm resistor somewhere in the memory components. Or as mentioned, some other damage somewhere when the board/BIOS was shorted at initial point of problem.

Please test, does this BIOS now remove the last (second) exit?
https://www.sendspace.com/file/2t05lb

Still, none of the new menu options help you get QST working? If no, next we can update ME version, but I don't think that will help since this version we use now used to work on this system.

Hello, I tested [MESwap 01 - ToQSTV 2 - BCPM - exittest.rom] immediately. But the second 'termination' was not deleted
"ME Error: No Channel 0 Memory; disabling HECI messaging" error is always present.

I have a request
Could you clean the ME region of [F02_FMV - D5380-2.bin] and [F03_FMV - D5370.bin] in [FMVESPRIMO.ZIP]?
I would like to initialize the repository and do cleaning without doing any special settings.
Of course I plan to use it for FMV-D 5380 NO.2 (machine 2), FMV-D 5370 (machine 3). Also, a problematic system, FMV-D 5380 NO.1 (Machine 1)
But it seems to be available. Even if it can not be used temporarily, clues may be obtained.

My current view
Currently I am looking for a short circuited resistor or capacitor using a tester. Because small resistors and capacitors, or SSOP or TQFP
Package IC can be exchanged for me. But I can not replace IC in BGA package. Therefore, failure of MCH or ICH is obvious
In that case, unfortunately we are going to give up this system any longer. Also, you can not exchange the motherboard.
This is because these systems FMV - D 5370 or FMV - D 5380 are not very available because of low production volume. In addition, manufacturers discard all inventory parts
(Japanese manufacturers never sell parts to users, it is awful). Therefore, I will do all I can to my self.
The situation at the time of short circuit of the EEPROM was fortunate to be noticed in a short time, and it did not result in complete burnout. Although the data seems to be broken,
Short circuited EEPROM (MX 25 L 3206) It itself has been cooled sufficiently, passed erase, blank check and can be reused.

There is no circuit diagram of the motherboard, but the method by measurement with the tester is not accurate, but it can be read about short circuit or disconnection.
Generally, it seems that overvoltage was applied to the line branched from EEPROM to ICH 10D.
It is hard for me to think that it breaks extensively from the MCH to the memory connector. If there is such a thing, it is somewhat
I think that the impact will occur. However, recognition and operation of the memory itself does not affect AMIBIOS, Windows, Linux (UBUNTU). Only for ME subsystem
It seems to be affected.
If you have something you know about repairing from motherboard damage in a similar model (such as Dell Optiplex 780) please let me know.

Additional notes
In the system made by Fujitsu Electric, we have found an example that "ME Error: No Channel 0 Memory; disabling HECI messaging" error is not limited to memory problems, but also caused by BIOS defects.
Although the model is described as FMV-DESKPOWER CE 55X / 50X series, LX 70X / 65X / 55X / 50X / 45X / 40X series, these structures are very similar to the FMV-ESPRIMO series (FMV-D 5380 etc.).
http://azby.fmworld.net/app/customer/dri...Number=E1006819

Second "Exit" was not removed? If so, OK, please clarify if that is what you meant, and I will edit again (Trying to find last entry, is why I wanted to remove second exit first as a test, then I will put it back and remove the other one with less options)

On the cleaning of ME. Did you already try that yourself and fail or cause issue? I ask because we have very detailed step-by-step guide here to do that, so you can do yourself if you follow the guide.
If it's too much detail for you to follow, or you don't understand that's OK and I understand and can do for you, but would rather you do yourself I you can because it's easy to follow guide and clean ME with the guide.

Maybe some part of ICH was burned through in the silicon? I know some parts of ICH or PCH can burn/break and chip still then function for all other areas (Due to how the B3 P67 chipset bug was found and what happens when it burns through)

"BIOS Defects" to some regarding the memory error may be all they understand, so they call it BIOS defect or BIOS issue etc, but in this instance it's ME which is part of the BIOS giving that error either due to ME setting/corruption, or actual missing
memory stick or memory detection due to other physical problem in the memory detection process.

I wonder if ME 5 can be updated to ME 6 to get around this problem @plutomaniac?

Thank you for your reply.
The second "exit" in the BIOS menu is not used on a daily basis, but there is a possibility to use it in the future, so you do not need to erase it.
What I am trying to do is prepared to be a very difficult repair work that is unprecedented in the past (at least not in Japan).

I have already failed to clean up the ME. The cleaning method of ME I went was [Guide] Clean Dumped Intel Engine (CS)ME/(CS)TXE Regions with Data Initialization
Or the method described in https://www.youtube.com/watch?time_continue=1&v=9Wli7_9v7CM movie. I think that these methods are probably inappropriate for Fujitsu Electric's system.
Besides that, please let me know if there is the most reliable method.

As I mentioned earlier, I already tried the following things although the sentences will be long.
Building the BIOS of [F01_FMV-D5380-2.bin] + "5.0.3.1126_DO_PRD_EXTR.bin" using FITC.EXE, FMV-D 5380-NO.1 (Machine 1)
Memory errors and ME subsystem problems are all resolved. However, the operation of IntelQST becomes abnormal only.
Therefore, in order to confirm the success of ME cleaning, I did the same work in FMV - D 5380 NO. 2 (Machine 2) and FMV - D 5370 (Machine 3), without suspected burnout on the mother board.
Even in these cases, there is no problem in the memory and the ME subsystem, and only the operation of IntelQST is abnormal.

[F01_FMV-D5380-1.bin] + "5.0.3.1126_DO_PRD_EXTR.bin" → FMV-D5380. NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST: NO
[F01_FMV-D5380-1.bin] + "5.0.3.1126_DO_PRD_EXTR.bin" → FMV-D5380. NO.2 (Machine 2) → Memory: OK, ME: OK, IntelQST: NO
[F01_FMV-D5380-1.bin] + "5.0.3.1126_DO_PRD_EXTR.bin" → FMV-D5370 (Machine 3) → Memory: OK, ME: OK, IntelQST: NO

[F02_FMV-D5380-2. bin] + "5.0.3.1126_DO_PRD_EXTR.bin" → FMV-D5380. NO.2 (Machine 2) → Memory: OK, ME: OK, IntelQST: NO
[F03_FMV-D 5370.bin] + "5.0.3.1126_DO_PRD_EXTR.bin" → FMV-D5370 (Machine 3) → Memory: OK, ME: OK, IntelQST: NO

So I tried to clean the ME region manufactured by FUJITSU ELECTRIC CORPORATION in detail. (Japanese users are not interested in BIOSMOD, there is no information on ME Firmware at all)
I was fumbling, I did a lot of experiments to erase and start data in the ME region one by one on HexEditter, but I can not solve it.
In the enumeration of data in the ME region, no rule was found and the data changed or moved every time the system was started, and nothing was possible.
Another possible thing is to disassemble ME Firmware using IDApro, but we do not yet know how to use IDApro and how to decode disassembly.
Old version of [IDApro Ver 5.0] It owns it. I need to learn from now.

Based on these points, I made a hypothesis.
FMV-D 5380. NO.1 (Machine 1) may have burned IC and resistance, but there is a high possibility of a ME region problem.
The ME region in the system made by Fujitsu Electric is not limited to the version, but is different from the ME region made by another manufacturer. Especially IntelQST is a special specification.
[F01_FMV-D5380-1. bin] [F02_FMV-D5380-2. bin] [F03_FMV-D5370.Bin] Looking at these binaries on HexEditter, we can find the section labeled [QstParams.NVKR]or[QstConfig.NVKR].
But I do not know how to edit or create new [QstConfig.NVKR], or the software for that.
However, there is a possibility of extracting special specification data which is indispensable for Fujitsu Electric's system based on this part.
For example, [FI 01 _ FMV - D 5380 - 1.bin] + '5.0.3.1126 _DO_PRD_EXTR.bin' + 'QstConfig.bin' in FITC.exe with the extracted data as 'QstConfig.bin'
It is thought that if you build as BIOS you can produce BIOS suitable for Fujitsu Electric's system.
However, however, I do not know the address of data to be specifically extracted and the processing after extraction.

Today, after doing some experiments again, there was a big discovery.

1. [F01_FMV-D5380.bin] + [5.2.0.1009_DO_PRD_RGN.bin] → FMV-D 5380. NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST: NO
2. [F01_FMV-D 5380.bin] + [5.2.10.1023_DO_PRD_EXTR.bin] → FMV-D 5380. NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST: What ???
3. [F01_FMV-D5380.bin] + [5.2.20.1030_DO_PRD_EXTR.bin] → FMV-D 5380. NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST: NO
4. [F01_FMV-D5380.bin] + [5.2.40.1037_DO_PRD_EXTR.bin] → FMV-D 5380. NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST: NO

We confirmed that only the second of these BIOS binaries has behavior different from the past.
There was no problem with memory and ME subsystem, and it was confirmed that the intelQST function started partially in normal operation.
The CPU fan runaway stops and the MCH sensor recovers. However, other multiple temperature sensors are not detected. Only the power supply fan is running out of control.
When Everest and Speedfan were started from Windows, it appeared a bit different from the display before the problem occurred, but speed control was possible only for CPUFAN.
Considering these, I think that some setting related to IntelQst exists beforehand in binary [5.2.10.1023 _ DO_PRD_EXTR.bin].
Looking at binary [5.2.10.1023_DO_PRD_EXTR.bin] on HexEditter, we could confirm the existence of the section called [QstConfig.NVKR].
Next it was ascertained whether this phenomenon could also occur in burnout-free normal systems FMV - D 5380 no. 2 (machine 2) and FMV - D 5370 (machine 3).
As a result, I confirmed the exact same phenomenon.

[F01_FMV-D5380-1.bin] + [5.2.10. 1023_DO_PRD_EXTR.bin] → FMV-D5380. NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST: What ???
[F01_FMV-D5380-1.bin] + [5.2.10. 1023_DO_PRD_EXTR.bin] → FMV-D5380. NO.2 (Machine 2) → Memory: OK, ME: OK, IntelQST: What ???
[F01_FMV-D5380-1.bin] + [5.2.10. 1023_DO_PRD_EXTR.bin] → FMV-D5370 (Machine 3) → Memory: OK, ME: OK, IntelQST: What ???
[F02_FMV-D5380-2.bin] + [5.2.10. 1023_DO_PRD_EXTR.bin] → FMV-D5380. NO.2 (Machine 2) → Memory: OK, ME: OK, IntelQST: What ???
[F03_FMV-D5370.bin] + [5.2.10. 1023_DO_PRD_EXTR.bin] → FMV-D5370 (Machine 3) → Memory: OK, ME: OK, IntelQST: What ???

If necessary, upload photos and log files, binary files. please let me know.

ME Version 6 is said to be dedicated to the intel-6 series chipset. Does it work well with the intel-4 series (Q45 and ICH10D)?
However, I think IntelQST must be Fujitsu Electric specifications.

Well my point was to erase it to find which 01 enables it, to find the end of the menu string. Then I would put it back, and remove the other limited exit instead, leaving you the better exit. So point on that is only to find where the 01 01 01 etc stops for the menu tabs

Yes, I would follow the cleaning ME guide you mentioned too, first link. I already did that for you in several BIOS, no change you said. Intel QST has settings in the ME while cleaning, you can further enable QST options.

System1 ME is fine, in all BIOS I returned to you for testing, ME cleaned and settings transferred in some cases too, all good but no change on your end - due to system one physical damage I think.

Fujitsu does not make any ME specifically, only Intel makes ME. You, or anyone else like Fujitsu, or Dell or HP etc, can put settings in ME, that is all you can do. 100's of settings to change

If you want some specific QST.bin file put in ME, you have to give me that file. Yes, Fujitsu can make a QST file to go into ME @plutomaniac may be able to tell you how to extract the QST from a working stock dump, I am not sure.

To @Lost_N_BIOS
I looked up the second "exit" menu again. However, whichever item you select, you will reach "Save settings and exit" (Save Restart Computer). This can not be used.
Optionally, to display hidden menus and delete unnecessary menus Where do you change the AMIBIOS address? AMIBCP 4.55 was not available for these purposes.

Additional information on ME ERROR
Apparently when it is a dirty ME region, "ME Error: No Channel 0 Memory; disabling HECI messaging" error is displayed even though there is no problem with memory, IC, resistance etc.
For example.
[F02 _ FMV - D 5380 - 2. bin] → FMV - D 5370 (Machine 3) → Memory: No, ME: No, IntelQST: No
[F 03 _ FMV - D 5370. bin] → FMV - D 5380. NO.2 (Machine 2) → Memory: No, ME: No, IntelQST: No
In problematic FMV-D 5380.NO.1 (Machine 1), I think that the ME region got so dirty due to ROM burnout. Maybe CRC error, or digital signature is invalid.

Please tell me @plutomaniac
Please tell me if you know the method of extracting the Intel Qst configuration file [Configfile.BIN] from the BIOS SPI-ROM dump and using it.
I will incorporate the configuration file [Configfile.BIN] with FITC.EXE with a clear ME Region.

The dump file is attached to the file
FMVESPRIMO.ZIP. (Download URL = https://drive.google.com/file/d/1ld-LSmz...7-gYduOuHC/view)
Included in [F01_FMV-D5380-1.bin] [F02_FMV-D5380-2.bin] [F03_FMV-D5370.bin]

Briefly explaining the history of the past,
ME Region problem occurred in Fujitsu Electric Corporation's system [FMV-D5380] in Japan, trying to clear the ME region, it succeeded, but intel QST is unusual behavior only.
The system of this manufacturer is generally special, apparently it seems useless unless the manufacturer's Intel QST setting exists with the ME region.
About 3 binaries.
[F01_FMV-D5380-1.bin] is a dump of the machine causing the trouble, [F02_FMV-D5380-2.bin] is a machine dump of the same model, and [F03_FMV-D5370.bin] is a similar machine dump.
However [F01_FMV-D5380-1.bin] is suspected of the ME region itself being broken, there is a possibility that it can not be used.
What I think is that looking at these binaries on hex Editor looks like the proprietary Intel QST settings around the string [QstParams.NVKR].
Also, please let me know if you know the software to create or edit a new Intel Qst configuration file [Conigfile.BIN].

to @plutomaniac
First, the defective system is FMV-D5380 NO.1 (Machine1). The system without problems is FMV-D5380 NO.2 (Machine 2) and FMV-D5370 (Machine 3). In case
Taking the complete SPI dump of a healthy system and flushing it with a defective system without any modification to it does not solve the problem as follows

FMV - D 5380 NO.2 (Machine 2) [F02_FMV-D5380-2.bin] → FMV-D5380 NO.1 (Machine 1) → Memory: NG, ME: No, IntelQST: NG
FMV - D 5370 (Machine 3) [F03_FMV-D5370.Bin] → FMV-D5380 NO.1 (Machine 1) → Memory: NG, ME: NG, IntelQST: NG

However, these problems are known to be caused by dirty ME regions, so [Guide] Clean Dumped Intel Engine (CS)ME/(CS)TXE Regions with Data Initialization
Or replaced it with a clean ME region using FITC.EXE according to the method described in the movie of https://www.youtube.com/watch?time_continue=1&v=9Wli7_9v7CM. Then it becomes as follows.

FMV - D 5380 NO.2 (machine 2) [F02_FMV-D5380-2.bin] + [5.0.3.1126_DO_PRD_EXTR.bin] → FMV-D5380 NO 1 (machine 1) → memory: ok, ME: ok, IntelQST: NG
FMV - D 5370 (Machine 3) [F 03_FMV-D5370.Bin] + [5.0.3.1126_DO_PRD_EXTR.bin] → FMV-D5380 NO 1 (Machine 1) → Memory: ok, ME: ok, IntelQST: NG

Also, if you do the following method, IntelQst will recover. But incomplete. It seems that [5.2.10.1023_DO_PRD_EXTR.bin] contains IntelQST settings for another model.

FMV - D 5380 NO.2 (machine 2) [F02_FMV-D5380-2.bin] + [5.2.0.1023 _ DO_PRD_EXTR.bin] → FMV - D 5380 NO.1 (Machine 1) → Memory: OK, ME: OK, IntelQST:incomplete
FMV-D 5370 (Machine 3) [F03_FMV-D 5370.bin] + [5.2.10.1023_ DO_PRD_EXTR.bin] → FMV-D 5380 NO.1→ Memory:OK,ME:OK, IntelQST:incomplete

Describing these incomplete IntelQst,
Control of the CPU fan recovers, but control of the power supply fan does not recover. MCH and CPU temperature sensor will recover. However, ICH and motherboard temperature sensors do not recover.
It seems to misunderstand CPU temperature and MCH temperature on the screen. It also seems that the MCH temperature is abnormally high.
These were confirmed with SpeedFan version 4.49 on Windows 7 32 bit in the defective system FMV-D 5380 NO. 1 (Machine 1).
The same symptoms were reproduced in these BIOS binaries even for systems without problems, FMV-D 5380 No. 2 (machine 2) and FMV-D5370 (machine 3).

As a result of the above, we have determined that it is necessary to load the Qst configuration file [QSTConfig.BIN] suitable for Fujitsu Electric's system.
If necessary, take a picture of the screen and upload it. Please notify me
You can also upload BIOS binaries, such as [F02_FMV-D5380-2.bin] + [5.2.10.1023_DO_PRD_EXTR.bin] etc).

On the exit menu, that is probably because I only "Enabled" the menu for you to see it as a test, and did not enable any settings inside there yet.
This is done in a Boolean string edit on the setup module via hex edit, then once main "menu item" is enabled you can adjust permission to User in AMIBCP and it should work, if not further hex edit would be done instead.
I would work on that, but you keep going back to ME and other issues, so I've set aside the menu stuff for now until you were ready again.

For now, please lets let Plutomaniac help you figure out the ME issue, he is pro at Intel ME! Thanks @Plutomaniac Lets see if he comes up with same summary I have

All files I have given you have clean new ME, except for a few tests you specifically requested with just copy from one to another, even some of those I cleaned first before putting in, so most 99% have been cleaned setting transferred ME.

So you know, this is very confusing and I doubt anyone would understand, maybe best to use another word or explanation in those instances instead >> IntelQST: What ???

OK, thanks, I will do that. I read the PDF and couldn't understand how to get anything QST related from a saved BIOS or from a BIOS dropped into FIT either. It read to me like I either had to make from scratch with the tools, or use the tools on a running system to setup some of the ini and qst files.
None of it looked like anything I could do something with from my end with only the BIOS dump and no system. @kanta - I will send you QST Software package, please do not upload and post around forums, thank you.

I'm glad it's not just me I thought maybe I wasn't familiar enough with it, but that was what I could gather, I needed the system in hand. There is a place in FIT to input a QST Config file

@Lost_N_BIOS, @plutomaniac, everyone else.
I'm really thankful to you. I was able to completely solve the problem of ME and QST.

I worked by reading the QST Tool document (pdf). We have created Qstconfig_D5380.INI by executing the command [QstINI - i Qstconfig_D5380.INI] with FMV-D5380 NO.2 (machine 2).
Next, the command [Qstcfg Qstconfig_D5380.INI] was executed on the defective FMV-D5380 NO.1 (Machine 1), and Qstconfig_D5380.INI was loaded.
This made it possible to solve the defect. Next, we could also create a bin for FIT with the -w option.
In addition, I loaded some standard QST settings, but they all got incomplete QST.
Also, as I looked at Fujitsu Electric's QST setting, there was obviously a strange part, so I'm working on fixing it now.

Currently, we use the new BIOS.bin, [F02_FMV- D5380.bin + 5.2.0.1009_DO_PRD_RGN.bin + Qstconfig.bin] for all of machines 1 to 3.

Three machines still have problems with memory limitation, but I would like to ask this with another thlet.
I'm really thankful to you.