Solar panel charge controller

You don't want to *cook* the battery (which is invariably at the output of the panel) by overcharging it (i.e., charging it past FULL CHARGE or charging it too *quickly*).

The controller can also block the battery from discharging into the "panel", act as a convenient place to tie in a mains powered charger, etc.

It's there for a number of reasons. As Don pointed out, it's to prevent overcharging the battery, and to prevent driving the panel from the battery. Good ones will also perform maximum power-point tracking, finding the optimal current to extract from the panel for maximum power given the current level of illumination, temperature, phase of the moon, etc.

Because the panel voltage varies with current and temperature, any decent charge controller will have a switching topology, to present the battery with the correct voltage and current given the voltage and current available from the panel.

and operate the solar panel at the load that maximizes its output power.

Yes. That makes the panel look like a more *useful* power source (kind of like heat pumps make latent heat more "useful")

This is known as the maximum power point tracker (MPPT)

Thus, a good controller has to do two things:

When i was researching this, previously, I stumbled on an amusing MCU-based design where the processor did the actual switching (as well as monitoring/tracking operating point, etc.)

Sort of like those designs where the MCU controlled the switch supplying its *own* power. (obviously boils down to a bootstrap problem)

Some MPPT charge controllers can also work with wind turbines or micro hydro souces (not all at the same time of course) like the MidNite Solar Classic....

boB K7IQ

As long as the MCU is getting its power off of the feed side, what's the matter with it doing the switching?

A processor acting as a switching controller is pretty darn feasible these days.

"Daku"

Do you think a *capacitor* can even out the variations in light coming from the sun ??

.... Phil

Well, it'll do a good job with the shot noise, but as you point out, that's not the main problem. ;)

Cheers

Phil Hobbs

(Assuming you are responding to my *last* comment/example,) the point was that the processor was running off the *output* of the switcher that it was controlling. I.e., bootstrapping was just enough to get it started controlling that switch (that ultimately provided its running power).

Of course, forgetting to turn the switch off was suicidal!

Here is a novel idea. If someone not skilled in the art asks a simple question, perhaps one could answer the question without insulting said person.

There is a battery connected to this "battery charger".... It will power the microprocessor AND act as a capacitor. A nice BIG capacitor too, at least for the output.

There has to be some kind of capacitor right on the output of the controller though because the battery is a few feet away, electrically.

boB

Daku wrote in news:b0b486d5-66f0-4730-a5e5- snipped-for-privacy@q8g2000pbb.googlegroups.com:

Solar panels have two bad habits.

1. A panel in darkness has quite a lot of leakage current, and that discharges your battery. This is stopped by the charge controller.
2. Once the battery is full, the panel does not stop pushing current trough the battery, thereby sulfating the battery and destroying it slowly. Also stopped by a smart charger.

"miso"

If you were not such a colossal, posturing WANKER

- you might see the point in getting rid of him.

coming

That's wonderful, thanks miso.

Or, (in the word's I always heard from Granma), "If you can't say something nice, then say nothing at all."

George H.

=A0coming

Why 'get rid' of people? I don't get that part.

George H.

I'd be very interested in how much is "quite a lot of leakage current" as I am investigating this. I once measured only about .1mA of current from the battery back into the panel under laboratory conditions.

Thanks, John

"George Herold"

Why 'get rid' of people? I don't get that part.