More info about NiMH accus

[izua]

I've bought some NiMH accus for my future robot. They're rated at 2.5 Ah, and 1.2 V. A week after charging (i haven't used them), I measured their voltage with no load, and it said 5.4. I thought they would discharge faster. If i hook them up to a modded servo, my multimeter (cheap) says anything between 4.5 and 5 (it reached 2.78 a few times:P).

When should I consider them discharged? If at max, after a complete load an accu is close to 1.5, what should be the minimum value?
Also, what effects should the charger have over the accus if it's a one-hour charger? The accus got hot after the first (and only) charge. Both are produced by energizer.

[Soeren]

Hi,

They're rated at 2.5 Ah, and 1.2 V.

Some manufacturers says 1.2V, others say 1.25V - they aren't different, just a matter of definition.
Sadly, some cells are considerably less than their marked capacity (mAh).

A week after charging (i haven't used them), I measured their voltage with no load, and it said 5.4. I thought they would discharge faster.

Don't mix up single cells and batteries, if you wanna talk about voltages.
When you charge a nickel based cell, it gets up to 1.40V (or 1.45V).
When disconnected from the charger, it'll fall to around 1.35V.
For the larger part of its discharge curve, it will be around 1.20V to 1.25V.
They are considered flat at 0.9V to 1.1V, depending on the discharge current. (They will raise their apparent voltage when the load is disconnected).

An unloaded voltage reading on a charged but unused cell is worth squat.

If i hook them up to a modded servo, my multimeter (cheap) says anything between 4.5 and 5 (it reached 2.78 a few times:P).

Again, please use either the voltage for a single cell, or the voltage for the battery (and if so, be sure to state the number of cells) - don't mix them freely

A nickel based cell CAN be drained down to 0.0V without damage. However, if it's part of a battery, all cells won't reach 0V concurrently and so the weaker cell(s) will be charged in reverse as long as a load is connected and the remaining cells holds a bit of charge.
A safe measure of End Of Charge (EOC) is between 0.9V and 1.1V depending on the discharge current - the larger the current, the lower the EOC-voltage.

When should I consider them discharged? If at max, after a complete load an accu is close to 1.5, what should be the minimum value?
Also, what effects should the charger have over the accus if it's a one-hour charger? The accus got hot after the first (and only) charge. Both are produced by energizer.

Fast chargers, if properly designed, is the best way to charge nickel based cells, but if poorly designed, can break down a cell in a very few charge-discharge cycles.
If the cells are too hot to hold (i.e. >35°C) they are being damaged.

Don't start recharging them before you have used a good part/most of their capacity, if you want them to last.


Buying battery cells way ahead of actually using them is a waste, so better start building

[izua]

I'm trying to build, but i'm still figuring out why the two servos spin at different speeds if both motors are supplied with +5 VDC (electronics removed)

Ok, so, in order to safely measure the battery state in the robot, how should I do it? Definitely the sum of series battery ain't worth that much, if they discharge at different rates. The question is, can they discharge at extremely different rates, if they are bought at the same time?

I doubt the charger is of such a bad quality. It's got a cartload of electronics x 4 (interesting enough, all charger channels have controllers, some inductors, and a bunch of chips, all multiplied in 4, hehe ) and the whole set (charger + accus) was ~ $60. But that's later to decide, anyway.

[Soeren]

Hi,

I'm trying to build, but i'm still figuring out why the two servos spin at different speeds if both motors are supplied with +5 VDC (electronics removed)

You won't find two motors that are exactly alike, but they should be somewhat equal if they're from the same production batch.

Ok, so, in order to safely measure the battery state in the robot, how should I do it? Definitely the sum of series battery ain't worth that much, if they discharge at different rates. The question is, can they discharge at extremely different rates, if they are bought at the same time?

If you treat the cells properly and they're not very old, they should discharge at the approx. same speed (within reasonably tolerance). After each charge discharge, the difference will be more pronounced, if you don't equalice the cells at least now and then. If your charger has separate circuits for each cell and provided each is adjusted to the same specs, you should be in the green about that.

I doubt the charger is of such a bad quality. It's got a cartload of electronics [...] and the whole set (charger + accus) was ~ $60.

This reminds me of the transistor radios of the sixties, where they sometimes had a sign saying "2 transistors" and when opened, one of the transistors just sat there, connected to nothing - just a market ploy
A $h!tlo@d of components is not allways a sign of quality, more often a sign of poor design - simplicity rules
And sadly, the pricetag doesn't allways reflect the quality either.

That said, I don't suspect your charger to be bad, but if the cells gets too hot, it WILL shorten their life.

Just measure the terminal voltage under load and charge when the voltage goes below 1.1V/cell if the current is not much above C/10 (i.e. 200mA for a 2000mAh cell). If the current is around C/2 or higher, use 0.9V/cell. Anything in between, just interpolate the value.

[izua]

yes but no servo or pic will run on a single cell, so i have to wire them in series. when i'll do the measuring, i'll have to wire them individually, to break the series. a circuit that will do this (with individual sockets for each cell) that can toggle between series and individual connection is too complex, plus the mcu has to be powered from somewhere else.

can't i measure the sum? will that be efficient?

[Soeren]

Hi,

yes but no servo or pic will run on a single cell, so i have to wire them in series. when i'll do the measuring, i'll have to wire them individually, to break the series. a circuit that will do this (with individual sockets for each cell) that can toggle between series and individual connection is too complex, plus the mcu has to be powered from somewhere else.

can't i measure the sum? will that be efficient?

You just measure the serial voltage of course. If you've got say 4 cells, just multiply the values accordingly (terminal voltage 0.9V/cell to 1.1V/cell is thus 3.6V to 4.4V for the 4 cells serially connected). If you could measure them one cell at a time, the End Of Discharge voltage wouldn't matter very much, since you should just stop a bit before any cell reached 0V and vent into reverse charging.

[izua]

Ooh, i thought measuring the sum doesn't work.
Probably on some future bot I'll do this, measuring individual cells. But that's a bit too complex for now, considering I can't even drill the holes right, in the wheels