flexible solar panels?

[obrien creek farm]

Current flow (amps) is a function of potential (volts). When continuity exists between two dissimilar potentials, current will flow from the greater potential to the lesser. The greater the difference, the greater the flow. An mppt controller will manage the charge rate by adjusting the charge voltage. The ocv of the battery bank does not change in the time it takes to swivel the pv array from east to west, so if the charge voltage goes up, you know more power has become available from the array.

[nanoplane]

So Solar panels are effectively constant current devices. This means that when they are producing power, they produce the same amount of current regardless of the load you put on them (limited by their open circuit voltage and internal resistance).

For example my Flex panels have the following characteristics:

Open Circuit voltage: 18.5-19V
Max Current: 6.5A
"Optimum" voltage under load: 15.5V

Optimum means that if you load the panel output such that you're getting 6.5A of current, with the voltage across the panel of 15.5V, Then the Power output of the panel is maximized to : 100W. .. and Power is what you really want to harvest.

Batteries, on the other hand are essentially constant voltage devices. It doesn't matter much how much current you pump into them, their voltage doesn't change much.

So, the optimum charging situation is if we can take the maximum power we're generating in the panel and feed that power into the battery.

If we connected the panel directly up to the battery, the Battery Voltage won't change and the charge current is limited by the panel, so Power going into the Battery would be: (assuming 13V for Battery)voltage:

6.5A * 13 = 84.5W ... We're leaving about 15 watts on the table.

Here's where an MPTT charge controller comes in..
An MPTT charge controller is a power "converter" and it converts one voltage to another with close to 100% power conversion efficiency. It 'Knows' what the optimum load voltage is on the solar panels and with take current from the panel until that voltage is reached.. so in the example above, the current going to the battery will be 7.75A so...

7.75A * 13 = 100W ... no loss.

Note the Battery voltage doesn't really change because it's a constant voltage device...

------
So much for theory, back to reality:

In my configuration, I have my 20 panels split into two arrays of 10 panels each.

Each of these arrays is configure as two groups of 2 panels in parallel and two groups of 3 panels in parallel.
The 2 panel groups are then connected in series and the 3 panel groups are connected in series and then these two sets are connected in parallel again at the charge controller. This gives me the following for each array: (measured today, a sunny day here in the Pacific Northwest:

Open Circuit Array Voltage: 37.5V
Loaded Array voltage: 31 V
Array Current: 15A

Current at the output of the Charge Controller is 37A. Battery Voltage is 13.7V (float mode). most of the current is being used to feed the MiniSplit AC.

So...

Key points:

- MPPT Charge Controllers Rock.. use them.
- The only way to check the efficiency is to look at the full load current coming out of the panels (at the "Optimum" voltage), the OC voltage won't really tell you anything.
- minimize serial connections to make sure all power from every panel is used.. and if you must serialize (to get the voltage up to where the charge controller likes it), connect in parallel first and then serialize groups of the same size.

[john61ct]

Originally Posted by nanoplane

Current at the output of the Charge Controller
…
The only way to check the efficiency is to look at the full load current coming out of the panels (at the "Optimum" voltage), the OC voltage won't really tell you anything.

Yes.

But only when a depleted bank and/or well-matched testing loads are actually pulling that current from the system.

A full bank and only light loads will tell you nothing about what the SCs are capable of delivering in those insolation conditions at that point in time.

And everything between the panels and their SCs should be ignored, treat it as a Black Box never mind internal details unless things go wrong.

[nanoplane]

Yes, you need to have a partially depleted battery or a heavy load that draws enough power that it exceeds the solar generation at the time. it doesn't have to be a fully depleted battery though.

If you have a battery monitor, it's easy to tell. if the battery is charged and the direction of current flow is still out of the battery, then you're using more power than you're generating.

just run something overnight to draw down the battery a bit. (in my case, I just run the MiniSplit to keep the bus warm..) The next day, I can do a variety of tests to determine charging efficiency based on different configurations..

[john61ct]

Yes by depleted bank I mean keeping it between ~30-50% DoD during testing

[CMORGANSKOOL]

My flex panels are screwed down to the roof. No air gap. And it has to be 150 degrees up there. I’ll record the data Thursday when I’m on the bus. My panels perform to spec.

[john61ct]

They output much higher power and last much longer when kept cooler.

[nanoplane]

Solar flat panels typically have a temperature coefficient of 0.5% / degree C above 25 degrees C. So if the temperature of the panels really is 150 degrees F, that would correspond to about 20% loss in efficiency.

Now, it's debatable how much lower the temperature would be if you add an air space. After All, the panels are black

[nanoplane]

I don't think it matters much how depleted the battery is, as long as you have:
- A big enough load (AC is good ) OR
- have the battery be below the charge controller "MPPT bulk" charging point for long enough to do your tests.

In my system, I have a switch that disconnects the solar from the battery (between the charge controllers and the battery). This lets me forcibly discharge the battery for a while to get it into a somewhat depleted state. AC will take about 65A so running it vigorously for an hour or so depletes the battery enough for about an hour of testing.

[john61ct]

Some SCs get damaged doing that.

As a rule better for such a switch to disconnect the panels from the SC.

[nanoplane]

Quite possibly. I use Midnite Solar charge controllers and they don't seem to have a problem with it. In fact, the manual specifies a disconnect on the output and the input side. I have fuses on the input and the switch on the output.

[gabrielah77]

**** yeah!

[Mekanic]

Originally Posted by obrien creek farm

I want to put flexible solar panels on the roof of my bus, but it looks like there about a dozen different brands available, so I am interested in any opinions/experience as to which brand is the best. thanks in advance for any feedback.

I know a guy that put 2 on his camper and THEN found out they are (like all solar panels) much more efficient if they are cooler in temp. In other words they need a air gap on the back side.

[Camalot]

I veto the bending panels. (I'm an electrician)

You have to get rid of heat below... preferably not into the bus frame. When they get too hot they drop efficiency.

If some cells don't get sun like the others they drop efficiency.

Reliability ratings of bendable solar is low as-well.

[obrien creek farm]

OK. After all is done and said, I have decided to put rigid panels on frames that tilt and swivel. We do all our camping in the winter, so this way I can tip them up and point them south, no matter which way the bus is pointed. Thanks to you all.