Continuously charging NiMH - best approach?

[jirij]

Hello,
I've built a simple project based on an idea from a video, https://www.youtube.com/watch?v=wVnAH17f4jg, essentially using simple resistors to limit charging current to ~20mA (C/100) or even lower, to continuously trickle-charge many NiMH batteries 24/7, so there are always fully-charged batteries available when you need them for some device.




The long charge time (in days/weeks) is compensated for by there being a lot more batteries than needed for any one device.

The project was based on the idea that slow trickle-charge is safe for NiCD/NiMH due to self-catalysis, where any generated gas is recombined back to electrolyte, as long as the current is low enough.

And many legit-looking articles, blogs, and anwers on the webs suggest that this is a good idea.

...

However, turns out the world is not simple.

A post on electronics.stackexchange.com suggests that modern high-capacity cells should never be trickle-charged. Not even at very low currents corresponding to C/100 or C/300:

When cell capacities exceeded about 1800 mAh AA packaged cells this mechanism was removed to allow more room for active material.
As a consequence modern NimH cells must not be trickle charged even at very very low rates. It is posisble that modern low capacity cells still have this mechanism inlcuded, but this is not certain and should not be relied on.
Most manufacturers indicate cells should not be trickle charged at all. A very few say a very low level of trickle charge (maybe C/100) may be applied for an extremely short period.

The poster suggests that this depends on the specific manufacturer chemistry, and another, older, post suggests this varies even across Eneloop battery ranges.

I tried to find what the specific change in chemistry is, how can one tell if a battery is safe to trickle-charge or not, but the only mention is "LSD", or low-self-discharge, and several sources (blog posts, forum posts) seem to agree that LSD is either always damaged by trickle-charging, or at least they LSD batteries have significantly lower recombination capabilities.

Since I'm using IKEA LADDA batteries that some sources mention to be LSD, I looked around and found out that IKEA's own charger seems to use PWM for trickle-charging, with a frequency of around 0.5 Hz and 10-20% duty cycle.

...

However, an interview with an Eneloop engineer suggests, amongst other interesting things (over-discharge being far worse than over-charge), that it's not the current that's the problem. It's the continuous 24/7 charging part.

36.  Does trickle charge damage regular eneloops? lite and pro?

Trickle charge doesn’t by itself, but continuous charge does.

This could explain why the charger is doing PWM when trickle-charging.

Looking into some datasheets for ICs, U2402B suggest a similar thing with its charging logic (~20s periods for top-off charge pulses and ~20 minute periods for trickle charging), though some chargers/ICs seem to completely turn the charging off after a few hours of trickle-charging.

...

Even here on the forum, people seem to generally mention that trickle-charging NiMH is not good, though it may be less damaging if done with a lower-than-maximum voltage and with a voltage-limiting by the charger. But there is no universal consensus from what I could find.

There seem to be some special batteries designed for trickle-charging (as used by cordless phones), though it's unclear where these have genuinely different chemistry, or are simply de-rated regular high capacity cells.

...

So, is there an off-the-shelf charger, an IC, a DIY project, or even just an algorithm I could implement in an MCU that would keep a set of NiMH batteries always ready-to-use, without degrading their lifespan significantly?

Is really the best approach to do -dV/dt (or dT/dt) at a higher current (so one can detect the voltage drop or change in temperature), and then switch to a once-every-hour pulsed "trickle" (maintenance) charging? Even if I don't need the batteries charged quickly?

Thanks!

[wraper]

Don't do that unless you want to throw out those batteries in ~1-2 years. Grey LADDA are eneloops in disguise made at the same Japanese factory as Panasonic. In fact most of Pananonic eneloop are now made in China instead of Japan, so it's even preferable to buy LADDA unless you physically check where Panasonic eneloops were made. They are LSD cells that barely self discharge. Constantly trickle charging them is absolutely stupid as eventually you'll find those batteries became garbage with abysmal ESR and cannot supply high current. Batteries that otherwise would be perfectly good with 95% charge left after a year of sitting in a drawer.

[IanB]

I think continuous trickle charging has always been bad for nickel chemistry batteries, it's just a question of how bad, and how long before they are damaged.

As mentioned by wraper, constant trickle charging of cells like Eneloops is utterly pointless. I have had devices like an old digital camera sitting on a shelf unused for 3 to 5 years with Eneloops inside, and it still fires up and works if I switch it on.

I can believe high capacity cells get damaged faster than lower capacity cells, but they all get damaged eventually.

[coppercone2]

you can trickle charge them but put it on a timer so it turns on once a year to top it off. needs a watch attached to it. Maybe even get a load that switches on to drain 5% and then charge it, or whatever is healthy for the battery (not sure).

[floobydust]

I get short lifetime from NiMH that get a lot of constant charging. It seems to damage them, dry them out possibly and even after one year they show problems.
Ladda (especially), less so with Eneloops which I find are not identical to the Ikea, different V/I curve.
In a pack, imbalance also then shows up badly and then it gets worse, cells are low charge V or way over V and the pack quickly degrades. You have to go in manually and rebalance.
Also I find NiMh need that top-up voltage to over 1.50Vpc so voltage-limiting constant charge does not really work either.

[RoGeorge]

The charger looks very nice, congrats! 

However, the idea of continuously charging didn't work very well for me.  I am under the impression that it damages the batteries faster than normal charging, in the sense that the nominal voltage can become 1.5V or more with time.  I've stopped using that charger of mine, though I did no rigorous study.  It's anecdotal evidence, but I am under the impression that continuous charging without voltage limiting, with time, it's raising the internal resistance of the batteries, and the battery voltage too.

I know the annoyance of never having charged batteries because they self discharge in a couple of weeks.  To solve that, either use NiMH marked with Low Self Discharge, or use Li-Ion batteries instead.  Li-Ion self discharge rate is negligible.  Either prefer Li-Ion appliances (when you buy), or migrate the existing ones to Li-Ion.  I think there are, too, AA/AAA with a Li-Ion inside and a voltage regulator, such that they appear as normal 1.5V batteries, but I never tried them.

I simply use low-discharge NiMH (casual brand, Tronic, available here from any LIDL supermarket for less than $1/pcs or so - I still have usable Tronic AA/AAA from 2014, less capacity but usable and still low-discharge).

[wraper]

I get short lifetime from NiMH that get a lot of constant charging. It seems to damage them, dry them out possibly and even after one year they show problems.
Ladda (especially), less so with Eneloops which I find are not identical to the Ikea, different V/I curve.

Besides low capacity LADDA that isn't eneloop, there are two types, one regular grey eneloop (750/1900mAh), another grey eneloop pro 2450mAh that previously was white, and AAA 900mAh eneloop pro that is no longer available. Also not all Sanyo/Panasonic eneloop are the same, I know of at least 4 generations of them. Curently sold Chinese and Japanese Panasonic eneloop are not the same either.

[tooki]

I can only agree with what everyone has said so far.

There’s a reason they sell those special NiMH cells designed for trickle charging (e.g. Varta “phone” batteries). It’s never been the best way to charge NiMH, but modern high-capacity, LSD cells are even less tolerant of it. No, they’re not just derated high-capacity cells, they’re not just low-capacity cells, they are modified low-capacity cells. (I forget what the difference in construction is, but I read it somewhere a while back.)

But moreover, the reason for trickle-charging in the past, aside from it being extremely cheap to implement, is that NiCd and old NiMH cells were NOT LSD — meaning that they had high self-discharge, meaning they’d lose their charge fairly quickly once removed from the charger. So trickle charging was the only way to ensure they were always charged.

Modern LSD cells are designed to, well, have low self-discharge. The whole point of that is that you can charge them normally, remove them from the charger, and then they’ll retain that charge for a long time, so they are ready when you need to use them. So you’re chasing a solution for a problem your cells don’t have.

Edit: and don’t forget that with your 20mA charge current, your 2450mAh cells will take 4 days to charge from empty, and you have no way of knowing when charging is complete. So the “convenience” of constant charging to have always-full cells means you have no way of knowing which cells are actually full, if any. I certainly find it easier to have a container for full rechargeable batteries: empty cells go into the charger, full ones come out and go into the “full batteries” container for use.

[floobydust]

I get short lifetime from NiMH that get a lot of constant charging. It seems to damage them, dry them out possibly and even after one year they show problems.
Ladda (especially), less so with Eneloops which I find are not identical to the Ikea, different V/I curve.

Besides low capacity LADDA that isn't eneloop, there are two types, one regular grey eneloop (750/1900mAh), another grey eneloop pro 2450mAh that previously was white, and AAA 900mAh eneloop pro that is no longer available. Also not all Sanyo/Panasonic eneloop are the same, I know of at least 4 generations of them. Curently sold Chinese and Japanese Panasonic eneloop are not the same either.

It is confusing and hard to keep up to date about LSD NiMH manufacturers. There are differences, including charging profiles.

Ikea puts COO on the backside of the Ladda packaging and it's also written on the batteries. Of the Japanese ones I have, closely looking at their construction shows they are not identical, different vent size and metal+plastic for Ladda vs. Eneloop white. I have no grey ones at all yet, some black Eneloop Pro.

Also, I was getting low capacity with all and traced it back to the Eneloop charger - it's old with no feedback opto for the SMPS and the output voltage either in the PSU drifted down, or the MCU A/D drifted up, so it undercharges. I did not troubleshoot and repair it. But for a while I was blaming the batteries for short lifetime.

[bdunham7]

So you’re chasing a solution for a problem your cells don’t have.

In the OP's case that's probably true, but what about older devices that are designed with float-charged cells (often NiCad) where I'd like to use better (less toxic, higher-capacity) NiMH?  Are those "floatable" cells available?  Will limiting the voltage with a shunt regulator work?  I have at least three devices with this issue that I'd like to restore.

[tooki]

So you’re chasing a solution for a problem your cells don’t have.

In the OP's case that's probably true, but what about older devices that are designed with float-charged cells (often NiCad) where I'd like to use better (less toxic, higher-capacity) NiMH?  Are those "floatable" cells available?  Will limiting the voltage with a shunt regulator work?  I have at least three devices with this issue that I'd like to restore.

OP and I both mentioned the special “cordless phone” NiMH cells designed for trickle-charge applications. E.G. https://www.varta-ag.com/en/consumer/product-categories/accus/phone-aa-1600-mah

[jirij]

Thanks for the replies.

The intention of the project was exactly to avoid the inconvenience of having to put discharged batteries in a regular charger, wait a day (hoping I'll remember tomorrow), and then retrieve and put them in a long-term drawer storage. That the charger would function as the "drawer storage", skipping the extra step.
It would go into a 3d printed box with a "slider" to mark the latest-placed battery, so I would have a "circular buffer" of batteries, always taking the oldest one I put into the charger.

But I suppose it's not that more inconvenient, and making a new project that tops-up the cells once every week/month is probably too much effort for the little benefit.

[David Hess]

I think if you really want to know if NiMH cells can be safely trickle charged, then you will have to take the time to run a test.  From multiple sources I have only heard that they cannot be safely trickle charged.

[tooki]

Thanks for the replies.

The intention of the project was exactly to avoid the inconvenience of having to put discharged batteries in a regular charger, wait a day (hoping I'll remember tomorrow), and then retrieve and put them in a long-term drawer storage. That the charger would function as the "drawer storage", skipping the extra step.
It would go into a 3d printed box with a "slider" to mark the latest-placed battery, so I would have a "circular buffer" of batteries, always taking the oldest one I put into the charger.

But I suppose it's not that more inconvenient, and making a new project that tops-up the cells once every week/month is probably too much effort for the little benefit.

Bear in mind that you ideally keep batteries in sets that get worn out together. Back when I did a lot more photography than I do now, I used NiMH in my flashes (in part to save $, but mostly because NiMH can deliver more current than alkaline, so flashes recharge faster using them). I still have most of them. To ensure they wear evenly, I keep them in matched sets of 4 (i.e. 4 cells from the same package, so they’re from the same lot) which I keep in 4-cell storage cases like in the image attached, and label each battery with an ID (e.g. A1 through A4 in battery set A, B1…B4 for set B, etc). They always get used together and always get charged together. This is much better than mixing and matching, at least for applications where you run them to empty. In essence, each set gets treated like a battery pack. I have separate 2-cell sets for my computer’s wireless trackpad, my bathroom scale, etc.

[tooki]

I think if you really want to know if NiMH cells can be safely trickle charged, then you will have to take the time to run a test.  From multiple sources I have only heard that they cannot be safely trickle charged.

Well, there’s certainly no risk to human safety. It’s just hard on the cells, reducing their lifespan.

[David Hess]

I think if you really want to know if NiMH cells can be safely trickle charged, then you will have to take the time to run a test.  From multiple sources I have only heard that they cannot be safely trickle charged.

Well, there’s certainly no risk to human safety. It’s just hard on the cells, reducing their lifespan.

For the test I would buy a bunch of cells, condition and measure their capacity with a battery analyzer, and then leave them on a trickle charger at different currents for months.  Periodically I would test and measure the capacity of each sell, maybe every week or month.

[wraper]

It is confusing and hard to keep up to date about LSD NiMH manufacturers. There are differences, including charging profiles.

Ikea puts COO on the backside of the Ladda packaging and it's also written on the batteries. Of the Japanese ones I have, closely looking at their construction shows they are not identical, different vent size and metal+plastic for Ladda vs. Eneloop white. I have no grey ones at all yet, some black Eneloop Pro.

Also, I was getting low capacity with all and traced it back to the Eneloop charger - it's old with no feedback opto for the SMPS and the output voltage either in the PSU drifted down, or the MCU A/D drifted up, so it undercharges. I did not troubleshoot and repair it. But for a while I was blaming the batteries for short lifetime.

I have compared LADDA with Japan made regular eneloop and eneloop industrial. White LADDA were exactly the same, plastic wrap even has the same reflective dotted pattern on top. The only differences are print on the wrap and top insulator color. Grey LADDA uses wrap of different color but otherwise construction is exactly the same when inspected under microscope.

[bdunham7]

For the test I would buy a bunch of cells, condition and measure their capacity with a battery analyzer, and then leave them on a trickle charger at different currents for months.  Periodically I would test and measure the capacity of each sell, maybe every week or month.

I was hoping someone else had done the work already!  There's quite a few devices that used NiCad batteries as voltage regulators of a sort.  I have two of them,  Fluke 731B voltage reference and an HP403B AC voltmeter.  I think I might just try to track down the Varta cells that tooki mentioned for the 731B and just use the zener/capacitor method for the 403B.  Then I can do a multi-year test in the device.  I think I can reduce the charge current and use a removable battery holder instead of soldering up a permanent pack. 

[bdunham7]

OP and I both mentioned the special “cordless phone” NiMH cells designed for trickle-charge applications. E.G. https://www.varta-ag.com/en/consumer/product-categories/accus/phone-aa-1600-mah
[/quote]

Those look good and one of my projects needs AA cells, so now I just have to track them down. 

[floobydust]

It is confusing and hard to keep up to date about LSD NiMH manufacturers. There are differences, including charging profiles.

Ikea puts COO on the backside of the Ladda packaging and it's also written on the batteries. Of the Japanese ones I have, closely looking at their construction shows they are not identical, different vent size and metal+plastic for Ladda vs. Eneloop white. I have no grey ones at all yet, some black Eneloop Pro.

Also, I was getting low capacity with all and traced it back to the Eneloop charger - it's old with no feedback opto for the SMPS and the output voltage either in the PSU drifted down, or the MCU A/D drifted up, so it undercharges. I did not troubleshoot and repair it. But for a while I was blaming the batteries for short lifetime.

I have compared LADDA with Japan made regular eneloop and eneloop industrial. White LADDA were exactly the same, plastic wrap even has the same reflective dotted pattern on top. The only differences are print on the wrap and top insulator color. Grey LADDA uses wrap of different color but otherwise construction is exactly the same when inspected under microscope.

I checked again and the AAA's vent hole I think the plastic washer is higher up so the port looks smaller.
AA's are noticeably different, each end stamped cap. Mass differs as well.

eneloop white 1,900mAh AA 26.1g
eneloop pro black 2,450mAh AA 30.2g
ladda white 2,450mAh AA   30.4g

eneloop white 750mAh AAA 11.8g
ladda white 900mAh AAA 12.4g
no-name china 600mAh green AAA 8.7g (solar garden light)

For OP, there are charge controller IC's for NiMh. Maybe look at them because it's false economy to save on a charger circuit but have shorter lifetime on the expensive pack.
If I'm not mistaken, Linear Tech invented the LTC4056 Li-Ion charger IC and it's been copied to great success. Look at the lineup at ADI, there are also NiMH charge controllers that terminate charging when done. CN3085 (1-3 cell NiMH), LS2533A but I have not tried them.

[NiHaoMike]

I expect that the longest service life would be when the cells are kept in a middle charge level most of the time, that's how the Prius does it.

You can take a look at this project for an open source NiMH charger:
https://www.ultrasmartcharger.com

If you have a whole lot of loose cells to charge, it would make sense to build a single cell charger and multiplex it with MOSFETs.

[wraper]

I checked again and the AAA's vent hole I think the plastic washer is higher up so the port looks smaller.
AA's are noticeably different, each end stamped cap. Mass differs as well.

Grey 1900mAh LADDA has the same shape on the bottom and no visible vent holes on positive terminal as your 1900 eneloop. Also it's obvious that your eneloop and LADDA come from the same place as they use exactly the same wrap with dotted pattern I mentioned before. In any case, there is only one LSD NiMH factory  https://www.fdk.com/battery/nimh_e/ in Japan according to what I read.

[Siwastaja]

The intention of the project was exactly to avoid the inconvenience of having to put discharged batteries in a regular charger, wait a day (hoping I'll remember tomorrow), and then retrieve and put them in a long-term drawer storage. That the charger would function as the "drawer storage", skipping the extra step.

Now that we know what your "X" is, I think it would be a better idea to design a relatively well-engineered modern-day fast charger, with the twist that once done, it goes into very low leakage mode, so that you can be confident that leaving the cells on the charger does not discharge them even in long term; basically combining the pros of your initial proposal with the pros of the usual fast chargers.

Some of the commercial chargers could theoretically be that way already, but maybe it's hard to know for sure, or find those.

[soldar]

This thread is of interest to me. I have always been frustrated because my cordless drills with NiCd batteries were never ready to use when I needed them. To the point that I have fitted connectors so I can power them with external 18V PSU.

In the coming weeks I will be rebuilding a couple NiCd packs with NiMh cells and my plan was to leave them permanently trickle charging. I do not quite understand how can that be a problem. How can it be a problem to supply a trickle current equivalent to the self-discharge? Or even a bit less?

[wraper]

How can it be a problem to supply a trickle current equivalent to the self-discharge? Or even a bit less?

LSD NiMH cells have barely any self discharge, in fact usually less than li-ion cells. Less than self discharge current would be a few microamps.