Ni-MH charge efficency at low rate (0,1mA etc)

Hello, does anybody know how well Ni-MH (Eneloop) store energy when charging at low rates around 0,05-1mA? I am designing something with quite low power consumption and I want to add solar panel to the device to make it really maintnenace-free. I can easily measure solar panels output etc and choose adequate one but I have no data on how efficent Ni-MHs are when we are talking about these low currents - will they even charge? Or will the efficency be around 50%? Does anybody know something about it? Thanks a lot. I am also thinking about adding supercap (4F/2,5V or even more) to store solar energy and battery will be there only as backup in cloudy short days in winter, but I would rather choose solution with just AAs.

i doubt they would charge at that current at all

If your power requirements are low, you could try using a rechargeable watch battery, like a CTL1616. It charges fine at 0.1 ma or even lower rates. I have one in my watch, so it needs a very small solar panel to charge it. It's lithium-ion, not NiMH. I think it only stores something like 18mAh. That's lots of power for a watch or small electronics, but YMMV.

Yeah I was tinking about lithium based cells as well but this would require some charging circuit and I am afraid it is too complex for my use - also I am not designing brand new device - more like tuning existing one . Btw I think CTL1616 is not standard li-ion. It is some sort of lithium based rechargeable but not regular li-ion. I found it has nominal 2.3V and is lithium-cobalt-titanium based. I never heard much about it. I will remember this for further projects but for this application I would like to stick to standard AA cells in case it fails in 5 years etc it is easy to replace with random AA cell even by technically not educated user.

I think it calls for test - 2 identical Eneloops, discharge them, connect one to constant 0.5mA source and let it charge for one month. After this time discharge again both of them and check discharged capacity

Please post your test results. Sounds interesting to know just how little a trickle charge is needed. Of course, if you're using a AA eneloop, it's going to take half-a-year to charge it fully even if it's 100% efficient. At such a low charge rate (and similarly low usage), you may be better off just sticking a lithium primary in your device and forget about charging anything. It should last years.

czAtlantis said:

I am also thinking about adding supercap (4F/2,5V or even more) to store solar energy and battery will be there only as backup in cloudy short days in winter,

Click to expand...

Makes no sense for a solar-powered application with a small load like this. Supercap alone if that does the job, otherwise a battery. Supercap+battery just adds another failure point. Such a combo would normally be used when there's large power fluctuations that a battery can't handle. For example re-generative breaking a vehicle, where a (momentary!) influx of power is too big to charge a battery directly with.

Probably the actual charging process for a NiMH is efficient at low mA's, but there's the continuous self-discharge. The lower the charge current, the larger a portion is wasted on compensating for self-discharge.

So you're looking for a low self discharge chemistry, and a small cell at that. For example a small size LiFePO4​ cell, or (to keep things simple) an AAA Eneloop Lite? An AA size Eneloop would just waste more of your input current on the battery's self discharge.

Thanks for answers. I don't think self-discharge is big deal - Eneloops claim 70% remaining capacity over 5 years now - that means 600mAH is lost in 5 years - if my calculations are correct it means avereage discharge at 14uA. My application draws on avereage around 200uA- it also means it is not so "extra low current" - I mean it is not like LCD clock with 30uA discharge current where you change battery every 7 years. At 200uA it will discharge Lithium primary (2500mAh) in year and half. But also I take this as "project" - it means this is primary about experience etc - it doesn't have to be cost effective or you know what I mean...

Supercap alone can't be used I think - it doesn't have enough capacity to safely last over night (well maybe 10F one will work) but main reason I dont want to use it is during cloudy winter and/or solar panel covered by snow it would fail and I am not sure it would recover from low voltage into operation...I was just thinking about some simple "diode logic" to drain energy from lithium primary only when supercap voltage drops below battery voltage during long night/snow cover...but I think best idea would be test Eneloops at slow charge.

In few days I will start the experiment, I have to build some constant-current circuit - 0,5mA or 1mA over 14days or 1month (depends on my impatience :-D ) to Eneloop+Eneloop Lite and maybe 10mA over 7days for Eneloop classic only just to compare.


On the other hand how well Eneloops tolerate overcharging with small current - like 50mA? As far as I know it is impossible to detect full battery with this non-stable charging because Ni-MHs are charged with "current not voltage" unlike Li-Ions. Only thing I can think of is add zener diode or voltage reference set to I don't know 1,45V? So any current above this voltage will dissipate instead overcharging battery. But in this case I really dont know what the voltage of overcharged battery will be.

Why am I asking - I want to have solar panel scaled large enough to cover daily power need in not ideal case scenario - with cloudy winter. But of course this device will work during summer too and when direct sunlight hits the panel the current will rise from cloudy for example 2mA to 100mA easily.

If the battery retains charge well enough over long periods (like lsd Eneloop), you could size the solar panel + battery such that the battery averages things out over 1-year cycles, as opposed to day/night cycles. Like, in the summer mostly charge battery, in winter mostly discharge. Could make for a much smaller solar panel and lower charging currents (to the point where a NiMH or LiFePO4​ cell absolutely doesn't give a f$%k about being 'overcharged').

200 uA average = 4.8 mAh / day = 1.75 Ah / year. Looks like a good number to work with...
(as long as you keep reduced battery performance @ low temps in mind)

A 50mA trickle charge is about C/40 for a AA Eneloop. That's probably a safe current to use (considering it would only be for part of the day), but I wouldn't go higher. Well, you could probably go as high as 200mA and your battery would still last years, but it would likely shorten its life. If you want it to last 10+ years, probably better to stick to C/40 or less.

czAtlantis said:

In few days I will start the experiment, I have to build some constant-current circuit - 0,5mA or 1mA over 14days or 1month (depends on my impatience :-D ) to Eneloop+Eneloop Lite and maybe 10mA over 7days for Eneloop classic only just to compare.

Click to expand...

It'd be adequate to just use a spare PSU and a suitable resistor - some suitable old phone charger or USB-output power supply, or any old regulated wall wart of known voltage.
If I didn't seem to have a constant lack of permanently-unneeded sockets, I'd be tempted to try it myself.

czAtlantis said:

Why am I asking - I want to have solar panel scaled large enough to cover daily power need in not ideal case scenario - with cloudy winter. But of course this device will work during summer too and when direct sunlight hits the panel the current will rise from cloudy for example 2mA to 100mA easily.

Click to expand...

The available current will be higher, but if the panel nominal output voltage is high enough, adding some resistance between panel and cell could effectively limit the current to the cell at full illumination without drastically reducing current in winter, since the cell voltage is relatively constant with varying illumination.

A new D alkaline capacity is about 15000mAh. It will last 75000h =3125 days = 8.5 years. My expectation is higer for a nearly dead D cell alkaline (from the recycling bean =more than 7000mAh ) than to a constantly toped enelope. Just seal it against leak.
Make it as simple as possible, but no simpler .
A. Einstein

Perhaps consider to use one of those very long life, low discharge cells like Tadiron sells and skip the charging ?

czAtlantis said:

Hello, does anybody know how well Ni-MH (Eneloop) store energy when charging at low rates around 0,05-1mA?

Click to expand...

The charge storage ability of NiMH seems to depend on two factors: the existing state of charge, and the rate of self-discharge.

A good cell like an Eneloop that is less than 80% charged will store close to 100% of the charge supplied to it. The charging
efficiency only drops off when the cells are near full charge.

Similarly, an Eneloop has a very low rate of self-discharge (<< 1 mA). So as long as you charge it at a rate of 1 mA it should accumulate charge until it reaches a moderately full state of charge.

However, this is based on theory, and it would be very interesting to see the results of a practical experiment.

Sorry about late reply, i was busy with work. First I wanted to build some true regulated 1mA supply but..well..I am lazy and I realised the voltage of AA cell will change only slightly during charge - so I used just constant voltage power supply, variable resistor and current meter. I set te resistor to exactly 1mA and let it work.

I used 2AA classic Eneloop cells, discharged them in MAHA C9000 at 100mA and let them sit for day. I discharged them again. Difference in capacity between these 2 cells was only 1mAh so I believe they are really identical. Now I let one cell charge at 1mA for 10hours (extra short test just to prove it is working). After 10hours I discharged both of them:

Idle cell: 9mAh
Charged cell 13mAh

Soo...it charged slightly....using 9mAh from idle cell as "ghost energy " , it charged only 4mAh which is 40% charge efficiency.
Now it is time for at least one week test + add Eneloop Lite to the test


Another test(86hours at 1mA):
idle cell: 9mAh
charged cell: 78mAh

78-9=69mAh
efficency: 80%


Another extra low current test (0,1mA/282hours):
idle cell: 17mAh
charged cell: 38mAh

38-17=21mAh
efficency:74%


I think it looks just fine