Correct and safe way to charge Ni-mh batteries.

[NaxFM]

Hi everyone, for a product I'm designing (and that I will hopefully sell) I need an extremely clean voltage source of at least 12V, and the cheapest way is to use two 9V Ni-Mh batteries in series.
The problem is that these batteries need to be rechargeable, and this is my first experience with a commercial product which uses batteries, so I want to be absolutely certain of what I'm doing.
I understand that Ni-Mh are much easier to handle than Li-Ion batteries and not nearly as dangerous, but I can't find any information about how to correctly handle them in series.
I only found articles saying that they just need a constant current source (1/10 or less of the rated capacity to be safe) and the proper cut off voltage to stop the charge.
Is it really all there is to it? It seems way too easy and as I said before I want to be 100% sure.
I already handled different battery types before for my personal projects, but now that I'm designing a product for the market I want to design a safe and proper battery management system.

[Ian.M]

NiMH cells are *NOT* easier to handle than LiPO cells, its just that their various failure modes have a much lower fire risk. 

They are much harder to charge correctly than LiPOs - Read https://www.electronics-notes.com/articles/electronic_components/battery-technology/nimh-nickel-metal-hydride-charging.php and weep!

Also note that there is no such thing as a 9V nominal voltage NiMH battery.  '9V' NiMH batteries can have from six to eight 1.2V cells in series, giving a nominal terminal voltage of 7.2V, 8.4V or 9.6V, therefore if the batteries are user replaceable, your end of discharge operating voltage for two in series needs to be as low as 13V to use the full capacity, but if they have more cells you need to end discharge at a higher voltage.

Allowing discharge past the knee in the discharge curve causes rapid failure due to slight inequalities in cell capacity leading to cell reversal.  Also you need to monitor the battery voltages and charge them independently as although cells within a PP3 size battery should be well matched, you have no guarantee they will be matched between batteries.

[IanB]

Also you need to monitor the battery voltages and charge them independently as although cells within a PP3 size battery should be well matched, you have no guarantee they will be matched between batteries.

I'd say if you have two 9 V batteries of the same type and using a low current (C/10) charging regime then it would be OK to charge them in series. The only requirement is going to be that all cells see the same current and that the charging process stops at an appropriate time.

But also, for any NiMH battery with multiple cells in series, a fast charging regime with smart charge termination is going to be inadvisable, even for a single 9 V battery.

[jpanhalt]

Battery University:
https://batteryuniversity.com/
https://batteryuniversity.com/search/results?q=NiMH+charging
https://batteryuniversity.com/article/bu-408-charging-nickel-metal-hydride

[Gyro]

I agree that low constant current charging is the way to go when you have so many cells in series. NiCd batteries used to be able to withstand C/10 charging more or less indefinitely. NiMH are more sensitive, but I think continuous charging at C/30 or lower is probably ok.

Presumably your device is intended to run on battery power when operating and charging in standby. In this case, you need to work out how the operating current compares to the charging current and the typical on / standby duty cycle. Obviously you don't want the batteries to go flat in operation but maybe you can adjust the charge current so that it still provides sufficient battery life without spending too long trickle charging fully charged batteries.

[Ian.M]

I'd say if you have two 9 V batteries of the same type and using a low current (C/10) charging regime then it would be OK to charge them in series. The only requirement is going to be that all cells see the same current and that the charging process stops at an appropriate time.

But also, for any NiMH battery with multiple cells in series, a fast charging regime with smart charge termination is going to be inadvisable, even for a single 9 V battery.

Yes, but if they are user replaceable, what's the odds that a mismatched pair will end up in there?  Also, unless you discharge first, to get to a known state of charge, or implement Coulomb counting, its hard to terminate C/10 charging with reasonable accuracy, and as small NiMH cells have very little tolerance for overcharging, even at trickle charging rates, if you rely on a simple timed C/10 charge, I would expect relatively short battery life.    Some manufacturers recommend trickle charging rates as low as max. C/40.

Personally, if this is ever going to grow past a niche market product, I'd go for a 4S LiPO pack or 5S if more headroom is needed, as highly integrated 4S and 5S LiPO charger ICs and BMS ICs are normally readily available and currently still somewhat available, and individual low capacity pouch cells or 4S packs are cost competitive against reasonable quality '9V' NiMH PP3 batteries.

[voltsandjolts]

Hi everyone, for a product I'm designing (and that I will hopefully sell) I need an extremely clean voltage source of at least 12V, and the cheapest way is to use two 9V Ni-Mh batteries in series.

Shipping products with batteries installed is a PITA, especially so for international. Save yourself a ton of trouble and ship to customers 'batteries not included'. PP3 is probably easiest form factor here. If your product is low power, perhaps it operates long enough that the customer could charge them in an external charger, or optionally use alkalines.

[jpanhalt]

Way back 20 years ago, I started using eneloop NiMH batteries (made by Sanyo at the time) in everything that did not require high current (e.g., power for electric model airplanes remained NiCd).  They were recharged safely with ordinary hobby chargers, and for individual cells, I used the charger supplied by Sanyo that charged 4 in parallel.

Never had a problem.  The lasted effectively forever.   The could be fully charged and used months later as self-discharge was very slow, or they could be left at some intermediate charge and charged just before flying.  No nursing, no discharge to storage voltage, no puffing, no fires if fully discharged and then recharged.   In other words, they were easier to maintain than either NiCd or lithium batteries.

As voltage needs for hobby applications have increased, the common pack of 4 batteries in series (e.g., 4 NiCd = 4.8V) increased to 5 or 6 batteries.  As this application seems to be for some purpose other than studying battery chargers, I suggest getting two, 5-or 6-cell packs and an associated charger.  Hobby sources are probably cheaper.

EDIT: In your final product for sale, you might want to reconsider those choices.

[NaxFM]

Thank you everyone, I'm glad I asked this question. It was way too easy to be true...
I thought about a 4s lipo pack. With a battery management module it doesn't even cost that much, just about 15 euro and I wouldn't have the hassle to design and verify my own charger circuit.
My only problem was shipping, as I know that there are may problems with shipping lithium batteries.
I also thought about the possibility to use just two normal 9V non rechargeable batteries, but that wouldn't be great for the user experience. As the product would draw about 7mA continuously.
I may also make two versions: one to sell with the lipo battery and one with attachments for two 9V battery to ship outside Europe...

In any case, it's something designed for electronic hobbyists, so a very restricted market. I'll talk about it in this forum when I will have a first prototype. I'm just waiting for my salary to order everything...

[jpanhalt]

7mA in the "off state" is huge, particularly if the hobbyist goes a few months between uses.  Or, is that only while on?  Can you put the device to sleep?

[NaxFM]

7mA in the "off state" is huge, particularly if the hobbyist goes a few months between uses.  Or, is that only while on?  Can you put the device to sleep?

Of course it while it's on. It can be switched off completely.
It's just that it's something you may want to leave on for many hours

[Ian.M]

How important is it that it be very compact? 

2 off 6x AA cell holders with PP3 style snaps,  filled with dollar store alkaline AA cells would  give around 6 weeks of 8 hours/day operation down to 1V/cell, for only a few bucks.   Make the PP3 snap harness pluggable, and there would be plenty of room for the user to fit a locally sourced rechargable LiPo pack.

[tooki]

I only found articles saying that they just need a constant current source (1/10 or less of the rated capacity to be safe) and the proper cut off voltage to stop the charge.
Is it really all there is to it? It seems way too easy and as I said before I want to be 100% sure.

I don’t know where you were looking, because that’s definitely not correct. You could design a very naive charger around that method, but since it lacks any way to actually detect “fullness”, then to prevent overcharging, you’d have to set the cutoff voltage to something well below the typical voltage of a (new) full cell, so that worn cells (with lower capacity) don’t get overcharged (which would wear them out much faster).

The gold standard way to charge them is constant-current charging with “negative delta V/delta T” termination. The first half means that you monitor the voltage (which slowly rises as it charges), and as soon as you notice the voltage start to drop, you stop charging. Simultaneously, you monitor cell temperature, and once the rate of temperature rise increases, you terminate. Whichever comes first. (The two detection methods are actually related: they’re both measuring the same thing, by proxy. Heat is the result of the cell being full, so temperature is a proxy for overcharging. But the increased temperature alters the cell’s internal resistance, which in turn alters the voltage. Thus, the voltage drop is also an indicator of being full.)

The drop in voltage is tiny — a few mV — so fairly sensitive electronics are needed to properly terminate the charge.

In a nutshell, NiMH is easier to handle (in the sense of physical shipping and handling, and the lack of catastrophic damage if the cell is abused), but is actually quite challenging to charge properly.

Unlike NiCd, NiMH doesn’t tolerate trickle charging. The exception are special NiMH types designed specifically for trickle charging, which are sold as “cordless phone” (as in, landline) batteries.