PDA

View Full Version : CMOS Battery needs to be replaced frequently


MoonSword
06-25-2004, 05:57 PM
I have two Socket-A motherboards, Asus A7V333 (Rev 1.02) and an ECS K75SA Pro (Rev 5.0L)

The problem is that the CMOS batteries (CR2032 coin cell) need to be replaced frequenlty... is this normal for Socket-A/Athlon XP systems?

Actually, the Asus A7V333 board's original battery lasted a year before I needed to first replace it, and the replacement battery lasted six months. Ever since then, I've only gotten 3 months of life out of each battery.

The ECS K7S5A Pro was recently purchased new from Newegg.com, but its original battery only lasted two months.

I have been an A+ certified tech for years, and in all this time, I've rarely ever had to replace a CMOS battery. Heck, I've got an old Asus P3V4X slot-1 (P2/P3/Celeron) motherboard that I used for four years, and not once did I have to replace the CMOS battery. I've never had to replace a CMOS battery in my old Socket-7 systems, either. I find it odd and most annoying.

Obscure troubleshooting tip: My Asus A7V333 won't boot if the CMOS battery is no good. In fact, the computer acts totally dead, which would lead many to believe the problem lies in a DOA Motherboard, bad power supply, or a faulty case power switch, etc... Some motherboards act totally dead if the CMOS battery is no good, and it's an easy fix that only costs a trip to your local Walmart or Radio Shack and $3. I just wish the darn battery would last longer than a few months.

MoonSword

homestarmy
06-25-2004, 06:39 PM
That sucks... I built a computer for a friend with the K7S5A a while back and the battery went out not long after. I replaced it with one from my board (those boards are sh|t, mine is dead) and it hasn't been in there long so I can't comment on how long it will last now... it lasted a while in my board.

huesmann
06-26-2004, 07:15 PM
That's way too short a time for the CMOS battery to go bad. Something's wrong with the mobos. I have mobos that are 5 years old that are on the original battery.

BadThad
06-26-2004, 08:57 PM
What kind of battery are you putting in? Sounds like a bad batch of batteries to me being it's happening on two different systems.

MoonSword
06-30-2004, 03:25 AM
Energizer Lithium CR2032. I get two-cell packs from Walmart for a little more than $3. Next time I buy batteries, I'll get them somewhere else, and a different brand. That'll help clear up the possibility of a bad batch. But even still, that would be a very, very large bad batch of batteries as I have bought several packs over the last year. Oh well. I'll see what happens. Thanks for the advice.

MoonSword

Ryoga
06-30-2004, 03:38 AM
Well how many can one store go through? It's likely they're all from the same lot.

If the situation doesn't clear up, then you either got a bad mobo or a bad PSU (or a PSU with a lose connection). How often does the PC sit with no power to the PSU?

MoonSword
06-30-2004, 04:20 AM
I imagine that Walmart store tends to go through a lot of CR2032 batteries. This Walmart is not in some rural backwoods location. In the area I live, we do have more people than cows. :D Not to mention that lots of different electronics use CR2032 batteries: Graphing calculators, data organizers, some cameras, etc.

Anywho, I tend to leave the PCs running all day and night. Even when I do power either one down, I leave the master PSU switch to on, so the PSU itself has power in it. About the only time the PSUs don't have power is when I swap the battery, or open the case for some other "surgical procedure". One thing is for sure... When my computer won't power up (after having powered it down for some reason), I swap the battery, and voila! It's got life again. BTW, PSU voltage ratings are at normal.

MoonSword

Zepper
06-30-2004, 06:44 PM
Are you actually measuring the voltage on the batteries before replacing them, or just assuming because the time is off, that the batteries must be going. They should be good down to 2.8V or so. There are other reasons for the time being off - fancy screen savers that eat interrupts and having the clock set to sync from a network that doesn't exist (or isn't connected to very often) are just two...
.bh.

There's the :sun: !

MoonSword
07-01-2004, 03:07 AM
Are you actually measuring the voltage on the batteries before replacing them, or just assuming because the time is off, that the batteries must be going. They should be good down to 2.8V or so. There are other reasons for the time being off - fancy screen savers that eat interrupts and having the clock set to sync from a network that doesn't exist (or isn't connected to very often) are just two...

It is my understanding that a voltmeter is a poor choice to test a battery. Voltmeters do not put a load on the battery, and without a load, the reading isn't reliable to determine a battery's strength. And since I do not have an actual battery tester, I really can't test the battery itself.

But here is where it gets interesting (a.k.a. annoying :disgust: ). The problem isn't really my internal clock. Heck, that tends to be off by a few minutes at most given times, so I just make it a point to fix the clock every six months or so when the time is very noticably off. Actually, the reason why I am annoyed is that my Asus A7V333 Mobo won't even power on if the battery is no good. When the computer appears totally dead, I pop in a new battery, and voila! My system is alive again! Quite annoying and inconvenient, but at least its a very inexpensive fix. :roll: Even still, I've been a certified, professional technician for five years, and this is the first time I've had, or even seen, frequent CMOS battery problems.

MoonSword

LTC8K6
07-01-2004, 06:31 AM
You need to make sure that the batteries are actually dead as the first step to diagnosis.

Usually you notice a few odd things if you have a low battery. The clock is off, it doesn't keep BIOS settings, etc.

Does the board boot if you just clear the CMOS with the jumper?

How about if you put the same battery back in?

Does your BIOS read the battery voltage? If so, you could check it every once in a while to see if it is dropping.

If the batteries are actually dead, and it wasn't a bad batch of batteries, then it's gotta be the board. I have several computers where those batts lasted many years.

LTC8K6
07-01-2004, 06:36 AM
The ECS board's battery may already have been old.

Zepper
07-01-2004, 10:09 AM
I use the voltmeter to measure the battery IN CIRCUIT. And I do it after the AC power has been unpluged (or turned off by the switch on the back of the PSU) for at least 10 minutes. Even though that might not be quite enough load to do a proper test, it will give you the actual voltage the circuit is seeing. You could also measure at the backup battery pin on the southbridge in case there are any intervening diodes. On most that I've checked, there is no significant diff between readings at the top of the battery and the backup battery pin.
. Perhaps the sensitivity of the internal CMOS battery measurement is off - I've heard that Asus health chip's (and others) readings often aren't all too accurate. Perhaps a BIOS update would help. Or you could disable the CMOS battery check function. Whatever the cause, three months is far too little battery life on a 2032 - avoid Chinese made ones if at all possible (Chinese batteries are generally shiznits). I'd be rigging a dual AA-cell battery pack for it if I couldn't find any other cause...
. I got my basic electronics training in college back in the mid-60s and have been doing techie stuff ever since - so I know how to test batteries.
.bh.

lobadobadingdong
07-01-2004, 03:42 PM
Originally posted by: LTC8K6
The ECS board's battery may already have been old.they get cheap battery's that have been sitting on a shelf a while from time to time.