Hi,

I have a 1W solar panel (6V, 200mA) and 4 NiMH batteries each with a capacity of 2500mAh. If the maximum charging voltage per cell is 1.78V, is it safe to connect the panel directly to the batteries indefinitely? What are the practical considerations and how can I simulate this interaction? My research suggests the pack will never fully charge and will self regulate current as the voltages approach equilibrium.

Regards,
Mark

 

If you limit the charging current to NiMH cells to about .05C they probably won't be damaged but nickel based batteries (NiCd, NiMH) are hard to charge properly and if they get warm they will be quickly killed.

All of the charging strategies have downsides. Very slow charge causes "memory" problems. Fast charge causes heating. Using voltage differential methods requires a resolution of a few mV to detect.

Charing will hot or cold reduces service life noticeably.

All in all, they are not a good chemistry for solar power outside, possibly, solar garden lights where the current requirements are very low. Nonetheless, they seem to be able to do very well in that application, though they do incorporate charge controllers.

What are you trying to power? What NiMH?

[EDIT: Fixed a sentence for readability.]

 

No, when NiMH batteries are overcharged they start spewing chemicals. You should not leave them charging with more than C/20, in this case 125 mA.

Edit: fix the charging current.

Bob

 

If you limit the charging current to NiMH cells to about .05C they probably won't be damaged but nickel based batteries (NiCd, NiMH) are hard to charge properly and if they get warm they will be quickly killed.

All of the charging strategies have downsides. Very slow charge causes "memory" problems. Fast charge causes heating. Using voltage differential methods requires a resolution of a few mV to detect.

Charing will hot or cold reduces service life noticeably.

All in all, they are not a good chemistry for solar power outside, possibly, solar garden lights where the current requirements are very low. Nonetheless, they seem to be able to do very well in that application, though they do incorporate charge controllers.

What are you trying to power? What NiMH?

[EDIT: Fixed a sentence for readability.]

I have two Arduinos with 2.4GHz modules in the works for a remote control vehicle. The cells are AA. The plan is to have the panel serve as the charging source mounted to the vehicle. I suppose the question really is how to use power from the batteries and charge at the same time when sun is available. I have a 500F supercapacitor I would like to incorporate somehow for the high current awesomeness.

 

I have two Arduinos with 2.4GHz modules in the works for a remote control vehicle. The cells are AA. The plan is to have the panel serve as the charging source mounted to the vehicle. I suppose the question really is how to use power from the batteries and charge at the same time when sun is available. I have a 500F supercapacitor I would like to incorporate somehow for the high current awesomeness.

Well, I would think you'd have more success if you converted it to LiPo. A pair of cells in series will get you enough to put a buck converter after them for the expected 6V. It will be a lot easier to find an MPPT solar battery controller for LiPo and you'll get a bonus of better performance.

 

The solar panel will not provide a fixed 6V under all conditions. High sun and low load will, I suspect, give a much higher voltage and damage the batteries.

 

Energizer Battery Co. says not to trickle charge Ni-MH more than C/40.

 

Well, I would think you'd have more success if you converted it to LiPo. A pair of cells in series will get you enough to put a buck converter after them for the expected 6V. It will be a lot easier to find an MPPT solar battery controller for LiPo and you'll get a bonus of better performance.

This is my first crack at a RC. My Arduinos seem to work alright with 4-5V and I would like to implement my own charge controller to some degree. I have subjected these cells to some abuse so I would like to get the most out of them while empirically identifying their limitations. Resources say NiMH are charged by a constant current for the most part which is how I've been charging them. I theorize the closed circuit voltage of my panel minus any regulator dropout will not be able to deliver a constant current so what is a way around this problem? A DC-DC boost stage feeding a current regulator seems like an option since I already have the Arduino present. If it works out that I cannot charge and use the battery at the same time that's fine too.

 

This is my first crack at a RC. My Arduinos seem to work alright with 4-5V and I would like to implement my own charge controller to some degree. I have subjected these cells to some abuse so I would like to get the most out of them while empirically identifying their limitations. Resources say NiMH are charged by a constant current for the most part which is how I've been charging them. I theorize the closed circuit voltage of my panel minus any regulator dropout will not be able to deliver a constant current so what is a way around this problem? A DC-DC boost stage feeding a current regulator seems like an option since I already have the Arduino present. If it works out that I cannot charge and use the battery at the same time that's fine too.

If you want to pursue using the NiMH you are going to have to work out a constant current strategy that:

1) Can somehow detect SoC and know when to switch from ~C/10 to ~C/.05, or, if you have the power from the PVs, and you know the cells are low, ~C/3.33 for some time until S0C demands dropping it. And this is the rub, determining SoC for NiMH is not possible with battery voltage, so how can you know?

Normally, SoC is determined either at the end when the cell voltage droop is pronounced and so you can tell it is depleted, or during changing when negative delta V indicates a full charge. Neither of those things help you.

You might be able to work something out using a Coulomb counter but of course that would be open loop and easily disrupted. If you don't care about killing the cells, you could use a pseudo-strategy of a thermal sensor which would indicate if the battery was charging or overcharging by their temperature compared to ambient. This might be possible because NiMH cells stay cool to about 70% of charge, then start heating. It would also warn you about the batteries being too hot (or cold) from discharge or environment to charge safely.

 

If you want to pursue using the NiMH you are going to have to work out a constant current strategy that:

1) Can somehow detect SoC and know when to switch from ~C/10 to ~C/.05, or, if you have the power from the PVs, and you know the cells are low, ~C/3.33 for some time until S0C demands dropping it. And this is the rub, determining SoC for NiMH is not possible with battery voltage, so how can you know?

Normally, SoC is determined either at the end when the cell voltage droop is pronounced and so you can tell it is depleted, or during changing when negative delta V indicates a full charge. Neither of those things help you.

You might be able to work something out using a Coulomb counter but of course that would be open loop and easily disrupted. If you don't care about killing the cells, you could use a pseudo-strategy of a thermal sensor which would indicate if the battery was charging or overcharging by their temperature compared to ambient. This might be possible because NiMH cells stay cool to about 70% of charge, then start heating. It would also warn you about the batteries being too hot (or cold) from discharge or environment to charge safely.

Looks like I'm best off using a commercial charger for batteries. I spoke to an engineer who claimed I could connect 4 cells to my rated panel (and a diode) with minimal concerns and so I was exploring the idea. I would like to know what characteristics the cells will acquire if they are connected in this manner perhaps with the voltage clamped with a Zener diode.

 

Looks like I'm best off using a commercial charger for batteries. I spoke to an engineer who claimed I could connect 4 cells to my rated panel (and a diode) with minimal concerns and so I was exploring the idea. I would like to know what characteristics the cells will acquire if they are connected in this manner perhaps with the voltage clamped with a Zener diode.

They will lose capacity over time and may suffer memory if charged very slowly many times. NiMH is a great chemistry in a lot of ways, that’s why hybrid cars use them, but SoC determination is not one of them.

 

I have a few Duracell AA size Ni-MH battery cells.
The ones made in Japan are 2500mAh.
The ones made in China are only 1300mAh. They were purchased at Costco.
Does Duracell know about the knockoffs?

 

I have two Arduinos with 2.4GHz modules in the works for a remote control vehicle. The cells are AA. The plan is to have the panel serve as the charging source mounted to the vehicle. I suppose the question really is how to use power from the batteries and charge at the same time when sun is available. I have a 500F supercapacitor I would like to incorporate somehow for the high current awesomeness.

The sun will charge the batteries regardless as long as the panel has sunlight.
If the toy is working, and if there's sun, the sun will charge the batteries while it's working.
Power will be divided... some to the battery and some to the load.
Unless a circuit is designed to allow the sun to charge only the battery, regardless if the toy is working or not.
If the toy is working, it will draw power only from the battery, and not the solar panel.
The solar panel will only give power exclusivly to the battery and not any to the load, if there is a load at the time.


What is the reading of the output voltage & output current from the tiny solar panel in good sun?
If each cell AA battery is max 1.35v and you have 4 of them, then 1.35 x 4 = 5.4vdc.

NiMh is maximum 1.35v per AA for eneloop ( 4x 1.35 NiMh = 5.4v )

Forget the supercapacitor, what you need is more panels. Honestly.