New CMOS batteries lasting a little more than a week

Hender

Senior member
Aug 10, 2000
647
0
0
I seem to have a pickle on my hands, and I'm honestly not sure what could be causing it without spending hundreds of dollars replacing numerous parts in my computer. The basic problem seems to be this: my CMOS battery is running out after only a few weeks. I'm currently on my third new battery in the last month and a half.

The symptoms are this: when the battery fails, I'll return to my computer and find that the system has "half frozen." That is, the clock has stopped, there's no CPU activity (literally--0% according to Task Manager), and whatever was running is still running. Menus can be clicked on, the Start Menu responds, but the system doesn't respond to Shut Down options. I eventually have to hit the restart button, and if the system boots I get a CMOS checksum error. Most of the time the system powers back on but nothing appears on screen. The LED diagnostics hooked up to my motherboard are failing on the decompressing the BIOS step. At that point, if I replace the CMOS battery, reset the CMOS, and power back up, everything boots fine. About two weeks later, I have to start all over with the same symptoms and solution.

Now, I know that it may not necessarily be the CMOS battery itself--I have no way of testing the batteries I replace, but a new battery fixes it every time. I've honestly never heard of a CMOS battery running out so quickly, and in all my years building computers, I've never *once* had a CMOS battery die on me. Now I've had three go belly up in less than two months.

I should note that this is a motherboard that was returned from an RMA to MSI, so it wasn't purchased new. I don't know if they ship refurbished models or new ones on RMA returns, so that may play a role.

Possible culprits:
Motherboard - MSI K8N Neo4-Platinum, returned from an RMA. Not the original board I sent in, according to the serial number. The BIOS has been flashed to the latest revision, accoridng to the MSI website.

Power supply - CoolerMaster 350W, came with my CoolerMaster case. PC Wizard 2006 reports that the 3.3V rail is running at 3.09V, the 5V at 4.99V, and the 12V at 12.65V. Nothing is overclocked in the system.

UPS - APC UPS purchased last summer, 650W capacity. The site wiring LED is lit, but I live in a house built in 1912--it's to be expected. I have surge protectors in other rooms with the site wiring LEDs lit, too.

Other components in the system - Athlon 64 3200+, 2x512 MB Mushkin Basic PC3200, Creative Audigy 2 ZS, Netgear 802.11g wireless card, nVidia GeForce 6600GT PCI-E video card.

I'm at a complete loss to solve this one. I've encountered and solved many a problem in my day, but this one has me stumped since I've never seen anything like it before. Anyone come across something like this before? Anyone have any insight as to things to try?
 

jonnyGURU

Moderator <BR> Power Supplies
Moderator
Oct 30, 1999
11,815
104
106
First thing that "pops" to mind is power supply.

Second is motherboard. Not unheard of. When I worked at ASI (Asus's biggest distro) there was a "run" of boards that had a "problem" that translated to CMOS batteries not being charged while the PC was running. Solution? They sent me a floppy with a new BIOS revision that actually fixed the problem and a Ziploc bag full of CR2032 batteries!
 

Jiggz

Diamond Member
Mar 10, 2001
4,329
0
76
This sounds like more of a psu problem or bios problem rather than a battery. Next time you encounter the problem try not to replace the battery. Instead just reset the cmos through the jumper. Then see if the system boots fine. Remember try running the system at bios default setting first until you encounter the problem again. If it will not boot, then enter the bios when it post. Then check make sure the system settings are at default. If the system settings are at default, check the voltage settings on the cpu and rams. Make sure they are within range. If they are adjustable, raise the cpu and ram voltages on the next higher level. Try again. If the system fails again, then raise the voltage again on the next higher level. This should be your last voltage raise since you can damage both rams and cpu with too much voltage. If the system boots, check the cpu temp and voltages. If the system will still not boot, try a different psu of higher ratings or of better quality. The cmos battery is not rechargeable. It's not designed to be charged. Instead it only provides power when the PSU is turned off (not the system!) or unplugged or when the mobo is unplugged. In the old AT cases or mobo yes, the cmos battery is rechargeable but not since the ATX. If you want to find out if the cmos batteries you got rid off are good or bad, use a DMM to check for voltages and it should read at least 3.0 V. You might want to reflash the bios again using the older version just to see this could be a problem too.
 

Hender

Senior member
Aug 10, 2000
647
0
0
Well here's a twist to the story that I forgot to add.

When the system enters the state in which a new CMOS battery fixes it, reseting the CMOS via the jumper (rather, a button on the MSI board) does little to fix the problem, and 9 times out of 10 the system will power on when I plug it back in, but will still sit on the "decompressing BIOS image" LED error from the MSI Diag cable. Nothing on screen, no beep at boot, just my peripherals buzzing as they power up.

Now, on that 1 time out of 10 when the system does boot, I get a CMOS error onscreen that there was an error in decompressing the CMOS image, and it prompts me to enter the BIOS since it's reset everything to the defaults.

Once I enter the BIOS, however, I am unable to set my settings. On the "Standard" page where you set the date/time, the year jumps around. It'll start at 1999, then jump to 2099, then 2399, then 2000, and then sit with a couple of the digits skipping between two years. All this without touching the keyboard. Once I reset the system, it's back to a non-boot situation.

The consensus seems to be PSU related. That seems to be the first logical place to look. Ironically, I bought a new PSU to try and troubleshoot the problem that originally caused me to send back the MB, but I sold it once that was determined to not fix the problem.

Sigh.
 

modedepe

Diamond Member
May 11, 2003
3,474
0
0
My money would be on a bad motherboard. Maybe there is a short somewhere causing the battery to drain. I highly doubt it's a bad PSU.
 

JimPhelpsMI

Golden Member
Oct 8, 2004
1,261
0
0
Hi, CMOS batt should be good for something like 6-8 years. Take it to the local Radio Shack. They will test it with a voltmeter. New will be about 3.1 volts. It's usually OK down to about 2 volts. Good Luck, Jim
 

Hender

Senior member
Aug 10, 2000
647
0
0
Well, I decided to kill two birds with one stone. You guys couldn't agree on which is more likely to be causing the problem, so I just went ahead and ordered a new MB and a new power supply. I picked an Abit KN8-SLI motherboard and an Antec 400W power supply, so if that doesn't solve my problem, I don't know what will. And hey, it's only money, right? Thanks for all your help!
 

Jiggz

Diamond Member
Mar 10, 2001
4,329
0
76
Well, let me know what you will do with the old mobo. I would like to troubleshoot that sucker to see what's up. I'm not a big fan of MSI but I'm intrigued with your symptoms.
 

Hender

Senior member
Aug 10, 2000
647
0
0
I'd be willing to sell it to you--cheap, of course--if you'd like to play with it. I'll send you an email with details.
 

KoolHonda

Senior member
Sep 24, 2002
331
0
0
It's a problem with Neo4's, looks like they had a bad run of boards. The only solution I've seen on the MSI forums is to RMA the board.
 

Hender

Senior member
Aug 10, 2000
647
0
0
Sadly, I've done that twice already. I'm on my third MSI K8N Neo4-Platinum, courtesy of MSI. I had an unrelated problem lat e last year that made me return the board I originally bought for replacement, and when the "new" board came back I had a different problem, so I returned that one, too. I eventually discovered that my CPU was bad and replaced that, so I was stuck with the second board they'd sent me.

I'm not willing to wait around for MSI to RMA yet another board and possibly be stuck with another malfunctioning board. It took them four weeks to RMA the board each time including shipping time to and from, which was unacceptable to me. Since I did it twice, that meant I was without my desktop for two whole months.

To me, it's worth spending the money to get a new board this time. Who knows? The new board and power supply just might fix my CMOS battery/BIOS problem as well as my hard drive problem that I've been having, too. I lean away from that being a drive problem since I had the exact same problem on my other hard drive, but it went away when I got the new motherboard. Go figure.
 

SnoMunke

Senior member
Sep 26, 2002
446
0
0
It is KNOWN documented problem with that motherboard. I had the same motherboard with the same problem. Do some google seaches and you will find lots of discussion about it. RMA is back to MSI fully explaining what is going on...hopefully they will give you the revision that has fixed that problme.
 

Hender

Senior member
Aug 10, 2000
647
0
0
Originally posted by: SnoMunke
It is KNOWN documented problem with that motherboard. I had the same motherboard with the same problem. Do some google seaches and you will find lots of discussion about it. RMA is back to MSI fully explaining what is going on...hopefully they will give you the revision that has fixed that problme.

When I first started researching the problem, I Googled for generic terms but not my specific model. At this point, I'm not sure if it would have made a difference. I think I'd prefer to just get a whole new board than wait another four weeks for a replacement for them.
 

Fox5

Diamond Member
Jan 31, 2005
5,957
7
81
I would make sure your jumpers are set correctly. Normally there's a jumper config on the mobo to clear the CMOs, maybe it is currently set to that position and draining the battery?
 

Hender

Senior member
Aug 10, 2000
647
0
0
Originally posted by: Fox5
I would make sure your jumpers are set correctly. Normally there's a jumper config on the mobo to clear the CMOs, maybe it is currently set to that position and draining the battery?

No, that's not the case. The MSI board I currently have has a button rather than a jumper that clears the CMOS so there's no chance of a jumper causing the trouble.

Besides that, I've never heard of a system not booting with a drained CMOS battery. They've always turned on for me, but the BIOS settings have been wiped each time you unplug it. In none of the times I've run across a machine like that has it resulted in a no-boot scenario, which leads me to believe there are darker forces at work.
 

Hender

Senior member
Aug 10, 2000
647
0
0
Originally posted by: fire400
if you can't flash the BIOS, then you got a problem.

Agreed. I'm on the latest BIOS revision as it is, but the damning indicator for me was the year in the time/date part of the BIOS jumping around on its own...in my experience, if you're having trouble with BIOS locking up or whatnot, you've got some serious hardware problems somewhere.
 

JimPhelpsMI

Golden Member
Oct 8, 2004
1,261
0
0
Hi, The reset jumper disconnects the battery and shorts out the caps in the circuit to reset the BIOS SETUP to defaults. It should not run down a battery. Likely one of theCaps in the circuit is leaky. That would drain the batt very rapidly.
Hope this helps a little, Jim
 

piasabird

Lifer
Feb 6, 2002
17,168
60
91
If the motherboard is so bad that it drains the CMOS Batteries, them make the manufacturer send you a replacement of fix what is wrong with it. If you purchase a boxed motherboard with a retail warranty make them give you a new one.
 

SnoMunke

Senior member
Sep 26, 2002
446
0
0
All-

It is a DESIGN FLAW! Not a jumper setting, BIOS revision, etc. RMA the board or buy a different one.
 

Hender

Senior member
Aug 10, 2000
647
0
0
Originally posted by: SnoMunke
All-

It is a DESIGN FLAW! Not a jumper setting, BIOS revision, etc. RMA the board or buy a different one.

I bought a new one. I wasn't willing to wait for the 4-week turnaround time with MSI again and risk getting another bad board. I ended up going with an Abit KN8-SLI, which I like the design of better because it uses a heatpipe cooler for the chipset instead of a fan, which slightly lessens the noise in my system. The chipset fan on my MSI board was dying, too, and becoming quite annoying to listen to.
 

imported_hopeless

Senior member
Oct 29, 2004
777
0
0
Even though you bought a new mobo, you could still RMA the other one. Then sell it after you get it back. Might help you get back some of the money you just dished out.
 

Hender

Senior member
Aug 10, 2000
647
0
0
Originally posted by: hopeless
Even though you bought a new mobo, you could still RMA the other one. Then sell it after you get it back. Might help you get back some of the money you just dished out.

Not a bad idea. I'll have to take it out, anyway, and since I'll have a working system it won't matter that much.
 

JimPhelpsMI

Golden Member
Oct 8, 2004
1,261
0
0
Hi again, Power Supply should not be able to run down a CMOS battery. If it's a bad component it is almost certain to be the Mother Board. Did you test the old batts? Were they actually down? Jim
 

SnoMunke

Senior member
Sep 26, 2002
446
0
0
Originally posted by: hopeless
Even though you bought a new mobo, you could still RMA the other one. Then sell it after you get it back. Might help you get back some of the money you just dished out.

This is a good idea since MSI fixed the problem is the subsequent revision of that model.