I am used to charging NiCd or NiMH cells however I haven't done it from a solar panel before. I'm happy I can hold the solar panel near the point of peak efficiency but I am concerned about charge termination due to the variable power available from solar panels.
Normally for NiCd I would use negative delta V for termination with a back up timer. For NiMH I would use a thermistor for delta T termination. However I can see potential problems with solar as the power source.
In principle I could lay out a 60W panel in Arizona and charge the batteries without problems - until someone parks a truck and shades the panel dropping the charge current.
So what is the best way to terminate charge on nickel based batteries with a variable power source?
I do not know about charge termination, but a diode in series with the panel prevents the battery from discharging back into a non-illuminated (or poorly illuminated) panel.
Nickel-based batteries are fairly easy to keep track off if you can put a temperature sensor on the battery, and a reference temperature sensor fairly close by.
Once nickel cadmium bateries are fully charged, any further changing liberates hydrogen gas at the cathode, which diffuses to the anode and recombines with oxygen (from the metal oxide produced by the same current) to produce water and heat - quite a lot more heat than is generated by the same current when it is charging a less than fully charged battery.
"Interchangable" thermistors are quite stable enough to let you reliably detect the consequent extra heating of the battery.
You can determine state-of-charge by hammering the battery (>=3D 1C charge rate) and seeing if it gets hot, but that's pretty hard on the battery if, for example, your circuit wakes anew each morning, or after a cloud. And that gets messed up if insolation dips mid-cycle; then the battery's full, but doesn't heat as expected.
You might terminate (or at least stop blasting) at a temperature- compensated voltage. I think that works, the battery just won't be completely full.
NiCd could then be safely trickled up @ 0.1C, but NiMH hates that.
--
So it looks like, without expressly acknowledging your earlier
boo-boo, you've admitted that your reference was bogus since a PV
lead-acid charger won't work for NiMH.
Unfortunately, and to your detriment, you still haven't addressed the
OP's query, which was, basically, "What do I do if the sun goes behind
a cloud while the battery is charging?"
Instead, you parrot an expensive "Interchangeable thermistor" scheme
to guard against overcharging, which the OP has already laid out, but
which has nothing to do with PV shading during charging.
The way I'd do it would be to completely disconnect the battery from
the charging circuitry when the array was shadowed, and then reconnect
it when the sun came out again.
Got a better idea?
Not many options left. I've never seen a process to determine NiMH charge level without the use of a long and continuous high current. If you look at all the graphs from manufacturers, you'll see that there are no absolute values to use as triggers. They're all deltas and they're only valid for a range of currents.
- Put a LiFePO4 in the solar charger and transfer that to the NiMH.
- Let the battery hit thermal cut-off and derate the life expectancy.
- Use proprietary fast-charge NiHM cells that have a gas pressure sensor.
--
I won't see Google Groups replies because I must filter them as spam
That was the content of the EDN article I pointed to.
formatting link
y_charger_pdf.pdf
Why don't you read it? It didnn't address the specific problem posed by Ravinghorde's customer's choice of battery, but somebody with a little more imagination than Ravinghorde might be able to adapt the ideas presented.
For PV shading during charging, you need something like a ringing choke inverter, that uses a low voltage source to build up current through an inductor, then switches "charged" inductor to discharge into the battery bank at a voltage determined by the battery bank.
Ringing-choke inverters don't exploit core material particularly efficiently, and there are a variety of more complicated schemes around. Linear Technology makes a fuss about its ics that can be used to build buck-boost SEPIC converters.
I've never played with them, so I don't know much about the area.
Ravinghordes delta-T charge termination scheme only uses a single thermistor to detect the jump in temperature when the cells go over to hydrogen generation; this is cheaper than using and monitoring two thermistors, but there are situations where the delta-T can be masked or faked by the environment.
The ringing choke - or one of the elaborations of that basic idea - can be used to step up the solar cell output voltage more when the cells are shaded so that the batteries continue to be charged, albeit more slowly, when there is less sun around.
It's obviously a better idea than yours, but it is scarcely mine - people have been doing it for years.
That's not a bad idea--Ray-O-Vac made some of those. 15-minute charging, I believe, so you load them at about 4.5C.
The problem with temperature cut-outs and fast charging is that once a cell's full, it still takes a while to heat up. So, you wind up overcharging a bit every time, which cuts the cell's service life.
A little OT, but the NiMH comments here were interesting--
formatting link
are-they-so-expensive/
Limited service life. If repeatedly deep-cycled, (i.e., the charge is completely depleted and then recharged, especially at high load currents), battery performance begins to deteriorate after 200 to
300 cycles.
Limited discharge current. Repeated discharges with high load currents reduce the battery's cycle life.
Sensitive to high temperatures. Performance degrades if stored at elevated temperatures. NiMH batteries should be stored in a cool place and at a state-of-charge of about 40 percent.
Maintenance. NiMH batteries require regular full discharge to prevent crystalline formation.
They say they can make an NiMH pack last the life of a car.
I remember reading a paper, which I can't find, years ago. The rough idea was you could use short discharge pulses between charge pulses and the voltage depression during the discharge pulse was related to the state of charge.
Whether the technique actually works in practice is another matter, but if it works then it wouldn't depend on the charge current being constant.
Hi, I have designed a solar battery charger for an autonomous system much like the one you describe. It uses NiCad batteries.
The batteries used aren't much like typical AA Nicd's, they are ranged from 45A=B7h to several hundred A=B7h's. As you have already guessed, in a photovoltaic power system, the charging is made on a "opportunity charging" scheme. You don't have control on how much power you can get at a given time, the panels output power will change drastically on a cloudy day. You can't expect to have anything like constant-current charging, so dV/dt charge termination will give you problems, since you can have negative dV/dt due to the panels decreasing its output current.
AFAIK, there are three types of charge controllers, and all of them rely on final battery voltage for charge termination. - ON/OFF regulators: These have two threshold voltages, a higher one which disconnects power to the battery and a lower threshold which reconnects power. Much like a thermostat does. These are the oldest and less sofisticated ones, but you can see that power disconnection relies on battery voltage compared against these thresholds. - PWM regulators: These have a final charge voltage setting, and do a PWM on the transistors that control power from the panels to the battery. They can gradually control how much power goes to the battery. The way they work is, if the battery voltage is much lower than the final voltage setting, they let 100% of the power to the battery, as the battery voltage rises and it gets closer to the final voltage setting, the regulator lowers the power delivered to the battery. Normally, battery voltage will equal the final battery voltage value. - MPPT (Maximum Power Point Tracking) regulators: These regulators have a DC/DC converter and can adjust the voltage seen by the panel when connected in order to obtain maximum power from the panels. I can't give you much detail about these types of regulators, since I don't know them very well. But I can tell you that they are supposed to give you the full rated power from your panels, which PWM or ON/OFF regulators wont do.
Normally, the batteries used in these systems are Pb based ones, mostly because they are cheaper. But if high temperature behaviour and in some high reliability applications, NiCd's are used. If you take a look at Saft's Sunica + range of batteries, you will see how these batteries are oriented for photovoltaic applications. And if you dig a little bit in the technical manual, you will see that they recommend a final voltage-based charge termination. They even recommend different voltage settings based on daily depth-of-discharge.
So the short answer is; If your application looks anything like this one, use final voltage charge termination.
I am going to use a comparator to give an interrupt to the PIC which will disable the buck convertor.
This should keep the solar panel sitting close to it maximum power point.
I like the pwm termination method of lowering the current as you get close to the maximum battery voltage. It should be possible to get over 90% charge without risking overcharging.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.