NiMh cells - can I 'float' charge?

[electronicsuk]

nimh float charge

Hi all. I have an application where I need to construct a DC power supply with battery backup. My initial intention was to use lead acid cells for the backup, but there is very limited space available inside the enclosure the circuit will be mounted in which rules this out. Instead I'm leaning towards a pack of high capacity NiMh AA cells, they offer the benefit of a much better capacity to physical size ratio.

My question is, is it possible to charge NiMh cells with a constant low current in the same way as you would float charge a lead acid? Again, due to size limitations I need to keep the charging section as simple as possible, preferably just a case of a single current limit that will be suitable for both initially charging the cells from flat and providing a standby charge current to overcome the cells self discharge rate.

I'm worried that after a year or so this form of charging might cause some damage to the battery, and wondered if anyone had any thoughts on the issue? Would it be reasonable to expect a good service life from this sort of arrangement, and what would be a good charge rate to choose? Again, I'd prefer as little active circuitry as possible but if different bulk charge and float currents are the only way then I could design them in.

Thanks in advance

Matthew

 

[iggyboy]

nimh float voltage

The topics of battery charging can be a bit tricky. You can float charge any battery, but beware if you do it at the wrong voltage as this will dramatically shorten battery life and (in case of Li-ion etc. batteries) can cause even an explosion. And that is real cool to see if you google a little bit, btw. It is very important how you charge a battery as many of it's properties are dependant of its history. That is the way it was charged before and the way it was discharged. Was it exposed to temperature variations etc. Lead acid are the most forgiving of all, but of course the penalty is their charge/voluume and the weight. And health. (OT: does anyone think ROHS is a dumb move? Or is it just me?) Beware: same chemistry batteries can differ much from manufacturer to manufacturer so the best move is to design for a selected battery and let the consumer worry if the appliance explodes killing his neighbours' cat.

If I recall correctly the correct way of charging NiCD and NiMH batteries is constant current untill the voltage at the battery reaches a maximum and then detect a drop in the battery voltage (at the same current!) soon enough and change the regulation to voltage mode and let the battery float at a certain low enough voltage. The algorythm is the same for these two chemistries, except the voltage drop indicating full battery is more pronounced for one of them, don't remember which one... It is interesting to note that often a temperature measurement of the battery is very usefull as some actually drop a few degrees C when full. Temperature of course plays a vital role in the performance and charging of batteries, especially for Li-ion ones. (That's why many of them come with several connection terminals - they are the most susceptive...) The net is full of info about this topic, but be careful, much of those are questionable at best. When I did the design, I got most of my informations from more or less known manufacturers (Panasonic atc.) and I was even lucky enough to get datasheets of the 2 most used batteries on the field. Some papers are avaible on IEEE.

Now for the lead-acid batteries which are used in the field and are trickle charged at too high a voltage, those were replaced every two years for reliability issues. Again I say, the lead-acid ones are the more robust. They work many years if done right. Imagine what happens with some other chemistry batteries.

You never stated the way the batteries are used? Cycle use or stand-by use? Cycle are for example cell phone, laptop batteries. Stand-by are emergency lights. One of the safest way to charge a battery is in the stand by mode. Some manufacturers do not put a limit to the charging current of batteries when charged in st-by mode. Consider this, it is the simplest way.

Charging a battery can take between 4 and 24h, sometimes even more and for some chemistries much less. Keep in mind that a battery gets 70~80% of its charge relatively fast and that the last few percents are the lenghty ones.

Also consider the charge efficency. That is how much charge is transfered compared to the charge stored in the battery. Highly dependant of well, everything, really...

One thing that must be acounted for: if you have a 2Ah battery and you discharge it at say 1A? How long will it last? About 2h. What about if you discharge it at 4A? Much less. Consult datasheets!

Happy charging.

 

[electronicsuk]

float charge lithium ion

You never stated the way the batteries are used? Cycle use or stand-by use? Cycle are for example cell phone, laptop batteries. Stand-by are emergency lights. One of the safest way to charge a battery is in the stand by mode. Some manufacturers do not put a limit to the charging current of batteries when charged in st-by mode. Consider this, it is the simplest way.

Thanks for the heaps of information. I should have mentioned that the application is for standby use, the circuit is basically a DC output UPS. While I know its acceptable to float charge a lead acid battery with a constant voltage with no need to terminate the charge, I'm still unsure if there is an equivalent way of charging NiMh cells. I understand that the best way is, as you say, to design a proper peak detect multi-stage charger for this type of chemistry.

The question is, do you *have* to do it this way, or is it possible to trickle charge the battery for standby applications at the sacrifice of having to wait a lot longer for the battery to charge from flat? Would a NiMh even charge at all from flat at the trickle rate?

 

[iggyboy]

float charge nimh

Some info on charging techniques:

https://www.panasonic.com/industrial/battery/oem/chem/niccad/index.html

I say that at a low enough end current NiCD can be trickle charged at constant voltage (low voltage), but this is not the recomanded method. Constant current is much prefered and you do not have to worry about anything if the charging current is set to a low enough value (no peak in voltage detection!). However degradation will occur, question is how bad it will be. NiCD are supposed to be fully discharged from time to time. Worst effect on NiCD capacity is partly discharging them and then charging them back.

Generally speaking a constant voltage with an appropiate resistor might work, but maching this primitive charger to a general battery and than putting it to temperature variations etc. is probably impossible.

I recomand a simple low current source, a wisely low charging current and many many measurement and a lot of consulting with battery datasheets from several manufacturers. Be conservative in your design. If a thermal runaway occurs well, then you are...

Google on "NiCD charge trickle constant voltage/current". You can find lots of accurate information on all the biggest IC manufacturers (maxim, LT, ST, analog devices etc.) sites. And then ALWAYS consult the batteries datasheet.

EDIT: I just noticed you are corious of NiMH batteries and not NiCD... Well NiMH can't be trickle charged they say, but I think that the above still applies.

 

[electronicsuk]

nimh float capacity ups

Iggyboy,

A constant current charger shouldn't be a problem, I could make up something compact enough based around an LM317 and a few other components. I'll have to check the datasheets for the batteries I'm looking at using, but as a general conservative rule, what might you suggest as an appropriate trickle rate for NiMh cells? The panasonic charging guide for NiCd cells suggests 0.05C, but I don't know how far this applies to NiMh.

Thanks

Matthew

 

[iggyboy]

nimh float capacity

Panasonic says 0.033 to 0.05CmA trickle current charge, however they never say what should be the voltage. Maybe it can be done, but they do not want to guarantee anything.

Useful info:

**broken link removed**

Added after 27 minutes:

The more I think, the more I see trickle charging NiMH batteries is a daunting task. I think that any simple circuit just isn't going to do it. Even if you set the current limit to a low enough value and the unit is successfully recharged, what will happen after a month or so at elevated temperatures? :| Sorry, mate, not reliable enough. After all it is a UPS we are talking about.

 

[electronicsuk]

simple nicad float charge circuit

The more I think, the more I see trickle charging NiMH batteries is a daunting task. I think that any simple circuit just isn't going to do it. Even if you set the current limit to a low enough value and the unit is successfully recharged, what will happen after a month or so at elevated temperatures? :| Sorry, mate, not reliable enough. After all it is a UPS we are talking about.

That's always been my major concern as well, but the problem still stands; I only have limited space in which to fit the cells and associated charging/control circuitry, and cost is also an issue. I think it might be necessary to conduct some trials to see what affect this type of charging has on NiMh cells.

The charge voltage isn't really an issue, as to achieve constant current the output voltage of the charger will be constantly changing as the voltage across the batteries terminals varies with the state of charge.

 

[iggyboy]

float charging batteries

Well if you do have an mcu onboard it helps. I also was designing a dc UPS and managed to squeeze everything (except the power elements) into the mcu. It seemed that the previous design held the voltage of the battery to high for an extensive period of time. In the end the floating voltage and current are both a parameter for lead-acid but it is true you can not control them indipendently. Unfortunately it is unclear what should be done, if the cell voltage gets too high for some reason, at the given trickle current, because of some obscure effect I do not really understand.

Trials will be long... I do not envy you.