How batteries can achieve ‘der alte’ status

Article By : John Dunn

What are the chances of getting a really long service life if a battery is delivering only a very small current?

The Liechtensteiner Polka opens with, "Der alte Herr von Liechtenstein…" which means the old man of the town—the Mayor. There's also the Yiddish expression "alte kaker," which also means the old man and maybe with a little disparagement thrown in. However, a question recently arose about just how old some batteries could get, about how long the service life could be, for alkaline batteries in a particular application.

Yes, I know that each "battery" is really just one single cell and not really a battery, but with your kind forgiveness, I choose to adopt the colloquialism of "battery" just for the sake of easing this discussion.

What were the chances of getting a really long service life if my battery were delivering only a very small current? What were the chances of a battery getting really old and thus achieving the status of "der alte battery"?

Looking at a D-size battery datasheet, I wasn't learning what I needed to know. Published service life information was only available for up to several hundred service hours. I needed some idea of what to expect in terms of years, not weeks.

battery_der-alte_01 (cr)
Figure 1: Published battery service life data

It then crossed my mind that my bedroom alarm clock uses alkaline batteries and that over the course of time, those AA-sized batteries have delivered well in excess of one year of service life, repeatedly waking me up before dawn each morning. I figured if I could measure how much current the clock was drawing from its AA-sized batteries, I could extrapolate a plausible current draw that I could demand from my product's D-sized batteries. I made this little test fixture for that purpose.

battery_der-alte_02 (cr)
Figure 2: Current measurement test fixture

Diagrammatically, the arrangement was like this.

battery_der-alte_03 (cr)
Figure 3: Test fixture details

I inserted the fixture into my clock this way:

Figure 4: Test fixture installation

The result was the following:

Figure 5: Clock current consumption test

The clock was drawing a somewhat variable current as the second hand swept its way around the dial which made the DMM reading vary somewhat, but the reading seemed to be pretty close to 100µA. There was, however, just one teeny little complication in the clock that I had to deal with in connection with this measurement.

Look at this picture of the inside of my clock. Instead of just one single AA device, there are two of them and they are actually connected in parallel which is something I was once taught to never do. There it was though and I guess that's a story to be told on another day.

battery der-alte 06 (cr)
Figure 6: Clock batteries in situ and connected in parallel

I had taped a date label on the clock's case the last time I'd put in fresh batteries. That date was 13 months earlier and therefore a little bit in excess of one year. Rather promising, I thought. Since the clock could still run properly with just one AA-size battery installed, I was able to read the demanded current as I had originally planned by removing one battery and installing my fixture in series with the other battery.

Nominally, with perfect balance, the current burden being imposed on each battery would come to only half of my meter reading, or only 50µA. From that number, I extrapolated an estimation of what a D-size battery might be able to do as follows.

battery der-alte 07 (cr)
Figure 7: D-cell versus AA-cell comparison

From the datasheet for the alkaline cell devices I was studying, I found that on average, the milliamp-hours capacity of the D-size device comes quite close to seven times that of the AA-sized device. Thus, my admittedly blithe inference is that if my operating device(s) don't draw more than seven times 50µA or 350µA from my D-size batteries, I should be able to obtain a service life in excess of one year.

Admittedly, this is based on a lot of iffy and specious assumptions, but it's still a better estimate than I could make by merely taking a blind guess.

First published by EDN.

Leave a comment