How to measure the amount of energy stored in a battery ?

[Mike Harding]

On Sun, 23 Jan 2005 09:18:42 +1100, "Phil Allison"
<philallison@tpg.com.au> wrote:

Franc Zabkar said:

An alkaline battery has a fairly linear discharge characteristic
whereas NiCd and NiMH rechargeable batteries maintain close to 1.2V
full voltage up to 80-90% of usage. They tend to go flat very
abruptly. I can see how one could reasonably estimate the remaining
capacity of an alkaline battery by measuring its terminal voltage at a
certain discharge rate, but a rechargeable battery would be more
difficult.

** Impossible is the word.

Oh Phyllis....

"Impossible"? Can we quote you on that?
Or is that a qualified "impossibility"?
I think you really mean "very difficult", don't you?

Mike Harding

 

[Franc Zabkar]

On Mon, 24 Jan 2005 10:48:25 +1100, "Phil Allison"
<philallison@tpg.com.au> put finger to keyboard and composed:

The "impossible" question is to determine, with a brief test, the SOC of a
NiCd or NIMH battery that is presented for such test. The continuous
discharge curve is of **no use** as a predictor when a battery is used
intermittently - as is usually the case.

The URL linked to by RH shows that it is possible to assess the state
of at least one cell chemistry (ie lead acid). Cadex use some
proprietary algorithm to measure and evaluate the frequency dependency
of a cell's internal impedance.

As for intermittent usage, I would think that a smart chip could
become confused under such conditions. For example, how could a
coulomb counter possibly know how much capacity has been lost to self
discharge, such as when a cell remains unused for long periods? In
such a case, ESR may be a better predictor of remaining capacity.

Cells recover when at rest for some time, voltages rise and ESRs fall -
this * voids* the accuracy of tests that rely on those two parameters. The
temperature and age of the cells also has a major effect on those same
parameters.

An old battery with reduced capacity can also confuse a simple coulomb
counter.

As for temperature effects, just wait for the battery to cool to room
temperature before testing it. <shrug>

The upshot is that a NiCd or NiMH battery placed on a brief test will show
an ESR and voltage reading consistent with near full charge when in fact the
true SOC is as little as 10 % of capacity.

............ Phil

- Franc Zabkar
--
Please remove one 's' from my address when replying by email.

 

[Phil Allison]

"Franc Zabkar" = really - irritating - fucking - idiot
"Phil Allison"

The "impossible" question is to determine, with a brief test, the SOC of
a
NiCd or NIMH battery that is presented for such test. The continuous
discharge curve is of **no use** as a predictor when a battery is used
intermittently - as is usually the case.


As for intermittent usage, I would think that a smart chip could
become confused under such conditions. For example, how could a
coulomb counter possibly know how much capacity has been lost to self
discharge, such as when a cell remains unused for long periods?

** Part of the software routine to allow for that - self discharge can be
estimated.

In such a case, ESR may be a better predictor of remaining capacity.

** So might dissecting your recent faeces.

Cells recover when at rest for some time, voltages rise and ESRs fall -
this * voids* the accuracy of tests that rely on those two parameters.
The
temperature and age of the cells also has a major effect on those same
parameters.

An old battery with reduced capacity can also confuse a simple coulomb
counter.

** Now the PITA jerk off is posting my words back to me:

" Sometimes the calculated **model** and the battery get right out of whack
and the micro has to be reset while the battery is fully cycled -
likewise if the battery becomes defective or dies of old age the model
will be wrong. "

As for temperature effects, just wait for the battery to cool to room
temperature before testing it.

** Tell that to folk using batteries in the snow or a desert.

Let the peasants eat cake ........


BTW

Ever figure out how increasing the value of a filter electro does not
increase the rms current draw - Frank ??




................... Phil