Jump to content

Mystified - Bios Settings disappearing


Recommended Posts

Kinda scratchin' my head over this one.

 

I tweaked my overclock settings just a bit from 270 FSB to 275 and ran memtest from bios to see how stable RAM would be (9/10 divider, or 180/200). I got errors so I rebooted out of memtest, and it wouldn't post. I cleared CMOS, put in some conservative settings (220 FSB, same RAM settings that were stable at 270). It came up alright, so I pulled the power plug again, put the shell back on the case, plugged things in, and tried to start it up - wouldn't post again. Repeat, same story.

 

Seems like everytime I pull the plug to power down the board, I lose Bios settings, and it won't post. I've pulled the battery, put the pin in CMOS clear position for an hour or more, put the battery back in and moved the CMOS pin back, and it still does it. Does it sound like the battery could be dead? The board is only a year old. Any other thoughts on what could be wrong?

 

I cleared CMOS for the first time ever last week with my raptor problem. I don't feel like I've abused the board. Finally, I should mention that I have tried CMOS reloaded and setting everything manually because I know some say CMOS reloaded can be corrupted in those situations.

 

Please comment ......

 

Also, I'll check my Bios version and post it. Might it help to reflash the Bios, even to something different?

 

Edit: OK, as the machine posts, I see Award Bios Ver. 6.00PG. In Bios, the battery voltage reads 3.08 volts.

 

Another Edit: You know, I really didn't research this too much, so I apologize if this is a no-brainer I should have found the answer to ..... busy helping the wife with dinner, etc.

Share this post


Link to post
Share on other sites

Thanks. I'll look into flashing the bios. I must add at this point that I have taken a few minutes to search the forum. I lloks like there is a very specific procedure for clearing CMOS in a no-POST situation (http://www.dfi-street.com/forum/showpost.p...04&postcount=24)

that I should try. Of course, my system will post, it just doesn't like to be unplugged after I get it set up.

 

I guess the thing that really bothers me is that this is a tweaker's board. It's made to help us mess with things to squeeze extra speed out of all our components. If I can't do some relatively sane testing like I did with my modest bump 5 MHz FSB without getting into this much trouble, I might as well buy an ASUS or MSI and leave it at stock settings. I'm up and OC'd by a few hundred MHz right now, but if I can't play with this board, I'll buy a new DFI.

 

No bitterness here - it's just that two sessions of "is my PC dead?" in two weeks have me a lttle shook.

 

Edit: Disregard the preceding rant. I just need to go to bed and refresh my perspective. Looking forward to all of your input!

Share this post


Link to post
Share on other sites

Seems like everytime I pull the plug to power down the board, I lose Bios settings
Battery or CMOS inside chipset is bad most likely.

 

RGone...

 

Kinda scratchin' my head over this one.

 

I tweaked my overclock settings just a bit from 270 FSB to 275 and ran memtest from bios to see how stable RAM would be (9/10 divider, or 180/200). I got errors so I rebooted out of memtest, and it wouldn't post. I cleared CMOS, put in some conservative settings (220 FSB, same RAM settings that were stable at 270). It came up alright, so I pulled the power plug again, put the shell back on the case, plugged things in, and tried to start it up - wouldn't post again. Repeat, same story.

 

Seems like everytime I pull the plug to power down the board, I lose Bios settings, and it won't post. I've pulled the battery, put the pin in CMOS clear position for an hour or more, put the battery back in and moved the CMOS pin back, and it still does it. Does it sound like the battery could be dead? The board is only a year old. Any other thoughts on what could be wrong?

 

I cleared CMOS for the first time ever last week with my raptor problem. I don't feel like I've abused the board. Finally, I should mention that I have tried CMOS reloaded and setting everything manually because I know some say CMOS reloaded can be corrupted in those situations.

 

Please comment ......

 

Also, I'll check my Bios version and post it. Might it help to reflash the Bios, even to something different?

 

Edit: OK, as the machine posts, I see Award Bios Ver. 6.00PG. In Bios, the battery voltage reads 3.08 volts.

 

Another Edit: You know, I really didn't research this too much, so I apologize if this is a no-brainer I should have found the answer to ..... busy helping the wife with dinner, etc.

Share this post


Link to post
Share on other sites

Update - I followed the posted CMOS clear procedure referenced earlier, and used a new CMOS battery. No joy, same problem. At this point, fifteen- and thirty-minute CMOS clears wouldn't get the thing to boot, so I did an overnight (~8 hour) CMOS clear. It started up this morning just fine, but I knew the underlying problem was still there - pull the plug and I'd be back in the same no-boot situation - so I downloaded the BIOS CD from the BIOS factory, and flashed the BIOS to the 623-2 BIOS. Don't know if that was a good choice or not, but that's another topic.

 

Anyway, it flashed fine, and I had to come to work (from whence I am writing this) before I could do any more testing.

 

I write all of this to ask RGone: Is this what you meant by "CMOS inside chipset is bad"? In other words, is it possible that a BIOS flash will fix the CMOS inside the chipset? law_freak suggested flashing earlier, and I am just now doing it beause flashing feels a little dangerous. Heck, I guess everything we do here is a little dangerous, come to think of it.

Share this post


Link to post
Share on other sites

A bios flash from Dos with the correct switches will not make the situation any worse at all. Bios flash does not flash the cmos itself but only the bios chip.

 

In your situation, it is my suspicion that the cmos storage area and the bios information at boot are not in agreement because the cmos is likely corrupt for lack a better description.

 

"Often" but not all the time, pulling the battery and all power cabling from the board, including the memory and video card and leaving sitting for 20 to 24 hours with the cmos jumper to clear and no battery in the battery holder, will actually allow cmos to fully clear. No man don't ask me why? The engineers share nothing about the 'why' to me. I have to find cures/patches/fixes or workarounds like the rest of you for this mysterious stuff.

 

Along the lines of the extended cmos clearing being some miracle healer here is an example that has actually happened to me. I flashed one of my boards with another companies bios. DO NOT ask why and do not try it. Hehehe. I did it and the board was hung perpetually at the place to enter bios. I pulled the bios chip and flashed a bios chip in another board. No work. Flashed another bios chip with a different bios. No work. I cleared the cmos for 30mins; 3hrs; 8hrs and said to heck with it and pulled the board and replaced it knowing full well what a dumb move I had made but oh well. 3 weeks later I decided to send board in for repair but as always, I needed to double check the symptoms to report to repair facility and viola the board was working and continues to do. So I knew what the problem was actually but my situation required that I quit messing around with the non-booter and move for a while and I did to return later to a working board after the cmos 'healed' itself. Yeah, right but what else to say when I have had similar time and again with many boards I had pushed past their limits including a Via KT333 chipset board and NF2 boards and one ALi chipsetted board. Oddities happen and it is the time factor that forces most to hurry and not be able to sit the board aside as I did.

 

Bear one thing in mind though; pulling everything out of the board and leaving it for 20-24 hours; as I said above, is actually faster than an rma.

 

RGone...

 

Update - I followed the posted CMOS clear procedure referenced earlier, and used a new CMOS battery. No joy, same problem. At this point, fifteen- and thirty-minute CMOS clears wouldn't get the thing to boot, so I did an overnight (~8 hour) CMOS clear. It started up this morning just fine, but I knew the underlying problem was still there - pull the plug and I'd be back in the same no-boot situation - so I downloaded the BIOS CD from the BIOS factory, and flashed the BIOS to the 623-2 BIOS. Don't know if that was a good choice or not, but that's another topic.

 

Anyway, it flashed fine, and I had to come to work (from whence I am writing this) before I could do any more testing.

 

I write all of this to ask RGone: Is this what you meant by "CMOS inside chipset is bad"? In other words, is it possible that a BIOS flash will fix the CMOS inside the chipset? law_freak suggested flashing earlier, and I am just now doing it beause flashing feels a little dangerous. Heck, I guess everything we do here is a little dangerous, come to think of it.

Share this post


Link to post
Share on other sites

Thanks, RGone. Your experience with the 3-week CMOS clear, and suggestion on a stripped 24-hour clear give me hope that if I'm not all better now, I might still get that way with a little more work. 6 or 7 hours from now, I'll see what my current situation is.

 

As far as RMA, I don't think that's an option for me - I've had the board for about a year. I am willing and able, however, to buy a new board if I can't get this one back up. If DFI makes a better board for a non-gaming, non-SLI-needing tweaking geek like me, feel free to suggest it. Gee, just writing that makes me feel a little better. It's not the end of the world, it's FUN!

 

Edit: Problem solved. Flashing the different bios did the trick. It no longer loses settings and refuses to boot after being unplugged. Of course, the dang thing is getting harder and harder to clear CMOS in. It also wants to hang up more when it does not like an overclock setting, forcing me to clear that stubborn CMOS, but - hey - these are issues for another discussion.

 

Thanks RGone and law_freak for the help!!

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...