Re: BIOS change killed my computer.

John Doe wrote:

Conor <conor.turton@xxxxxxxxx> wrote:

John Doe says...


I think using the jumper to clear the CMOS/BIOS will do just as
well.


Then you think wrong in many cases. On an ATX motherboard, it is
still powered even though the computer is turned off.


Of course it is still powered. When things are not powered,
switches don't work. 

That depends on what the switch is.

What this
means is that even though you use the Clear CMOS jumper then
because power still isn't interrupted, the settings aren't
reset.


A reset input resets the device. The device should be powered at
the time reset is used.

Although it's possible (anything is) that someone made one, the typical clear CMOS jumper is not a 'reset input' as there's very little reason to go to the effort of making a 'reset input' when simply removing power to CMOS will kill (reset) whatever is in it.

I have had many ATX mainboards. The instruction has been to turn
the system off before clearing the BIOS/CMOS. None of them have
suggested unplugging the power supply or removing the BIOS/CMOS
battery.

And we all know how explicitly clear and intuitively obvious the typical Sinoenglisheese manual is.

I have no idea what your particular manuals actually say but it's possible that when they say to turn the system off they mean to really turn the system off, like with the AC mains switch on the PSU (a switch, btw, that doesn't need 'power on' to 'work). Pulling the AC plug is simply another way of doing it, and some PSUs don't have a mains switch.

My K7S5a Pro manual says to do it the hard way and first remove all power plugs from the motherboard before operating the jumper.

I think the reason the system should not be on when the CMOS
jumper is cleared is probably because BIOS data is being accessed,
not because power is being supplied to the BIOS chip. The system
being on probably does not mean some parts of the mainboard are
still powered.

The typical CMOS circuit *IS* powered by the PSU when the system is 'on' and also in the 'front panel power switch off' state because that saves battery power. You have AC mains and a multi-hundred watt power supply there so why, in heaven's sake, would you leave CMOS on the battery to drain it?

How the jumper is wired varies with manufacturer. Some just break the battery connection and you need to wait a while for it to drain, with the 'second' location being just a place to hold the jumper (often shipped in that position so the battery isn't drained during storage). Others use the second location to short the CMOS device to ground for a 'quick' discharge. Those are the ones you sometimes see people wondering why their shiny new motherboard won't power up because it shorts the standby power as well (since that is primary CMOS power), preventing the PSU from turning on.

Lastly, you could break the CMOS device's power pin itself off the entire circuit, so that neither standby nor the battery are on it, but they tend to not do that because it would float the power pin and engineers don't like floating pins on CMOS devices. Second, it falls into the category you mentioned. You'd have, at the very least, the 'power on switch' logic waiting to debounce the switch and 'decide what to do' when it's pushed as, suddenly, the CMOS settings telling it what to do go to a nonsensical state. As well as the wake on LAN, timer settings, et all. No, you don't want standby power on even IF it didn't matter to clearing the CMOS chip itself (and it does).

Regardless of how it's done, however, you are assured of the CMOS being cleared when you operate the jumper if you first remove all power to the system by either the PSU mains switch or pulling the plug. And pulling the battery removes any 'confusion' about jumpers.

I cannot imagine how removing the battery would be a good idea if
your mainboard is still powered.

That won't accomplish anything either as standby power remains on the CMOS. It's doing both that work, unplug it (or use the main switch) and remove the battery (or use the jumper). As long as you wait long enough.

I can imagine how unplugging the
power supply might help, but I'm sure mainboard manufacturers
would say that if it were necessary.

It isn't 'necessary' if you have a mains power switch. It's just 'guaranteed' to remove power whether you do or not and when giving instructions to people, when you have no idea whether their PSU has one or not, it's simpler to just say remove the power plug and be done with it. Same with "remove the battery."

.

Relevant Pages

  • Re: to clear CMOS
    ... Now that the jumper is in the 'clear CMOS' pins, for how long should I power up the computer before shutting down to replace the jumper to its original position? ... I might remove the CR2032 battery from its ...
    (microsoft.public.windowsxp.hardware)
  • Re: Locked Harddrive due to Windows error!
    ... Actually, to clear the cmos, you have to power the system back up with the ... then have to shut down again, un-jumper the pins, and power up again.. ... > Have you tried clearing the CMOS via the jumper on your motherboard? ...
    (microsoft.public.win2000.general)
  • Re: Dumb soldering Q
    ... I can solder a jumper across 2 contacts on the switch board. ... Then add an in-line switch to the power cord. ...
    (alt.home.repair)
  • Re: Tech EM Gottlieb Pop A Card
    ...  Is there a way to jumper a switch to get ... power to the playfield and in doing so isolate where my problem is? ... I would gladly help you isolate how to jumper to get power going if ...
    (rec.games.pinball)
  • Re: XP Wont start at all.
    ... Switch off the computer, including the power switch on the back ... look for the battery which powers the CMOS. ... Close the jumper for 30 secs ...
    (microsoft.public.windowsxp.perform_maintain)