power versus charging voltage and voltage drop solar

[terramir]

Ok I live in Los angeles and right now I have two solar panels on the roof they produce an open circuit voltage of around 18 to 18.9 V at my charge controller with the battery disconnected almost 20 volts on the roof. The voltage drop is calculated current times resistance (0.27 Ohm x 4 amp=1.08 volt for example) now the power loss that equals to is about 4.32 W at that moment in time. The 4 amps is a in-line measurement at a good sun time of the day.

But what really confuses me here are two factors

#1 depending on the charge level of my battery the voltage of the entire circuit is drawn down this means that when the battery is charging from near empty let's say it's at 12V charging @4 amps that would equate to 48 W while when it's near full at 14 volts it would equate to 56 W.

#2 if right now the solar panel is providing an open circuit voltage of near 20 volts at let's say 6 amps short circuit what voltage and current could this thing actually provide under load, I mean if I didn't have the storage drag down.

That is the big question since the battery drags down the voltage aren't we losing more power than need be ? Has anyone ever done any load experiments that could really tell us what voltage would actually be the optimal voltage current yield under full sun I'd really like to see that data anyone can point me in the right direction?

terramir

[commanda]

search the board (or google) solar mppt.

Amanda

[Flux]

Yes that is what it is all about.

A panel has to be able to charge a battery that is at 14v when it is hot during summer months.

In cold weather the panel will put out more volts and when charging a low battery it is significantly mismatched.

The maximum power point is usually near 17v but depends on the panel temperature.

By using a converter to match the panel to the battery voltage you can gain power, mostly in the winter when panel voltage is high.

For one panel it is not likely to be worth the bother but for larger installations the gain can be significant.

Flux

[Opera House]

It is getting to be camp time and I have the same problem with wire losses and voltage matching problems.  I'll be putting my two panels in series and building a switching regulator to convert 24V to 12V.  This will reduce line losses and match the panel to the battery.  The design of the regulator will be to load to a constant voltage on the panels.  Having a fixed voltage on the input gets you about 90% of the advantage of a MPPT.  A 12V input to a MPPT has almost no advantage over just connecting it to a battery.  Start thinking of changing your panels to 24V.

[terramir]

Hi thank you for that info, but I see these things are interesting to say the least in design, using an dc to ac back to dc conversion, anyone ever design a MPPT based on PWM (pulse width modulation)? I'm just curiuos about that

terramir

[Opera House]

If you use another switcher with two inputs like TL494, 38xx - 35xx series, one input can monitor the input voltage (starts at minimum voltage) and the second can monitor the battery voltage (over voltage limits current).  This is fine for solar panels because the MPPT voltage stays about the same for different light levels.  For wind generator this can raise the cut in speed and adding some slope with the feedback/gain resistor should prove usefull.  A second PWM dump would be needed at the input to prevent overspeed.  Or just have very over rated batteries that can take a lot of charging.  The normal limit on current by the second input is used for battery monitoring so there is no protection.   So, the circuit must be overdesigned to maximum current possible.

[terramir]

Ehhh one more question how exactly would I determine the maximum power point, My open circuit voltage is like 20 V or more in full sun. So I could you actually determine the maximum power point or is the maximum power point just the voltage that you get after pulling a load that will take all available volts and amps?

terramir

[commanda]

There are 2 approaches for solar panels. First uses a switchmode converter where the input voltage is fixed. I have a circuit somewhere which uses a 78s40 regulator chip, but can't quickly find it online with google.

Second is a true "hunting" mechanism. See my diary entry for an explanation of the modus operandi.

http://www.fieldlines.com/story/2006/3/12/14840/1315

There's another circuit floating around, a link to it was posted here in the last couple of months.

They're also an off-the-shelf item to buy for solar panels. Shortest point from A to B if you're not heavily into DIY electronics. Why re-invent the wheel?

Rule of thumb; if your open circuit volts is 20, your mppt will be about 17, educated guess. In full sun, of course.

Amanda

[Opera House]

In the charts I have seen, the maximum power point for all useful light levels occurs at approximately the same voltage.  In the circuit I have described you would just watch a amp meter on the output and adjust the input voltage for maximum current.  A periodic adjustment would have to be made as the average outside temperature changed.

[terramir]

Ahhh a questions but first a comment, I can't really convert to 24V because my panels deliver mismatched current and it would cause additional power losses, although the cells are identical in both panels, one panel is a 32 cell plus a 4 cell panel this happened when I didn't know as much, also that panel has cells with more severely chipped corners than my other 36 cell panel, they also due so something I don't understand have an about 0.9 Volt differencial and the patched (32 + 4 cell panel) has about 0.5 A less current. These panels however in parallel deliver about 19-20 V (18-19 in my apartment) and 4 amps of charging current with the battery hooked up with a cheap overcharge controller. The short circuit current even though the wire has a total 2-way resistance of about .27 ohm is about 6 amps. I think I could benefit from some sort of circuit a variance of the mppt.

How would you make a circuit draw an ic adjustable determined voltage?

The rest of the circuit is pretty clear to me what I would do. I would use a pic-processor to steer two transistor as switches one rail would use the pwm principal to create a voltage approxamitly 0.5 volts over what ever the battery is the second rail would get all the pulses the other one doesn't get and be run through an adjustable booster in order to reach the same voltage level. The code is doable I'm just a little confused about how to get the circuit to draw exactly what ever voltage is the mpp is.

terramir