Impact of mixing panels with different specs.

Tricksailing · June 19

There seems to be a widespread opinion on websites and Youtube that if you have an array of mixed output panels the output current of a series array is limited to the amperage of the lowest panel and the output voltage of a parallel array is limited to the voltage of the lowest panel.
This doesn’t seem correct to me and doesn’t seem to be supported by comments in these forums.
If a series panel array has appropriate bypass diodes, won‘t those diodes allow higher currents to bypass the low output panels whether output is restricted by design or shading? (although I suspect panels with significantly lower output current specs would never contribute).
I was puzzled enough to check a pair of parallel connected mismatched panels from a grid-tied system and an off-grid marine system, open circuit and with a glowing light bulb - which is an admittedly tiny load. The voltage in both cases remained above the Voc of the lower voltage panel.
In parallel panels I’d thought that the highest panel voltage would dominate until its I-V properties dropped the output voltage to the Voc of the next highest voltage panel when that panel will begin to contribute current. I expect there could be some impact of reverse current flow through the second panel but doesn’t a blocking diode prevent that?
I accept that mixing panels negatively impacts maximum power point, power point tracking and efficiency but I’m addressing a much simpler issue of possible widespread misinformation. I’m very open to correction and amplification if I have things wrong.

BB. · June 20

You are correct, mixing diffent spec'ed panels on a MPPT controller means you should be following a specific set of rules... Rules which may mean that some (or even all) of your mixed panels may not be "optimally" connected together into one array.

If you have a PWM (pulse wave modulated) solar controller (basically, just turns on and off very quickly to give you the average "needed" charging current) operates in a way, that, in theory, you can connect all of your different panels in parallel and get a pretty useful array...

The downside here is, if you have a PWM controller and (for example) a Vmp=18 volt panel and a Vmp=36 volt panel, with the PWM controller connected to a 12 volt battery bank, the output power of the panel will be (roughly) the Pmp of the 18 volt panel + 1/2 of the output power of the 36 Vmp panel... Because:

Power = Voltage * Current

For a PWM controller:

Power (18 volt panel) = ~ 15 volts (operational voltage) * Current from panel
Power (36 volt panel) = ~ 15 volts (operational voltage, not Vmp=36 volts of panel) * Current from panel

When you measuring Varray~14 volts, that would seem to be typical of a PWM controller (battery voltage +1-2 VDC controller drop).

Again, without knowing your setup details (PWM or MPPT controller, panel1-x Vmp+Imp specifications, how you have panels wired parallel, series, or series parallel, etc.) I am only making guesses here.

I can go down the rabbit hole of mixing/matching rules for MPPT controllers--But without "knowing" that you have a MPPT controller, and what the specifications of each panel is--It is pretty much impossible to give you any "accurate" setup information.

That is why I asked if you can give us the Voc and Isc of each panel you have (if you don't have the specifications off the rear mfg. label for each panel).

You can estimate the Vmp of each panel by counting the number of solar cells in the panel... Each cell provides around 0.5 Volts Vmp... So an Vmp=18 VDC panel would have 36 cells in series. A Vmp=30 volt panel will have 60 cells in series. And a Vmp=36 volts would have 72 cells in series....

Sometimes it is obvious how many cells a panel has--But not always. You could have 72 cell panel with each group of 36 cells connected together in parallel, for a Vmp=18 volt panel. And even then, sometimes solar panels are not "obvious" in their cell count/wiring/etc.). And we have even seen mfg. labels that were flat out wrong.

If your system is "working" and doing what you need--I am not going to "argue" that your setup is wrong.... Again, I don't have the details to really make a coherent statement about your system.

Among the many reasons it is difficult (and frustrating) to try and put a system together would flea market parts. If you have the knowledge and test equipment to figure out the good from the bad parts, and verify specifications--Sure, it can be done.

But with beginner's knowledge and less than basic tools (such as an AC+DC Current Clamp DMM)--The job becomes just that more difficult to do.

In the "olden days", there where pretty standard panel sizes and configurations for "battery only" operation. Today, with the Grid Tie panel market much bigger than the off grid market, and the proliferation of MPPT (and the Legacy PWM) controllers--There are few "standard systems & products" anymore... And every system becomes a "bespoke" design and installation.

-Bill

Tricksailing · June 20

Bill - thank-you for your response and please forgive my apparent continuing ignorance.

In your comment you say:

"For a PWM controller:

Power (18 volt panel) = ~ 15 volts (operational voltage) * Current from panel
Power (36 volt panel) = ~ 15 volts (operational voltage, not Vmp=36 volts of panel) * Current from panel"

which reaffirms the statement which puzzled me - that the voltage of an array of parallel connected panels is drawn down to the voltage of the lowest voltage panel. Can you explain why this happens? I'm really not trying to argue, just learn how solar panels work.

I've added a diagram of the test I did, along with simulated I-V curves of the two panels and my (apparently incorrect) prediction for the I-V curve of the two panels connected in parallel (green line with a small offset to show the blue line) assuming a blocking diode on the smaller panel would pretty much eliminate any shunt current through the smaller panel. I did the power curve too, just for kicks. Apparently this entire paragraph is wrong and so I'd just like to understand better.

This isn't anything to do with a proposed system - just a request for help to better understand solar panels. Thanks.

Image: https://us.v-cdn.net/6024911/uploads/editor/j1/xgufkbl8ptei.jpg

BB. · June 20

The PWM controller will draw "any solar array" down to Vbatt+1 or 2 VDC (controller voltage drop)...

If you have a MPPT controller, it "depends". The array voltage depends on both the Vmp-each panel, and the MPPT controller itself.

If all the panels have Vmp equal (or at least no more than 5% to 10% difference between "high and low" Vmp panels in parallel), then the Vmp-array is actually controlled by the MPPT controller itself.

The MPPT controller attempts to solar the equation Pmp=Vmp*Imp where Imp is varied to find the peak Pmp (peak maximum array power).

When all panels are "equal" Vmp, then Vmp-array is equal to Vmp-panel (with upwards of 20% Vmp-panel/array drop as the panels get very hot in full sun).

When you have multiple Vmp-panels in parallel, there will probably be 2-3 or more Vmp-array peaks. The low Vmp-panel will be the first peak, a second peak somewhere between low and high Vmp-panels, and a third Vmp-array peak at high voltage Vmp-panel voltage.

Which of those (in this example) 3 peaks is the "actual peak, depends on solving the Pmp=Vmp*Imp equation.

And not all MPPT controllers are designed to "find" the maximum peak--Will it find the "first peak, middle peak, or last peak"--Cannot tell without testing or knowing the software.

In all cases... 3 pair of 100 Watt Vmp=18 volt panels connected into an array will output 300 Watts total (within specs/temperature/etc.).

If you have a 100 Watt 18 volt vmp panel, a 100 Watt 30 Vmp panel, and a 100 Watt 36 volt Vmp panel--There is no "peak Vmp-array" that will equal 300 Watt total... There will always be a panel "mismatch" with less than "optimum 300 Watt array output).

-Bill

Tricksailing · June 20

Thanks again Bill. I think I'm getting there. Your mention of multiple power peaks in a parallel mixed panels seems to confirm my feeling that voltage isn't limited to the lowest voltage panel. My confusion arose because almost all the websites I've found seem to come up with some variant of the following diagram, either in words or pictures. Note the voltage here is 17v, despite parallel 19v panels. This one is https://solarpanelsvenue.com/mixing-solar-panels/

From your comments, it seems like the argument would be moot in most instances of PWM systems since the voltage is most likely drawn down to less than than the max voltage of all parallel connected panels. It seems a bit more complicated for MPPT systems. If the graphs in my previous post are representative, the higher voltage panel in this specific example would define the system MPP voltage, contrary to the apparently common assertion that the voltage is set by the lower voltage panel.

NANOcontrol · June 27

In your first example: A panel is just a string of diodes. Each diode has a forward voltage of about 0.65V. Times the number of cells, this will be the approximate voltage the panel will start conducting at sun or no sun. If you had snow, you could back feed the panel and melt the snow. IT TAKES A LOT OF POWER. No damage unless you exceed panel current. The higher voltage panel could feed into the lower one.

littleharbor2 · June 28

NANOcontrol said:

If you had snow, you could back feed the panel and melt the snow. IT TAKES A LOT OF POWER. No damage unless you exceed panel current. The higher voltage panel could feed into the lower one.

Midnite Solar had originally planned to add a reverse power flow feature in the Classic controller line which could have been activated remotely to melt snow off your array. I understand it was potentially too risky to implement so they never went through with it.

Impact of mixing panels with different specs.

Comments

Categories