Charging 3s NiMH (solar power)

[bocat]

Greetings hobbyists and professionals!

Apologies for the long post in advance - if you don't feel like reading the whole thing there is a TL;DR at the bottom.

I am working on a solar powered sensor where I use a 900mAh 3cell NiMH battery charged using solar charger board made for LiPo. Since charge and discharge voltages of 3s NiMH battery are almost identical to lipo, I thought using a lipo solar charger would be ideal.

I chose CN3065 based charger bought from dear friend in China and when testing theory all worked out perfectly. However, when I have hooked it up to a 6v solar panel the output current of the charger varied significantly with brightness of the sun and ranged from 10mA to 200mA on a bright day. Since I read that you should charge a NiMH at a lower current (generally not exceeding 1/10th of rated battery current capacity), I have looked at ways of limiting current of the charge controller and indeed in the datasheet found that max charge current can be limited by soldering appropriate resistor to one of the controller pins.

I did this mod to the controller and it worked out perfectly, however then I faced another problem. The charger used to shut off when the battery voltage would reach 4.2v and turn back on when v would drop to ~4.10v to top it off. After the modification I have noticed that the controller no longer shuts off charging when the battery reaches 4.2v but instead drops current while still constantly pushing around 20mA into the battery at the mentioned voltage. The circuit itself consumes about 6mA, so net current is still exceeding consumption at the mentioned fully charged voltage level.

Is this behavior of the charger dangerous for the battery or as long as voltage does not go beyond specifications I should be fine? Would be great to hear advise from those who had experience with NiMH charging.

Due to nature of the projects, I cannot use any other type of batteries such as LiPo or lead-acid.

TL;DR - is it dangerous to constantly push current into NiMH battery which is already fully charged based on voltage reading (voltage at which current is pushed is the same as fully charged voltage, so battery V does not increase)?

[jremington]

For such a low current project, I would not use the charger, and simply trickle charge the NiMH battery. They have completely different charge characteristics than LiPo cells, and I'm not at all surprised that the LiPo charger doesn't work.

Use a resistor and blocking diode to limit the charge current during full sun to 0.05C (45 mA for 900 mAh cells) and the batteries should last for years.

That has worked very well for me in the past, but do use high quality cells. The cheap ones don't last.

[bocat]

I have thought of this approach but the challenge is due to location of solar sells, I can only get 3-4 hours of the sun in the afternoon, but if it is a stormy week, then pretty much none.

This is why I went the charge controller route to basically use sun power to the max when it is available. and it was working perfectly for several weeks until I've decided to limit max charge current on it. Trickle charge runs a risk of draining battery at some point, plus I don't exactly understand how to control voltage and current with just a resistor and diode. Solar panels are 6v rated - 200mAh at full brightness and I'm afraid it would fry the battery or not charge it enough with variable voltage and current and limited charge time.

[jremington]

plus I don't exactly understand how to control voltage and current with just a resistor and diode.

The diode is required to prevent the solar panel from discharging the battery at night. It will drop about 0.6 to 0.7 V when conducting current in the forward direction.

Ohm's Law works extremely well to pick R, but it is easier to use your multimeter. Wire a resistor, a diode and two fully charged cells or three discharged cells in series with the solar panel and the multimeter, and pick R to limit the current to 45 mA in full sun.

R will be in the general vicinity of 50 to 80 Ohms, depending on the solar panel voltage.

[bocat]

Thanks for details! I do understand the theory of this and actually tried this circuit before. The problem with assumptions for ohms law is that the resistance limiting current is applied at full sun (voltage and power) while the output current will drop significantly when the sun is at 50% or 25% of its brightness, and since I only have a limited time to charge the circuit (in reality not more than 3 hours a day) and given the unknowns of the weather, this would not really work in my application. I do agree this would not have been a problem if I had full day of sunlight to work with and the weather was more or less predictable in the area where I live.

Additionally this circuit above does not guarantee that the battery would not be overcharged if charging at a rate of ~45mA, so the current would have to be dropped even lower -which makes it safer but not efficient on not so sunny days.

[bocat]

Basically my goal is to find a solution that will utilize solar energy to the max, but limit charge current to under 90mA, but still pull full potential from the sun when it is available. I partially achieved this, but one thing i'm not sure is the behavior of charge controller that still pushes ~20mA into the battery while fully charged at 4.2v. Would not trickle charge do the same to a fully charged battery, except you would have even less control on the voltage?

[ocrdu]

Just some general info:

NiMH can be easily be charged at C/2, so no problem there for a 900mAh battery pack at the currents you have. GP call this fast charging.

The trick, however, is knowing when to stop fast charging. Your charger stops at 4.2V, which through some cosmic fluke is usually about the charging end voltage for 3 NiMH cells when charging at 0.1C. At higher charging currents, the end voltage is higher so in that case the cells will not charge fully when charging stops at 4.2V.

That aside: voltage is not a good way of determining end of charge for NiMH, and a Li-ion charger can't charge NiMH the way they should be charged. Still, it seems you got away with it in your un-modded set-up because of the 4.2V cut-off and relatively low charging current.

Unfortunately, the standard ways of determining end-of-charging for NiMH cells (dV/dt and/or dT/dt) don't work well or at all when there is a solar panel on the input; the charging current varies too much.

Standard charging for NiMH is at 0.1C, and at that charging rate, they can take some abuse: "GP NiMH batteries can endure 0.1C continuous charging for about one year."

NiMH doesn't like overcharging, but GP do suggest a trickle charge of 0.05C.

There's more info on charging NiMH here: GP - NiMH technical

And no, I don't work for GP 8-).

[jremington]

It is possible to use a supercap instead of a battery with low power solar projects. No charge control or worries there!

This is pretty thoroughly explored in Nick Gammon's fun solar powered Arduino tutorial.

[bocat]

ocrdu:
Just some general info:

NiMH can be easily be charged at C/2, so no problem there for a 900mAh battery pack at the currents you have. GP call this fast charging.

Thanks for your great input! I guess all the sources I read from had kind of old information regarding NiMH but quick googling on more recent articles and indeed faster charge rates should not be a problem for modern NiHM cells. The only thing is I've already modified the resistor on charge controller and don't feel like messing with those 0608 smd parts fixing it back, lol. So from a "fast charge" controller it has now been converted to a more of a "constant current" charge controller.

ocrdu:
The trick, however, is knowing when to stop fast charging. Your charger stops at 4.2V, which through some cosmic fluke is usually about the charging end voltage for 3 NiMH cells when charging at 0.1C.

This is exactly why I chose to use a 3s battery, because it has the same full charge voltages as single cell LiPo, charge controllers for which are abundant.

ocrdu:
Standard charging for NiMH is at 0.1C, and at that charging rate, they can take some abuse: "GP NiMH batteries can endure 0.1C continuous charging for about one year."

Do I understand this correctly - a GP cell can be "plugged into" charger that continuously supplies constant current for a year and still survive? I've done some research on this and it sounds like NiMH start to dissipate extra current through heat when fully charged with constant current charge method and a little extra current pushed in won't hurt them much unless it sits like this for long periods of time, however nowhere it defines what long period stands for - a minute, an hour, weeks?

ocrdu:
NiMH doesn't like overcharging, but GP do suggest a trickle charge of 0.05C.

Based on what I've done to the controller, sounds to me I've created an efficient trickle charge set-up so generally should be good.

As mentioned earlier, I was concerned about continuous constant current that was forced into battery, but sounds like it should not be a problem.

I'll try to post a photo with Current/Voltage graph with the part I was mostly concerned about.

[bocat]

jremington:
It is possible to use a supercap instead of a battery with low power solar projects. No charge control or worries there!

This is pretty thoroughly explored in Nick Gammon's fun solar powered Arduino tutorial.

Tried that as well on a 1.5 Farad cap. The problem with those is that although they will survive overnight providing charge to atmega and few other components (whole set up is running on both 3.3v and 5v) it gets too complicated managing different voltages as well as if there are 2 days in a row with cloudy skies, my circuit would be dead.

[bocat]

Attached are photos of the charging profile graph.

V is is the blue line with scale on the left and mA is the black line with scale on the right.

The part I was concerned about is in the red circle where voltage has reached 4.2v and stayed there for about an hour, while the battery was still pushed current in by the controller. It actually looks like the current that was pushed in was up to 70mA, not the 20mA I've mentioned earlier (which I've observed while simulating solar output with 5.5 DC from a switching adapter)

Also attaching the whole device if anyone is interested. This is a custom designed circuit board with Atmega328P, BME280 which measures conditions inside an airtight enclosure, Geiger counter circuit and VEML6075 (UV index) I've added for fun.

[ocrdu]

If changing battery chemistry ever becomes an option, you may want to look into LiFePO4 cells (a little less finicky than Li-ion) and an IC like the LT3652.

It will let you harvest solar energy more efficiently. I have seen a few LT3652-based charger PCBAs out there like this one, but I don't know how good they are; I have rolled my own.