Trickle Charging NiMH LSD AA Cells Test

[WalkIntoTheLight]

Looks like just the end point is off. Yeah, probably some smoothing, like a spline where the end doesn't match too well. Raw data might look too ugly for a graph, and isn't much use anyway if the error bars make it noise.

 

[AA Cycler]

See image/development above. Something or somebody is smoothing the curves...

Good observation NiMHi!

I draw all lines with bezier smoothing otherwise they would be too jumpy. Test yourself, here are the raw data - http://aacycler.com/post/trickle-charging-lsd-cells.csv.

Cheers,
AA Cycler

 

[hc900]

Hi, AACycler is doing a new test! This time to see the trickle charging effects on NiMH AA LSD cells.

Link to test: http://aacycler.com/post/trickle-charging-lsd-cells/

I'm curious how the lines will progress. Currently it looks like the 100mA trickle charge is even going up in capacity.

Anybody dares to make a prediction?

Perhaps someone also got other test data or papers on the effects of trickle charging LSD cells?

Things are a little distant between the trickle charge of the SKYRC MC3000, which delivers a true constant current. But this is the only one who does things so well.

The Maha C9000's Trickle Charge is pulsed at 2000mA
While that of the Opus BT-C700 is pulsed at 700mA

The thermal effects on 100 mOhm are really very different, much suggests that with AAA batteries shooting pulses with peak at 2C or 4C is not very good for their health.

How about moving the survey on these much more used chargers?

 

[NiMHi]

Good observation NiMHi!

I draw all lines with bezier smoothing otherwise they would be too jumpy. Test yourself, here are the raw data - http://aacycler.com/post/trickle-charging-lsd-cells.csv.

Cheers,
AA Cycler

Ah yes, I understand now. I tried to make a graph with no smoothing (see below).

@hc900 I see what you mean. The Panasonic BQ-CC55 charger is even worse with 3A pulses every 10 seconds (see link for HKJ review)!

I think the possible different effects of pulse or constant current trickle charging might need one day its own separate test(s).

 

[IonicBond]

Generally trickle charging a discharged battery often enough damages it. Since the charge is not efficient, only a small fraction of the available current does the charging - most of the rest of the dinky current goes into chemically heating the battery, even if you can't feel it.

This damage from enefficiency is a problem with many batteries regardless of chemistry. Ever see someone try to trickle-charge a dead car battery with a wall-wart? They take days or weeks to charge (if the charger doesn't do a safety timeout first), and have only achieved a faux surface-charge. Not enough efficient current to get deep into the active material. But in the meantime, that active material was "slow cooked" with a different chemical process (heating mostly, even if miniscule) than a normal charge.

So yes, I think that if you are going to do this at all, then high-current pulsing is the better answer chemically.

But now we run into the complexities of the "end game" and deciding when to realize an end-of-charge scenario so all your pwm pulsing hard work turns to naught.

I guess it's an application question rather than a charge question. If using a more efficient charge rate (usually 0.5 to 1C from most manufacturers, not the 16 hour slow charge technique) along with a reliable end-of-charge detection system will result in good battery health, what's the point of using longer term pulsing if the net result is the same?

 

[IonicBond]

Here's an application for you - with solar in mind....

Due to the problem of a non-steady source of power causing resets of battery chargers, and the problems that poses, perhaps try this for nimh:

1) Use pulses to charge cells from 0.5 to 1C to get as much energy in as fast as possible initially. Because the sun may be varying in intensity, instead of slow-cooking cells, use pulses - to a point:

2) Have the circuit measure the cell voltage regularly. When it reaches say 1.2v, then skip out of using pulses, and just slow charge the rest of the way. Perhaps designed for 1/10th C.

3) If the cell measures 1.4v, then still use a slow trickle, but this time at 1/100th C.

4) Overdischarged cells? If the circuit measures a cell at less than 1v, then use the trickle rate of either #2 or #3 to gently bring the cells out of the discharge knee, before applying the higher-current pulses. This low rate would also help prevent damaged cells from getting out of hand.

This might be a way to handle the variables of an unsteady source like the sun, and keep the batteries healthy without having to resort to any major EOC event.

Now there's a handy use for an RPI, arduino, etc etc..

 

[hc900]

@hc900 I see what you mean. The Panasonic BQ-CC55 charger is even worse with 3A pulses every 10 seconds (see link for HKJ review)!

I think the possible different effects of pulse or constant current trickle charging might need one day its own separate test(s).

The great work of AACycler is really excellent, my observation is not a criticism against him, rather an aid to better understand how to improve his very useful investigations.


I realize that many things written by others, although true in those times, are no longer valid with the new rechargeable batteries.
The AACycler site provides many explanations, proven by the facts, of how the batteries are destroyed using certain charging systems, In principle has confirmed almost all my fears about how many of my Ni-Mh batteries have died.


Question:
Can we say that the new batteries are more fragile than those of 10 years ago?


Question for AACycler:
I had many of the batteries you tested, but maybe you did not write how many had died before the end of the tests or DOA, it seems to me that some types had tested 4 or 3 pieces, then on the charts of your site there are only 2 left.
Of these here on 8 that I had bought, 4 were good and 4 were very poor:


http://aacycler.com/battery/aaa/everactive-silver-line-800/

 

[AA Cycler]

Question for AACycler:
I had many of the batteries you tested, but maybe you did not write how many had died before the end of the tests or DOA, it seems to me that some types had tested 4 or 3 pieces, then on the charts of your site there are only 2 left.
Of these here on 8 that I had bought, 4 were good and 4 were very poor:
http://aacycler.com/battery/aaa/everactive-silver-line-800/

Hi hc900,

beginning this year (2018) I changed by test setup and for sake of consistency I am re-testing the cells that were done with the old setup. Usually I test only 2 cells. If there are 3 lines in the graph it only means that re-testing is in progress. Once the 3rd line is finished I remove the 1st one, ending up again with 2 lines. This is how I gradually phase out old data and replace them with new data. Sorry if it creates confusion

And yes, I have the same experience with everActive Silver Line 800 cells as you. They are shite, and one even shorted during my tests Other bad cells would be the Camelion Ni-MH ones with occasional DOAs, but much more inconsistent capacities.

Cheers,
AA Cycler

 

[NiMHi]

So far, after around 50 cycles, the capacity and internal resistance haven't changed that much. I expected worse results. Let's see what will happen when we reach 100 cycles.

 

[WalkIntoTheLight]

Yeah, if I look at AACycler's Eneloop tests, IR starts to go up around 200 cycles. But that's with normal charging. The trickle-charging should see that much sooner, if we're to believe the general wisdom that trickle-charging is bad for LSD cells. I'm a bit surprised it hasn't shown up yet, but I suppose 50 cycles isn't that much.

 

[NiMHi]

Very interesting results after more than 100 cycles:

from: http://aacycler.com/post/trickle-charging-lsd-cells/


It looks to me that the trickle charging a LSD NiMH AA at a moderate 20mA is not that deadly as I read/thought.

 

[WalkIntoTheLight]

Very interesting results after more than 100 cycles:


from: http://aacycler.com/post/trickle-charging-lsd-cells/


It looks to me that the trickle charging a LSD NiMH AA at a moderate 20mA is not that deadly as I read/thought.

Yes, I'd say that is "myth busted".

 

[NiMHi]

It is not looking good for the 100mA trickle charge (battery)....

 

[NiMHi]

Thank you for the cool test AACycler.

Previous tests with these batteries showed 261 and 288 cycles to 100mOhm impedance.

Quick overview/guess:

Trickle - cycles (up to 100mOhm)
100 - ~160
050 - ~190 (guess)
020 - ~210 (guess)
000 - ~270

Based on my guessing that will mean that trickle charging will give 75% or less of the possible cycles. Wow!

Lets see how it will go.

 

[WalkIntoTheLight]

Very cool test, which I believe is coming to and end shortly!

And the winner is....

Don't trickle-charge. But, it doesn't seem to get bad until you're up to 100mA of very-long trickling, and using more than 150 cycles. That's not as bad as what we've been told over the years. This test is very enlightening.

 

[Grijon]

...That's not as bad as what we've been told over the years. This test is very enlightening.

Agreed!
Thank you to those doing the work and sharing!

 

[NiMHi]

This interesting test has come to an end, AACycler posted this:

Conclusion (Ongoing)

This test took 82+ weeks (1 year 7 months) to finish.
I am going to assess the impact of trickle charge based on how the internal resistance developed in each cell. The controll cell without trickle charge defines the baseline.

Cell	Cycles to 100 mOhm	%	Change
0mA control	248	100%	0%
20mA test	210	85%	-15%
50mA test	190	77%	-23%
100mA test	160	65%	-35%

Even the smallest 20mA trickle charge causes 15% drop in cycle count.
Conclusion: don't trickle charge LSD NiMH cells, or do so with less than 20mA current.
Cheers,
AA Cycler

Thank you for this nice test, AA Cycler!

[edited my own wrong calculation]

If you look at total mAh extracted it looks less bad. For 20mA trickle only ~4% less extracted Ah and for 80% of the cycles (up to 100mOhm) you have more (~8%) or at least the same amount of capacity as the 0mA trickle per cycle. So 20mA trickle doesn't sound so bad... Am I missing something or calculating something wrong?

Cell	Cycles to 100 mOhm	total Ah	normalized
0mA	248	488	1,00
20mA	210	468	0,96
50mA	190	435	0,89
100mA	160	384	0,79

Data from here: http://aacycler.com/post/trickle-charging-lsd-cells.csv

 

[sbj]

If you then assume that hardly anyone would "trickle" charge their cells for 44 hours, but maybe only 2 to 4 hours?
Then one could assume that the damaging effect of "trickle" charge would be much less.

 

[HarryN]

To me, it would seem quite natural for many AA battery users to just leave the cells in the charger until needed.

Battery enthusiast might be more cautious, but I am not sure that most consumers would pay attention - or care.

Nearly all of these AA cells are providing 100 - 200 cycles vs the 1000 cycles indicated by the literature.

 

[WalkIntoTheLight]

In any case, a gentle trickle for a few hours doesn't seem very hard on the cells. I wouldn't do it myself, since I'm not interested in squeezing in maximum capacity, but I wouldn't call it a bad way to charge, either.

And yes, the cells don't get anywhere near the advertised cycles. But if you read the way they do the cycle tests, it's only partial cycles. If NiMH are anything like lithium-ion, it's the bottom and top of the charge/discharge cycle that is the most damaging. Maybe that partially explains why trickle charging is harder on the cells: you're essentially charging to the very top of the cell capacity.