Most of the info I find around the web relating to solar charging of NiMh batteries assumes trying to charge them quickly to then be used in another device. Issues discussed include delta V, timed charging, current measurement, C as it relates to the battery's capacity, etc.
What I'm interested in is charging / maintaining a NiMh battery pack in a device with relatively low current consumption, similar to the solar powered garden lights. I've read that NiMh batteries can be continuously charged at 0.004 C, although 0.1 C and .05 C are often used. The non-continuous nature of solar makes a higher C value acceptable. Etc, etc...
My question is mostly about solar panel choice as it relates to available light in the environment the device will operate in and the voltage of the battery pack, 4.8V. (four AA NiMh). The environment is inside a greenhouse covered with two layers of polyethylene plastic. This cuts the light by 20% to 30%.

What is the recommended voltage of the solar panel above the battery voltage? Will a 7 volt panel provide nearly the same charge current as a 9 volt panel rated at the same current output?
I'm currently experimenting with a 9 volt 210 mA operating (230 mA shorted) solar panel and am getting about 125 mA charge current when the sun hits it through the plastic. When it clouds over, charge current drops to ~15 ma to 60 mA depending on how cloudy. Even this low charge current is great for my application because the circuit draw is only about 5 to 10 mA. I went for overkill on the solar panel due to short days in the winter months. I'm planning on incorporating a charge controller to cut off the solar panel when the battery pack hits 5.8 volts (maybe less, 5.7V?)
Input welcome, TIA
Neo

 

A 9V panel is likely to put out slightly more current at 7V then a 7V panel rated for the same current.

The maximum recommended trickle charge for NiMh batteries seems to be somewhere around 1/30 or 1/40 of the battery capacity. Thus for a 2000mAH NiMh the trickle current should be limited to 50mA to 66mA. So for best battery life you may want to add a current limiter, such as one using an LM317.

High voltage cutoff is not a reliable method to regulate the charge of a NiMh battery.

 

http://forum.allaboutcircuits.com/blog.php?b=536

Just so happens I posted the above blog today about solar arrays and how they work - I'd love some feedback.

Trickle charging really isn't my favorite method of charging batteries after they've been discharged. Trickle charges are really meant to maintain the battery state-of-charge. Charging a discharged battery at a slow rate can reduce the life of the battery - as long as you're okay with that, then go for it. It will probably take a long time to charge your batteries though.

With NiMH batteries, you want to sense the drop in battery voltage before turning off charging. The drop in battery voltage corresponds with a increase in battery pressure, indicating that the battery is fully charged. Sensing a peak voltage is not really the best way to terminate charging.

You can buy high current solar array chargers on ebay for $30 or so.

 

I'm hesitant to limit current due to the short days in the northern hemisphere during the winter. But, I see your point. It's a trade off of battery life. I can live with some for sure... annual battery change...

Please elaborate on high voltage cut off not reliable... Is it due to the inevitable deterioration of the batteries to achieve 1.2 volts per cell? I should mention that I'm using a MCU (PIC16F886) and monitoring the battery pack on an ADC. Point being, I expect to monitor the battery pack condition and provide warning prior to failure.
I don't have any real world info on NiMh life expectancy under these conditions. But, annual battery change would be acceptable...

crutschow, thanks for your insight on the 7v vs. 9v it is quite helpful.

Neo

 

Tindel,
I read your blog post and it is spot on. I had not even considered the thermal effects on the solar array (better term) because I figured I'd simply over engineer it to compensate for my ignorance.
I noticed that the solar array acts as a constant current source more or less when the sun hit it (only the current varies). It seems to always be about give and take... in my case, battery life vs. level of charge. As I type this I realize that battery life is less important (in my application) and AAA size may be a better choice replaced annually.
As mentioned above, for my application, the MCU will monitor the batteries and warn of failure before it's a problem.

Neo

 

I'm hesitant to limit current due to the short days in the northern hemisphere during the winter. But, I see your point. It's a trade off of battery life. I can live with some for sure... annual battery change...

Please elaborate on high voltage cut off not reliable... Is it due to the inevitable deterioration of the batteries to achieve 1.2 volts per cell? I should mention that I'm using a MCU (PIC16F886) and monitoring the battery pack on an ADC. Point being, I expect to monitor the battery pack condition and provide warning prior to failure.
I don't have any real world info on NiMh life expectancy under these conditions. But, annual battery change would be acceptable...

Slight overcharging for short period of time such as in winter likely won't have a large effect on battery life. You might adjust the limit for a higher value during winter than summer.

Peak voltage detect doesn't work well since voltage does not give a good indication of the state of charge of NiMh batteries.

 

Use 2 resistors and a LM317. That's the simplest way of doing it

 

Right, this is an old blog post but I would like to add a couple of considerations in case there are folks out there who are planning a similar project; as am I.

I am planning what I call a self-powered "Dark Detector". It could be called a light detector as well but I like the idea of inverting the naming logic. This is for a gardening project to measure the number of sunlight hours vs dark hours. I would like to build a small circuit into a mason jar with a small profile round solar cell in the lid of the jar. All the rest of the components will be protected from moisture inside. I realized that NiMH would be a good choice for maintaining voltage during dark times, I plan to use a 6V, 100 mA round cell. Four AA's ought to do it for the batteries because I would like a few days of overcast backup power and my circuit draws about 20 mA peak. I am also aware that solar PV is not the ideal in battery charging profiles. Add to this the concern about high ambient temperatures inside the jar and things start getting complicated.

Anyway, I can live with the shortened battery life and my plan is to provide a relatively high charging current when sunlight available and use both voltage and current sensing. Hopefully I can learn know when the batteries are topped off. I may experiment my monitoring with the theoretical current bump but combine this the product of V x I along with V / I to calculate the real time power of the charge and the real time self-impedance of the battery. My guess is watching these profiles, I can figure out when to fold back charging current to protect the battery from overcharge.

For some level of heat sinking, the batteries will be positioned low in the mason jar and glued to the bottom. This way I can bury the jar part way in the ground and heat conduction will help a little.

If anyone tries my VxI and V/I sensing idea, let us know on this thread how it turns out for you.

All the best, Doug