How much current draw on a CMOS battery?

I lost the CMOS settings on this computer, and since the 2032 button cell was a few years old, I decided to change it out. When I looked at the old one, I found that it was still measuring 2.8v open circuit, and it's still able to drive an LED to upwards of 40ma. So how much current draw is there from the CMOS chip in a PC? I'm thinking that it should be in the nanoamps range, especially when it's quiescent. Anybody know offhand?

I don't know how much current there is but my dads computer drained a 2032 down to about 1.5v over a period of about 10 years. I am guessing self discharge and age accounts for a lot of that voltage drop.

I just dug a VERY old 386 motherboard out of my closet. It has a 3.6v AA size lithium battery soldered directly to the board. It still reads 3.66v! /ubbthreads/images/graemlins/eek.gif Pentiums have been around since the early '90s...when did 386's come out?

it depends on the 386 chip. The drain on cmos is rather low, I am thinking perhaps self discharge may drain the battery more than the cmos does. The idea of it lasting over 5 years.. probably at least 7-10 most of the time on older simpler computers (386,486) with smaller cmos chip memory sizes is not unthinkable... perhaps the larger 1-2meg cmos memory chips have larger ram areas. only a small amount of memory is backed up. most is a burned rom chip.

Good point. The machine that lost it's CMOS marbles is a newer style Athlon mobo with lots of feature settings to back up. Apparently the internal resistance of the old cell went up somewhat. I re-tried the LED test, leaving it on longer this time, and it started to fade pretty quickly.

It has nothing to do with the type of processor. The CMOS memory is not part of the chip. Over time however the die has been shrunk, and the transitors got smaller, which means lower leakage currents and lower switching currents, and longer battery life.

CMOS is actually short for COS/MOS, Complimentary Symmetry Metal Oxide Semiconductor. That means every cell in in fact two MOS transistors in series, one is always on, the other always off. the result is tha other than changing from states, the current draw is exceptionally low. A button cell has a capacity on the order of a few hundred MAH, so if it last 3 years, that is about 27,000 hours. If the cell was say 270mah, that would put the current draw at about 1 micro ampere. In many COS/MOS devices the current draw is so low, that there isn't even a lot of point in putting in a on/off switch, you are going to get pretty much shelf life!

Most likely the battery in your mobo has been sitting in stock for too long. They have been using 2032 cells in mobos for over 15 years. A simple voltage test will probably show up as well below 3v. I am unsure how low the voltage can go to sustain memory of todays and yesterdays motherboard but I recall the battery in my dads system had dropped to about 1.6v when it started losing its mind... or settings. I haven't changed the battery in my computer for a long time but I don't turn it off and have it on a UPS so the battery rarely is keeping the memory up.

uh, come to think of it, i have never changed a CMOS battery.... no, hang on, i did it once to an old pentium that had been sitting idle for 13 years (happened a few months ago) - probably just died of self-draining.

meh. just my 5 cents (we dont have no stinkin pennies down here!)