Remaing Charging Cycles on Rechargable NiMH AA 1.2V Batteries

[carloszoom3000]

Hi,

I'm trying to find out the remaining charging cyles on NiMH AA 1.2V rechargable batteries.

I'm placing a battery in series with a the parallel of 3 resistors of 2.2 ohms which are 5, 10 and 20 watts respectively.
To log the discharge and using an Arduino UNO and with some visual basic code i can see on every second, the voltage values measured (it also shows a graph).

When checking a functional battery it takes 15327 seconds before the curve shows it's discharge and it takes 9000 seconds on the non functional.

I would like to know what parameters i should take into account to know the remaining charging cycles.

Regards,

 

[jpanhalt]

I don't understand this sentence:

I'm placing a battery in series with a the parallel of 3 resistors of 2.2 ohms which are 5, 10 and 20 watts respectively.

I did a quick calculation assuming a resistance of only 2.2 Ω, not three of that value in parallel. With 2.2 Ω, you have 545 mA for 4.25 h = 2316 mAh, which is a reasonable result for an AA NiMH cell under the conditions of your experiment. Of course, if you have 3, 2.2 Ω resistors in parallel as the load, then the calculated cell capacity is quite high for that size cell and discharge rate.

As for your specific question, the number of charge-discharge cycles remaining depends on how you define a still usable cell and the conditions for charging and discharging. For example, heavy or deep discharging will shorten the battery's life. Check out Figure 3 here:
**broken link removed** (Figure 3)

Here are some other relevant discussions from Battery University:
**broken link removed**
**broken link removed**
**broken link removed**

You might also check literature for major manufacturers, such as Sanyo.

John

 

[audioguru]

Energizer also has an excellent Ni-MH Handbook And Applications Manual available for free on their website (look for Tech Info at the bottom of their home page).

 

[carloszoom3000]

Thanks for replying and also for the links.
Yes, if the battery handles a load that makes the current go high, that could reduce the battery life.
With loads that show low current, the discharging process takes much longer.
It would be great to connect the battery to an specific circuit (that does not harm the battery) and compare performaces from brand new and already used but still functional batteries, so the remaining charging cycles or estimated life of the battery divided in let's say levels that go in counts of 10 from 100 to 10% can be determined.

Regards,

 

[audioguru]

When my AA and AAA Ni-MH cells are fully charged and I measure their maximum current for a moment with my multimeter (its leads are 0.15 ohms) then a new cell reads almost 10A and an old cell with a low capacity reads 2A or less. Try it.

 

[jpanhalt]

I stumbled on this today: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=27&ved=0CEgQFjAGOBRqFQoTCP_nnsiIz8gCFQlQiAodhiMLug&url=https://www.springer.com/cda/content/document/cda_downloaddocument/9781461401339-c1.pdf?SGWID=0-0-45-1364113-p174121858&usg=AFQjCNEpm3AccQX_mmQPKCme7Ncz0MWF0Q&sig2=7p4JCc3JdeTNNtnczwhxsQ&bvm=bv.105454873,d.eXY

Its focus is electric vehicles and it discusses both current state of technology and future (e.g., silicon batteries). It discusses in depth State of Charge, but only mentions cycle life as an issue. Obviously, the fact that cycle life is not discussed in depth doesn't mean it can't be measured. However, my search that led to the link above didn't turn up any good discussions for accurately measuring cycle life. One might measure state of charge at a set time after a standardized discharge of a "fully" charged battery to estimate its health and from that you might find a loose correlation between state of health and remaining cycles for a particular brand, chemistry, and size of battery.

Regards, John

 

[MrAl]

Hi,

As mentioned by others in this thread, cycle life is a sort of foggy issue in that it partly depends on the application. If the application is medium to heavy load though then you can figure out what the approximate cycle life will be per cell, provided you take the time to do some measurements when the cell is new.

If the new cell has capacity of say 2000mAhr and 6 months later it is down to 1500mAhr, then you might expect another 6 months before it reaches 1000mAhr which is then only half of what it was when new. When you have to change the batteries manually that's when it starts to get annoying because you end up changing them and recharging twice as often as when they were new.

To measure cell capacity on discharge, you have to log the current and the time and the voltage so you know when the cutoff voltage has been reached. If you are using equal increments in time then all you have to log is the start and end times and the current measurements at each sample instant. The current sample is then weighted by the sample period, and since the sample periods would be all the same, it makes it a little easier.
For example, measurements of 1.0, 0.9, 0.6, 0.3 amps at times 1 hour, 2 hours, 3 hours, and 4 hours respectively (equally spaced by 1 hour each until the cutoff voltage has been reached) means the ampere hour capacity is:
1.0*1+0.9*1+0.6*1+0.3*1=2.8 ampere hours

and since the samples are all spaced at 1 hour each we can do it more simply:
(1.0+0.9+0.6+0.3)*1=2.8 ampere hours

The accuracy is better for shorter sample periods of course like 10 seconds each.
If you use 1 second for the sample time spacing then you would have to divide by 3600 to get the result into units of ampere hours because there are 3600 seconds in one hour.

 

[jpanhalt]

If the new cell has capacity of say 2000mAhr and 6 months later it is down to 1500mAhr, then you might expect another 6 months before it reaches 1000mAhr which is then only half of what it was when new.

That does not seem intuitively obvious to me. Do you have any data to support that proposed linear degradation?

John

 

[jpanhalt]

Addendum:
There are lots of details missing, but in general, cycle life is defined/limited by when the battery capacity is 80% of its new capacity:

(Source: https://www.mpoweruk.com/life.htm )

That is a rule we have followed for years with model airplanes. Some of our chargers will reject batteries with less than 80% capacity.

Thus, a 2000 mAh battery with a capacity of 1600 mAh would often be considered shot, not not halfway to its end of service life.

John

 

[MrAl]

That does not seem intuitively obvious to me. Do you have any data to support that proposed linear degradation?

John

Hi again,

No, in the absence of data we always assume a linear relationship until we get data that suggests otherwise. If you can find data we can do some calculations.

Why doesnt it sound intuitive to you, and what would sound more intuitive to you?

 

[audioguru]

When my Ni-MH batteries drop to about half their capacity I phone the manufacturer, tell them the date code and they send me a cheque so I can buy more.
The new ones last longer and have a longer guarantee. They also hold their charge for one year.
But some stores have them on the shelf for years then the guarantee is nearly finished.

 

[jpanhalt]

Hi again,

No, in the absence of data we always assume a linear relationship until we get data that suggests otherwise. If you can find data we can do some calculations.

Why doesnt it sound intuitive to you, and what would sound more intuitive to you?

My rationale is pretty clear from the addendum. There are plenty of data on NiMH battery charging and discharging, including in the link I included in post #6. Not having looked at any of the data is not justification for assuming there are no data.

John

 

[MrAl]

My rationale is pretty clear from the addendum. There are plenty of data on NiMH battery charging and discharging, including in the link I included in post #6. Not having looked at any of the data is not justification for assuming there are no data.

John

Hi,

Ok, so based on your reading of the data sheet or other data what do you suggest is a reasonable derating factor or method for the loss of cycle life after any period you choose under any conditions you choose?
You've stated that you dont like my quick estimation, so let's hear your version

 

[jpanhalt]

In my profession, being wrong or uninformed is far worse than not knowing the answer. When experts talk about cycle life but are unable to predict it accurately from such measures as charge capacity, we must assume there are reasons for that gap. Most people have experienced cells that were seemingly good and then deteriorate much more rapidly that other cells in a pack. That has been a problem with EV systems using NiMH cells.

The fact that a cell which accepts less than 80% of original charge is considered bad (i.e., at the end of its cycle life) by many experts is clear evidence that the loss of cycle life versus state of charge is non-linear, even if it cannot be predicted with any certainty. Your linear prediction simply ignores the available evidence and was wrong.

I am comfortable with saying that someone, e.g., the TS, might still wish to study prediction of cycle life and should not be discouraged, but they need to do background reading to avoid replowing old and infertile fields.

John

 

[MrAl]

In my profession, being wrong or uninformed is far worse than not knowing the answer. When experts talk about cycle life but are unable to predict it accurately from such measures as charge capacity, we must assume there are reasons for that gap. Most people have experienced cells that were seemingly good and then deteriorate much more rapidly that other cells in a pack. That has been a problem with EV systems using NiMH cells.

The fact that a cell which accepts less than 80% of original charge is considered bad (i.e., at the end of its cycle life) by many experts is clear evidence that the loss of cycle life versus state of charge is non-linear, even if it cannot be predicted with any certainty. Your linear prediction simply ignores the available evidence and was wrong.

I am comfortable with saying that someone, e.g., the TS, might still wish to study prediction of cycle life and should not be discouraged, but they need to do background reading to avoid replowing old and infertile fields.

John

Hello again John,

So you are saying you dont have ANY prediction then?
I am not sure what you are trying to say. If you have some idea what you would use for this then please share. Or is it that you cant find any data that shows a better estimate?

Another estimate could be based on the volumetric degradation but i'll hold off on that for now

Here is a graph i found for NiMH.

The 'digital' red line goes from 100 percent to 80 percent (20 percent decrease) in about 300 cycles, then another 20 percent in another 300 cycle, then another 30 percent in another 300 cycles. You can see there is a rough linear relation ship here.

The other curves are not quite as matching, but then again they dont show the entire drop in capacity either so it's hard to tell.