Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Battery matching solution?

0 views
Skip to first unread message

Bruce W.1

unread,
Aug 18, 2003, 12:11:14 AM8/18/03
to
Please see previous thread.

I'm trying to match AA battery capacities so they can be used together
in battery packs and I think created a low-tech way to do this. I'd
like to hear more informed opinions.

This would apply to NiCads and NiMh but it should work with lithium-ions
too.

Take all your batteries of the same type and charge them all in parallel
to the same voltage. Then one at a time remove them from the float
voltage and attach them to a fixed size resistor (say 6 ohms) for a
fixed period of time (say 3 hours) then measure their voltage. Well
matched batteries should have similar voltages.

Does this wash?

The only thing I wonder about is the full charge state voltage. Say I'm
doing Ni-Cads which usually top-out around 1.65 volts (where they
experience a voltage drop), and I set the float voltage for the bunch at
1.5 volts. Would the maximum voltage of each cell make a difference
here?

Thanks for your help.

Roy Lewallen

unread,
Aug 18, 2003, 12:52:09 AM8/18/03
to
What's really important for matching in a battery pack is that the cells
have very close to the same capacity. If one has lower capacity than the
others, the others will reverse charge it after it discharges. This
further reduces its capacity, making it go flat even sooner the next
time. The absolute cell voltage at any point in the discharge process is
unimportant.

Unfortunately, you can't discern the remaining capacity of a NiCd or
NiMH cell by looking at the cell voltage. The only way to accomplish the
matching is to discharge each cell until its energy is essentially
completely gone (i.e., to 1.0 volt at moderate drains), and select cells
on the basis of how long they take to get there. Ideally, the discharge
rate would be about the same as in the device you intend to use the
battery for. But in practice, the capacity of NiCd and NiMH cells isn't
a very strong function of the load current, so you've got quite a bit of
leeway.

Many years ago I made up a discharger for this purpose. It's basically
an adjustable constant current load with an adjustable voltage
comparator. When the voltage reaches the comparison voltage, it
disconnects the load. An old analog electric clock is used to record the
time -- the comparator also removes power from the clock. Initially set
at 12:00, it directly reads the time it took for the cell to discharge
to the specified voltage level. These days it would make a lot more
sense to use an A/D converter computer board and/or a microcomputer chip
of some sort.

Roy Lewallen, W7EL

Bill

unread,
Aug 18, 2003, 5:59:37 AM8/18/03
to
Many years ago someone associated with NASA figured out that the
life expectancy of nicads (series or parallel use) could be increased
significantly by matching the internal impedances. The recharge/charge
life did not just double, but went up by a factor of ten (approx).
I think this was published in the some NASA Tech journal. It used to
be published about 10 times a year, it might still be.

Bill Cook

Bruce W.1

unread,
Aug 18, 2003, 10:59:57 AM8/18/03
to
Roy Lewallen wrote:
>
> What's really important for matching in a battery pack is that the cells
> have very close to the same capacity. If one has lower capacity than the
> others, the others will reverse charge it after it discharges. This
> further reduces its capacity, making it go flat even sooner the next
> time. The absolute cell voltage at any point in the discharge process is
> unimportant.
>
> Unfortunately, you can't discern the remaining capacity of a NiCd or
> NiMH cell by looking at the cell voltage. The only way to accomplish the
> matching is to discharge each cell until its energy is essentially
> completely gone (i.e., to 1.0 volt at moderate drains), and select cells
> on the basis of how long they take to get there. Ideally, the discharge
> rate would be about the same as in the device you intend to use the
> battery for. But in practice, the capacity of NiCd and NiMH cells isn't
> a very strong function of the load current, so you've got quite a bit of
> leeway.
>
> Many years ago I made up a discharger for this purpose. It's basically
> an adjustable constant current load with an adjustable voltage
> comparator. When the voltage reaches the comparison voltage, it
> disconnects the load. An old analog electric clock is used to record the
> time -- the comparator also removes power from the clock. Initially set
> a_pag:00, it directly reads the time it took for the cell to discharge

> to the specified voltage level. These days it would make a lot more
> sense to use an A/D converter computer board and/or a microcomputer chip
> of some sort.
>
> Roy Lewallen, W7EL
>
========================================================

Seems the RC car community matches batteries based on a number of
criteria like capacity, internal resistance, and average voltage. The
websites I found on this don't give specifics.

Buying a $200 pieces of test equipment to measure all this is slight
overkill. I think I came up with a way. What to you think of this?

Charge all the batteries in parallel to the same voltage, say 1.5
volts. Then partially discharge them all in series. At this point
quickly measure the voltage of each cell while still under load.
Matched cells would have the same voltage.

Tarmo Tammaru

unread,
Aug 18, 2003, 12:15:10 PM8/18/03
to
Why charge them in parallel? Having them all charge the same in the real
world should be part of the matching. In fact, if you start out with all
batteries in the same state, and charge them in series, you can quickly pick
out oddball by measuring the charging voltage across each cell.

Tam/WB2TT
"Bruce W.1" <b...@NoDirectMail.com> wrote in message
news:3F4051E2...@NoDirectMail.com...

Rob Judd

unread,
Aug 20, 2003, 2:14:19 AM8/20/03
to

I'd add a prior step: Fully discharge the batteries.

Rob

Kavall

unread,
Sep 4, 2003, 1:25:14 AM9/4/03
to
Battery matching.

Here is how I would go about it. Charge the cells individually untill
their voltage starts to drop, aka. peak charge them. Measure the peak
voltage. Then hook them to a fixed load and measure their discharge
time to 0.9V. Let the cell rest. The next day repeat. Take the
measurements and compare close discharge times and close peak
voltages. If you could measure internal resistance it would be nice,
but I don't know of a simple way. In my opinion the important
qualities are.
1. Close discharge times
2. Close and low internal resistances
3. Close peak voltages

Joshua K Drumeller

unread,
Sep 4, 2003, 11:47:40 AM9/4/03
to
The internal resistance of the battery changes as the battery is discharged.
Also things like temperature, age and other factors effect it.
I think the proper way to test your battery is to fully charge it to the
manufacturer specification and discharge it at a given amp hour rate.
Compare your discharge curve with the manufacturer. Choose the same
discharge rate that the manufacturer does btw. If your curve is a little
less than what the manufacturer says then your probably ok. If it is much
less then you need a new battery.

Josh


"Kavall" <kav...@yahoo.com> wrote in message
news:a393ed14.03090...@posting.google.com...

dloyd...@gmail.com

unread,
Jan 23, 2009, 2:36:58 PM1/23/09
to
On Mon, 18 Aug 2003 00:11:14 -0400, "Bruce W.1" <b...@NoDirectMail.com>
wrote:

dloyd...@hotmail.com

unread,
Jan 25, 2009, 8:22:57 PM1/25/09
to
0 new messages