Question old Fujitsu board cannot save any BIOS settings

[atan_shubashi]

Hi,

I am using this old Fujitsu board (model D3061) in an ESPRIMO P700. Works like a charm, unless I want to change something in the BIOS settings: I can enter the BIOS with F2, I can make changes (in my case to disable the low-power mode and then enable powered USB ports when turned off). But if I save&exit, reboot and later return, the setting has reverted (to "enabled" in this case). Funnily, if I make my change, and use the "Save Settings" menu (i.e. without exiting the BIOS), and go back to my changed settings, it has already reverted. Ergo saving is reverting for this board...

On the other hande I was able to change the boot sequence, and I could change the date and time. And the board does keep date&time even when off power; i.e. the battery is ok (and anyway tests at just over 3V).

I could swear it worked once, the USB ports were powered alright. But I once booted into Linux, shut down, rebooted, tried to make some more changes, and that's when it stopped accepting my changes. Since then it just seems to know better what the settings should be.

I have already tried to revert to defaults, and I have taken the battery out to null all the settings. Did not make a bloody difference.

Any idea?

 

[Grobe]

That is really weird because symptoms seems similar to empty 3V battery, but you tested the battery ok.

What happens if you clear cmos settings (normally by jumper on MB) ?

 

[atan_shubashi]

Can't find a jumper for clearing... The only MB manual I can find is horribly thin in content:
https://support.ts.fujitsu.com/IndexDownload.asp?Softwareguid=459DEC8B-02CE-4A6C-AB31-8B0B6C4D17F6
BIOS manual here: https://support.ts.fujitsu.com/IndexDownload.asp?SoftwareGuid=EC815A65-E33B-4A15-AF51-61227A8A8BA2


There's a jumper for "recovery" (manual says "Recovery inserted = The system starts from USB stick and allows a BIOS recovery. Details can be found in the BIOS manual.", BIOS manual describes BIOS flashing, not CMOS clearing), one jumper for "TPM enable" (which I tried; did not do anything), and one for "Desc Override Enable" (on which I cannot find any information in any of the manuals or on google, so I did not dare touching this one). No "clear cmos" or otherwise named.

When I measured the battery, I did take it out for a few mins, after which the RTC was reset to 1 Jan 2011, but otherwise this did not seem to do anything.

One thing I was wondering: could this be related to this "Intel Management Engine (MEBx)" thing for remote maintenance? But first of all I would expect the BIOS Setup to throw an error if it cannot save because of that, rather than just reverting my changes, and secondly there is nothing I can do with that, as it does not even react to its magic key (Ctrl-P).

 

[Grobe]

Are you able to find a newer bios version?

 

[atan_shubashi]

I just checked. Fujitsu support page lists V4.6.4.0 - R1.29.0 (18/06/2015) as the latest version. Which is exactly what I have on the MB right now.

 

[Grobe]

When you had the battery out - did you looked at it, and the contact points on the battery holder? Was it clean and free of dirt, fat or corrosion?

 

[Calvin7]

Replace battery - It's called routine maintenance.

 

[atan_shubashi]

First of all: issue solved
In the end it was an undocumented jumper. I found one on the "front panel" header, shorting two contacts (9 and 11 if I counted correctly) that -- according to the MB manual I quote above -- have no function. Since I removed that jumper, the settings save fine and of course are preserved fine across a power-off. So it seems this is some kind of write-protect jumper. Would have been very nice if the manufacturer had documented such a thing...

@Grobe: Thanks for your suggestions. Re the last one: yes, I had checked the contacts too, they looked pristine but nevertheless I had scratched them a little bit to remove any (invisible?) corrosion. But anyway this was not a promising lead, because as I wrote above, the MB kept the time across power-off, while removing the battery and putting it back in did reset the clock. I.e. the battery was having an effect.

@Calvin7: yes, routine maintenance is a good idea. Except in this case it would not have made the slightest difference.

 

[Grobe]

Sounds like you just found the "clear cmos" jumper and the documentation is just a mess.