Unusual battery tester.

[NW27]

Hi All,
This is my first post to this forum.
I need to make a Battery tester for 1v..12v batteries.
Basically I'm looking to place a short circuit (mosfet?) across the battery and measure a) The current and b) the voltage. I'm looking to determine the internal resistance of the battery.
I'm also going to measure the temperature rise of the battery.
C size batteries using a current clamp meter give approximately 160 amps.

Thoughts?

Neil.

[Delta]

C size batteries using a current clamp meter give approximately 160 amps.

Thoughts?

Neil.

I think you have made a mistake!

[Dago]

You do not have to short-circuit the battery to figure out the internal resistance... Hint: Ohms law

[NW27]

Why do you say i made a mistake?
Have you shorted out a c size nicad?
You are right about ohms law. How do you measure the internal resistance of the battery? Without applying an external current limiting resistance?
Neil.

[MK]

try using a pulsed current of known current and measure the voltage drop. shorting cells can sometimes be bad for batteries.

[NW27]

At the moment I'm just applying a direct short circuit to the battery to determine its temperature rise and what it physically does.
Determining the internal resistance is stage 2 of the measurements.
I only get one go with the batteries, after the above torture they are stuffed.

[NW27]

I  would expect the internal resistance at say 2 amps to be different to 50 amps or greater.

[Siwastaja]

Internal resistance is not a constant over all operating conditions.

It makes more sense to measure it at some relevant conditions. Very low-ohm external short is seldom relevant.

Don't go that far (direct short).

The most extreme test that makes some sense is to apply external resistance that is low enough so that the battery terminal voltage drops in half. This is the point where the battery supplies most power. This is also the point where the battery efficiency is 50%.

But I'd recommend that test the internal resistance using two different loads that are both near the intended operating conditions. Then calculate the internal resistance as R = (V2-V1)/(I2-I1).

[Fungus]

You are right about ohms law. How do you measure the internal resistance of the battery? Without applying an external current limiting resistance?

a) Measure battery voltage with no load.

b) Measure battery voltage with a suitable load applied, eg. 250mA for a C cell.

The difference between (a) and (b) tells you the resistance (with Ohm's law).

[NW27]

Internal resistance is not a constant over all operating conditions.

It makes more sense to measure it at some relevant conditions. Very low-ohm external short is seldom relevant.

Don't go that far (direct short).

The most extreme test that makes some sense is to apply external resistance that is low enough so that the battery terminal voltage drops in half. This is the point where the battery supplies most power. This is also the point where the battery efficiency is 50%.

But I'd recommend that test the internal resistance using two different loads that are both near the intended operating conditions. Then calculate the internal resistance as R = (V2-V1)/(I2-I1).

Thanks, I will give this a go and compare to the manufactures data sheet.
This will however mean i need more batteries if I'm testing alkaline cells as performing a short on this heavily loaded battery will have taken some of the charge.

[NW27]

What about applying the short in an electronic fashion?
As I mentioned, c size nicads can generate a lot of current. Far more than a normal relay can handle.
Hence why I was thinking of a mosfet rated at a few hundred amps.
Similar currents or better in Nimh batteries or SLA batteries.

[HKJ]

a) Measure battery voltage with no load.

b) Measure battery voltage with a suitable load applied, eg. 250mA for a C cell.

The difference between (a) and (b) tells you the resistance (with Ohm's law).

Except on a lithium battery where it will give a wrong value, the best is to measure with two different currents near the working current.

[Raj]

use multimeter (for amps, use some kind of hail sensor or other current sense attachment)

[tszaboo]

And that is not how you should measure IR of a battery. You feed in a 1KHz or 10KHz signal, calculate RMS current and voltage values and divide. Well, at least if you want to ignore charge transfer resistance. Also, you need 4 wire measurements for the voltage. In any case you dont need a dead short.

[HKJ]

And that is not how you should measure IR of a battery. You feed in a 1KHz or 10KHz signal, calculate RMS current and voltage values and divide. Well, at least if you want to ignore charge transfer resistance. Also, you need 4 wire measurements for the voltage. In any case you dont need a dead short.

Both AC and DC impedance/resistance can be interesting to measure, they are not the same.

[Siwastaja]

And that is not how you should measure IR of a battery. You feed in a 1KHz or 10KHz signal, calculate RMS current and voltage values and divide.

Do you really need to spread this misinformation myth stuff over and over again? We have gone through this once, it should really be enough. Please refer to those posts, I spent a lot of time trying to explain it to you. I'm happy to help if you have any questions regarding this matter. Or, if you really disagree, please provide proper counterarguments instead of ignoring the matter and going on trolling.

Well, at least if you want to ignore charge transfer resistance.

Why the heck would you want to ignore that? That's the whole point of measuring battery resistance (at least to anyone outside cell manufacturing / chemistry research), to see the resistance, not only a part of it, d'oh! You couldn't give any reason last time. Still waiting.

Again, 1kHz or 10kHz AC impedance (which should not be called "internal resistance") is used for battery analysis to predict changes in chemistry and analyze the internal chemistry or structure, but it's of no use for the user / battery system designer, at least not as a parameter called "resistance". Also, 1kHz/10kHz AC impedance is very outdated, primitive way for what it was for; today, we have EIS analysis, which is practically a sweep from a few Hz to maybe 100kHz, measuring complex impedance, not only magnitude but phase, too.

Also, you need 4 wire measurements for the voltage.

This is true.

[cat87]

Ok, so maybe I can squeeze in my 2 cents to this topic.
At work, we do some some battery tests on some Lithium-Iron-Phosphate batteries, 1Ah capacity and of course, the internal resistance of the battery is important, because we can know it it is healthy or not.
To do this, the voltage of the battery is  measured (V_{NO_LOAD}) Then, a 1A load is connected to the battery. After about a second, when everything is settled down, the voltage is measured across the load (V_LOAD). The calculation of the resistance  will be: (V_{NO_LOAD} - V_LOAD ) * R_LOAD / V_LOAD

Of course, the resistance , as some have said, will vary according to the temperature and the state of charge of the battery, so the calculated value is useless without knowing the characteristic of the battery, either from a datasheet from the manufacturer, or from doing real measurements across a range of temperatures and states of charge. For us, it's the latter.

The switching of the load is done by a MOSFET. 1A isn't that much, but paralleling some big ass power MOSFETs can make possible  the switching of some really large currents, of 10's of amps up to about 100 Amps or so.

We've never done any AC measurements, because, for our purposes that isn't useful at all.

[tszaboo]

And that is not how you should measure IR of a battery. You feed in a 1KHz or 10KHz signal, calculate RMS current and voltage values and divide.

Do you really need to spread this misinformation myth stuff over and over again? We have gone through this once, it should really be enough. Please refer to those posts, I spent a lot of time trying to explain it to you. I'm happy to help if you have any questions regarding this matter. Or, if you really disagree, please provide proper counterarguments instead of ignoring the matter and going on trolling.

It is how you measure battery IR. End of story. That is the standard, well, if you ignore the crazy german car manufacturers, who defined to short a battery with 10C.
And Im sorry that I cannot give you any more specifics. I worked on battery testers for the last five years (not anymore), and I had access to information under NDA, that you dont have access to. Also, spent months making equipment to be able to measure IRs in the microohm accuracy, you know, 1000+A 1KHz is not done just for fun.
The DC method is fine. Customers used it. Ive also seen customers using a slab of wood with some 100 Ohm resistors and battery holders butchered on it use for "battery testing".  And we are talking about companies worth billions here doing that. So yeah, there are links saying that the DC method is good, but it isn't because it is SOC dependent, time dependent, and it is not measuring what it supposed to in the first place.

[nctnico]

And that is not how you should measure IR of a battery. You feed in a 1KHz or 10KHz signal, calculate RMS current and voltage values and divide. Well, at least if you want to ignore charge transfer resistance. Also, you need 4 wire measurements for the voltage. In any case you dont need a dead short.

Both AC and DC impedance/resistance can be interesting to measure, they are not the same.

True. With a proper data set from aging batteries you could even use the complex impedance to say something about how healthy a cell is. But the proper method of testing largely depends on what the purpose of the measurement is. Usually 18650 cells are specified at 1kHz AC and DC using 1A and 4A discharge currents.