No POST after writing single bit to CMOS; any solution?

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

The battery is slightly weaker, 2,84V or so, but even if it is removed, once I save the BIOS, it won't POST again.

[SteveNielsen]

Re: No POST after writing single bit to CMOS; any solution?

Perhaps CMOS has been damaged by ESD. The ease and subsequent potential frequency of changing the battery by unknowing people it seems a high enough probability to consider. Regardless of the cause, I think your CMOS is damaged.

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

Nobody touched it inside, my parents are not capable of it. I just know previously the problem was the system was suddenly hanging on POST, than some guy took it, found nothing and gave back. I've reset the CMOS and you know the rest. I'd say the previous problem was already a sign of bad CMOS memory…

[SteveNielsen]

Re: No POST after writing single bit to CMOS; any solution?

I agree. Just thinking out loud about what may have caused it.

[Per Hansson]

Re: No POST after writing single bit to CMOS; any solution?

Why have you not tried doing the simplest thing yet, or did I miss it?
Replace the battery

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

It does not work with battery (2,84 V is OK for operation), it does not work without (running on +5 V SB), do you think using different battery will help?

[Agent24]

Re: No POST after writing single bit to CMOS; any solution?

It does not work with battery (2,84 V is OK for operation)

Is that measured in-circuit or unloaded? Either way it may be too low.

A brand new CR2032 I just measured shows 3.4v without load.

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

Most batteries I've replaced started to to loose time all not sooner than under 2,5 V, still plenty to go

[Sparkey55]

Re: No POST after writing single bit to CMOS; any solution?

Most batteries I've replaced started to to loose time all not sooner than under 2,5 V, still plenty to go

Geez, how hard is it? Try a new cmos battery at least to be able to rule that out as a possible cause of trouble.

[momaka]

Re: No POST after writing single bit to CMOS; any solution?

Just an idea, although it probably won't work (and you may have tried this already), but still...

Have you tried disabling everything in CMOS? That is, the floppy controller, IDE controllers, serial and parallel ports, onboard audio and LAN (if present)... even stuff like SMART and L2 cache, if such options exist.

Also, make sure RAM is set to run by SPD or underclock it to its minimum, just in case. Set its voltage to SPD or normal settings too. Same with CPU - normal voltage and underclock to the minimum FSB.

But leave the time and date same as before.

Geez, how hard is it? Try a new cmos battery at least to be able to rule that out as a possible cause of trouble.

Agreed.
It likely won't fix the problem, but it's still better to try it than not. After all, what if it really did fix the problem?

[Per Hansson]

Re: No POST after writing single bit to CMOS; any solution?

Most batteries I've replaced started to to loose time all not sooner than under 2,5 V, still plenty to go

All CMOS designs I've seen differ greatly.
The big name brands like Dell often manage to design it well so that the battery is run really low, and they also warn of low battery in time.
Other boards just say nothing and stop working properly.
(My Asus Maximus Formula is an example, the clock simply start running slow).

I'm not saying it's it, but changing the battery is a really easy thing to do.
Plus so far in your troubleshooting all we can deduce is that the 2.8v or whatever and 0v (no battery) results in a corrupt CMOS.

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

I told ya, storing single bit in there results in corrupt CMOS. No matter what I do, even if I don't change anything and save, it's bad. If I exit without saving, it is able to start. Does not seem to me like battery problem if it does the same without battery.

Even if it worked with battery, it is useless, that would mean the stand-by supply is not being used. In such case only battery will hold the settings and I'd have to replace it every half a year. I'd rather throw the board to river and replace it than this.

Will remove it and check the circuitry as soon as I get some time.

[Agent24]

Re: No POST after writing single bit to CMOS; any solution?

Check for bad connections or open parts in the standby supply line to the RTC etc then?

[Sparkey55]

Re: No POST after writing single bit to CMOS; any solution?

Just for shits and giggles try changing the keyboard out with another working keyboard. Could be a corrupt key signal or dirty off frequency signal geting into the bios registers. Also highly suggest using a NEW BATTERY for the tests.

[momaka]

Re: No POST after writing single bit to CMOS; any solution?

I'd rather throw the board to river and replace it than this.

Okay, that's fine. But just try a new bat. for the sake of curiousity / scratching everyone's itch here (mine too). It takes like 5 seconds anyways. Don't say you don't have time for that.

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

Different keyboard was used, I took only the case with me to Prague, no change in here.

Not really sure why you guys try to get most crazy explanation. I opened this thread because I have seen TOO MANY different motherboards with the same problem so far. One of them was some better 775 board with burned RJ-45 connector, probably from lighting so it strongly suggests CMOS or similar defect. The boards is long gone but I have seen more. Hell for sure at least once I have also tried replacing battery, last time it was some Socket 478 board and IIRC it was too weak already. Guess what? It never helped, otherwise I would not start this thread, right? I am trying to find if that is CMOS (and if, where the hell it is located), BIOS or something on the way (e. g. standalone controller responsible for writing into the CMOS).

So far best possible explanation was with some diodes, just did not get time so far to remove the board from the case and have a look after them. But the rest is just - pardon me - rubbish. Clearly the easiest expanation does NOT apply here, this is more serious problem.

[Agent24]

Re: No POST after writing single bit to CMOS; any solution?

I am trying to find if that is CMOS (and if, where the hell it is located), BIOS or something on the way (e. g. standalone controller responsible for writing into the CMOS).

In older machines the BIOS settings used to be stored in a separate battery-powered RAM chip, nowadays it is usually integrated into the Southbridge.

http://en.wikipedia.org/wiki/Nonvolatile_BIOS_memory
http://en.wikipedia.org/wiki/Real-time_clock

[Behemot]

Re: No POST after writing single bit to CMOS; any solution?

Been using HDMI graphics in this machine for testing plasma so I fed it with new CR2032, guess what…

[Agent24]

Re: No POST after writing single bit to CMOS; any solution?

Been using HDMI graphics in this machine for testing plasma so I fed it with new CR2032, guess what…

I don't know, you tell us?

[stj]

Re: No POST after writing single bit to CMOS; any solution?

suspense is killing me