Mixing Solar Panels (again)

[laika]

Well this *has* been asked before but I'm still not clear on the answer.

I have various odd sized deck spaces I don't need to walk on often. I reckon I can cram 250W of semi-flexible panels mixing 50s and 75s of various sizes. I'm aware of the conventional wisdom: ensure current is identical when chaining panels in series or voltage is identical when wiring in parallel. And the other stuff about mitigating different currents in series chains using bypass diodes and problems with MPPT controllers when using different panels which I might nod at and then quietly ask someone for a bit of clarification in a comedy stage whisper away from camera...

I've been looking at specs for panels from various manufacturers but it seems that Vmp is different for different sizes of panel even from the same range So I am now unclear how you're supposed to do this in practice. I'm guessing that plenty of people have installations which they'll say "this works fine" but are actually suboptimal. I do want to do this right. Thoughts?

And how come people don't make charge controllers with several inputs where electronic wizardry could solve the problem optimally?

 

[thinwater]

Before the yelling begins, also remember that with shading and differences in sun angles, the panels are never going to be "identical" anyway. On a home roof top, yes, but not on a boat.

 

[vyv_cox]

I have never had two identical panels. In my collection I have an 85 watt rigid, a 40 watt rigid, a 38 watt semi-flexible, a 30 watt flexible and three 10 watt, one semi-flexible, the others rigid. At various times many of them have been used in parallel through several different PWM controllers. All combinations seem to have worked reasonably well.

The point about shading is well made. My current two main panels, the 85 and the 40, are mounted each side of my radar scanner and antennae. At various times of the day one will inevitably have shadow across it whereas the other will not, and in port there is almost always a mast shadow across at least one of them.

 

[VicS]

I have never had two identical panels. In my collection I have an 85 watt rigid, a 40 watt rigid, a 38 watt semi-flexible, a 30 watt flexible and three 10 watt, one semi-flexible, the others rigid. At various times many of them have been used in parallel through several different PWM controllers. All combinations seem to have worked reasonably well.

The OP is asking about the effect variations in parameters might have on their performance when used with an MPPT controller. He appears to have discovered than panels of different sizes even when of the same type and from the same manufacturer can have different values for V_PMAX.

I suspect the differences will be small and in the overall scheme of things insignificant.

 

[laika]

The OP is asking about the effect variations in parameters might have on their performance when used with an MPPT controller.

Not necessarily with an MPPT controller, although that was a follow-on question and one that my subsequent reading has given me a better handle on.

The basic point was that stuff I've read seems to say that if you have panels of different size in series you limit current through each to the current that the smallest panel is able to handle and in parallel you'll be losing power if voltages don't match. Moreover maximum power point will be different for each panel, so an MPPT controller won't be as efficient as it would be with matched panels. However I've had a hard time matching voltages OR current with different panels even from the same manufacturer.

I *think* what is being said is that like navigation and epoxy work, boat PV engineering isn't the exact science the literature might suggest and that inefficiency due to couple of volts difference in Vmp will be lost in the non-idealness of the system as a whole. whilst having a panel rated at 17v in parallel with a couple at 24v might be an issue, a 17 an 18 and a 19 would not be so much of one. Is that correct, is parallel the way to go here (current differences seem much larger than voltage for different sized panels) and would I be right in thinking that in a heterogeneous parallel set-up I'd be just as well off with a decent PWM controller?

And why don't they sell multi-input controllers?

 

[Scomber]

Not necessarily with an MPPT controller, although that was a follow-on question and one that my subsequent reading has given me a better handle on.

The basic point was that stuff I've read seems to say that if you have panels of different size in series you limit current through each to the current that the smallest panel is able to handle and in parallel you'll be losing power if voltages don't match. Moreover maximum power point will be different for each panel, so an MPPT controller won't be as efficient as it would be with matched panels. However I've had a hard time matching voltages OR current with different panels even from the same manufacturer.

I *think* what is being said is that like navigation and epoxy work, boat PV engineering isn't the exact science the literature might suggest and that inefficiency due to couple of volts difference in Vmp will be lost in the non-idealness of the system as a whole. whilst having a panel rated at 17v in parallel with a couple at 24v might be an issue, a 17 an 18 and a 19 would not be so much of one. Is that correct, is parallel the way to go here (current differences seem much larger than voltage for different sized panels) and would I be right in thinking that in a heterogeneous parallel set-up I'd be just as well off with a decent PWM controller?

And why don't they sell multi-input controllers?

I'm also interested in this topic so can anyone tell me if they have diodes before each panel is connected in parallel with others to prevent any leakages/losses(etc). Thinking of an extra roving panel kept below until at anchor/mooring. Also a simple cheap way to connect and disconnect the extra panel? Don't fancy the MC4 connectors for regular use- seems the lugs can break and might be fiddly? Just bought a PWM dual battery controller and already using a 40W panel. Thanks for any help.

 

[Mistroma]

I have 1x65W and later fitted 2x40W which matched the first panel reasonably well, but not exactly. My 65W gave the predicted output and the 40W panels increased output pro-rata. I have daily output records since 2012. I think that the Vpmax specs. only vary by about 0.5V and that seems to be fine. I have panels in parallel with a PWM regulator.

I agree that small differences can be ignored and that it's better to add a panel that's not quite matched than leaving a space on deck. Laika seems to have a good handle on things and I don't see anything that looks wrong.

As already said, even matched panels will rarely be producing the same output because of angle, shading etc.

 

[Pete7]

I have a 60w and 45w panel going through one PWM and another 80w panel going through a separate PWM regulator. Maximum output is about right for the three panels. The reason for using separate PWMs is that it was easier to wire up separate ones rather than long wires running all over the boat.

 

[Mr Cassandra]

Polycrystalline or Monocrystalline Which would you chose I am looking for a 150 W panel, hoping to get 50 amps a day in the Aegean there`s plenty of space to place panel on the boat

 

[ianj99]

I have a 60w and 45w panel going through one PWM and another 80w panel going through a separate PWM regulator. Maximum output is about right for the three panels. The reason for using separate PWMs is that it was easier to wire up separate ones rather than long wires running all over the boat.

I think having a separate low cost pwm controller for each panel makes a lot of sense.
No worries about mixing and matching, built in redundancy for failure of a panel or controller.

Mppt controllers maybe better but if there is not a need to maximise solar power generation, multiple pwm versions would probably be adequate. Also, one of the smaller panels could be dedicated to the engine battery.

 

[noelex]

With parrallel connection you can connect panels of very different wattage without any problems.

With a non MPPT controller the panels will need to be nominally 12v panels. (Vmp will be around 18v).

With a MPPT controller the rule of thumb is that the Vmp needs to be within 0.5v for the performance to be reasonably effecient (ideally adjust the Vmp for each panel a more realistic temperature than the 25C specified for STC). If the discrepancy is greater than this there will be no damage, but the overall output will be less.

 

[Scomber]

Thanks everyone-all very interesting-now we just need the sun!

 

[Mistroma]

Thanks everyone-all very interesting-now we just need the sun!

Not a problem, it will be switching on again after 22nd April when I hope to be back to my boat. Forecast on accuweather starts with decreasing clouds then partly cloudy followed by a load of variations on sunshine, plenty of sunshine, abundant sunshine and so on. A bit like the Fast Show (Scorchio). Time to dig out the sunscreen.

 

[lpdsn]

I *think* what is being said is that like navigation and epoxy work, boat PV engineering isn't the exact science the literature might suggest and that inefficiency due to couple of volts difference in Vmp will be lost in the non-idealness of the system as a whole. whilst having a panel rated at 17v in parallel with a couple at 24v might be an issue, a 17 an 18 and a 19 would not be so much of one. Is that correct, is parallel the way to go here (current differences seem much larger than voltage for different sized panels) and would I be right in thinking that in a heterogeneous parallel set-up I'd be just as well off with a decent PWM controller?

With the proviso that I've not done much research on it, I can't see how even a small voltage difference between two panels (even if identical panels but subject to different levels of shading or a different angle of incidence) would not mean that the panel with lower voltage would provide no current.

At a common sense level, if one panel is providing 19V and the second is providing 18.5V, at that particular point in time, and they are connected in parallel, how would the 18.5V panel ever drive any current through the blocking diode?

So I just assume you'd need a controller per panel. If having an MPPT controller for each is more than you're willing to spend, then I guess PWM controllers are better than trying to use one MPPT controller for all the panels.

Unless anyone can point me at something which gives a good case for a single MPPT controller being able to draw current from all the panels regardless of small variations in voltage.

 

[noelex]

At a common sense level, if one panel is providing 19V and the second is providing 18.5V, at that particular point in time, and they are connected in parallel, how would the 18.5V panel ever drive any current through the blocking diode?

So I just assume you'd need a controller per panel. If having an MPPT controller for each is more than you're willing to spend, then I guess PWM controllers are better than trying to use one MPPT controller for all the panels.

If the panels are connected in parallel a single MPPT controller this will select the ideal voltage where the combined output of the two panels will be highest.

In the case you quote this would be 18.5v so both panels would produce plenty of power.

Two MPPT controllers (one per panel) could potentially extract slightly more power by operating one panel at 19v while the other is operated at 18.5v. However, you need to take into account the self consumption of the controller itself. So the real gain in this case would be likely to be very minimal and there may even a net loss when a second controller is added.

 

[William_H]

Each different solar panel has an effective internal resistance. This is somewhat related to the current taken out and the voltage of the panel at that current. So a typical 20 w panel will deliver near max current 1 amp while maintaining a panel voltage of around 17v. This is called the max power point and it does vary with sun shine and shadow even temperature.
Now it is this internal resistance that makes them sort of self balancing. So 2 different panels either different by design (max voltage ie no of cells) or by shadow wired in parallel will mean one produces more power under load so dropping the voltage by giving more current or by reducing the current of the weak one so raising the voltage the voltage is made the same. (about 80% of V max. So if feeding into an MPPT controller it will try to produce a load that averages tot he best power for the 2 unmatched panels.
Now MPPT controller has the big advantage that it can use more efficiently the excess voltage of a panel which is provided by more cells in series. The excess voltage is needed to provide enough voltage to charge a 12v battery ie 14v at limited light or shadow levels. But mostly this excess is wasted in the internal resistance if the panel is connected directly to a battery. Like a PWM controller. Which is a direct connection panel to battery until battery voltage shows full charge.
So the point is with panels wired in parallel even if somewhat different design and with different light levels efficiency will not be lost to any great degree by their difference. Only by the loss of efficiency in the shaded panel or the lower voltage panel.
Now with panels in series there is a similar averaging out. Yes the current will be equal to the lowest current panel by shade or design. However the higher current effectively forced through the weaker panel (up to max I) will reduce its voltage(provided it is in that range between Max I or max V and about half voltage at max current.
So the answer is no not a great concern of unmatched panels within reason and likewise with shading.
Using a PWM for each panel is OK. However when the battery voltage rises to fully charged the PWM cuts back current. Now on PWM is inevitably going to cut back before the other so you will get un equal loads on panels but that does not matter.
Why not a MPPT controller with multiple inputs. This would mean a tracking circuit and converter circuit for each panel.
In effect a complete controller for each panel. Might as well buy 2 controllers. Subject to the confusion of voltages as above and multiplied with stepped charging regimes.
Does that make sense? I hope so. olewill

 

[lpdsn]

If the panels are connected in parallel a single MPPT controller this will select the ideal voltage where the combined output of the two panels will be highest.

In the case you quote this would be 18.5v so both panels would produce plenty of power.

The obvious question is how?

The only weapon an MPPT controller has in its armoury, as I understand it, is the ability to force down the voltage at the input side of the controller. A pretty crude weapon and, whilst it can obviously be used reasonably effectively with a single panel there doesn't seem to be any way that it could be used to run two panels producing different outputs at the maximum efficiency of both of them.

I still suspect that the two panels controlled individually would output more power than the pair being controlled as a single entity, even by a better MPPT controller. (As I think someone has already said, it would be different on a roof where identical panels aligned and shaded identically can probably quite reasonably be treated as a single entity).

It would be interesting to see if anyone has tried experimenting with the two alternatives and has real data as a result.

 

[lpdsn]

Why not a MPPT controller with multiple inputs. This would mean a tracking circuit and converter circuit for each panel.
In effect a complete controller for each panel. Might as well buy 2 controllers. Subject to the confusion of voltages as above and multiplied with stepped charging regimes.
Does that make sense? I hope so. olewill

I guess there's no point in making a controller with two or more inputs. Better to mass produce them with a single input and keep the price down.

A quick glance at Amazon's website (prices in Sterling) shows that MPPT controllers cost between £30 and £80. There are cheaper, but I'd have to be convinced they are genuine. One of the £80 ones claims that it is 15% to 30% more efficient (than a PWM controller presumably). I doubt it is more than 1 or 2% more efficient than the £30 controllers, if any better.

PWM controllers seem to be available from £6.

I'm going to make a finger in the air assumption that even with the best MPPT controller you'll not get more than 150% combined out of two parallel panels that are affected by different shading/angle of incidence (one panel at max eff = 100%). I suspect it is a lot less.

Let's assume someone has spent £100 on a second panel. It does seem worth spending the extra on an additional controller, whether PWM or lower price range MPPT, to get the extra benefit out of it.

Interesting.

 

[noelex]

The obvious question is how?

The only weapon an MPPT controller has in its armoury, as I understand it, is the ability to force down the voltage at the input side of the controller. A pretty crude weapon and, whilst it can obviously be used reasonably effectively with a single panel there doesn't seem to be any way that it could be used to run two panels producing different outputs at the maximum efficiency of both of them.

As you correctly point out if the Vmp is different a single MPPT controller cannot run both panels at maximum efficiency. A single MPPT controller can only select the most efficient Vmp for the combined panels. An average if you like (although it likely to be much closer to lowest Vmp)

In the example you quoted where one panel has a Vmp of 19v and the other panel had a Vmp of 18.5v the combined Vmp would be close to 18.5v and this is the voltage the controller would select. It is easy to estimate the loss of this simple model. It would be 0.5/19= 2.6% for the 19v panel. The 18.5v panel will have zero loss. If both panels had an equal output the total loss will be around 1.3%.

This is somewhat simplified. In practice, the actual loss will be even less than this model suggests for a 0.5v difference.

MPPT circuitry is quite complex and controllers with good tracking have reasonable self consumption so we also have to factor this into account. In the case you quote the self consumption of the extra MPPT controller may well be more than the small gain produced. A single MPPT controller could still be the most efficient option.

This does not mean that multiple MPPT controllers are a bad idea. In some cases the Vmp can differ by much more than 0.5v, but if we want to fit one controller per panel it requires MPPT controllers with excellent tracking, but low self consumption. This is a difficult combination.

Even with this sort of controller the overall average gain will only be only very small. Given the limited real estate for solar panels such small gains are worth considering, but there are very few suitable controllers (Genasun are about the only option that I am aware of if you want to go this way) and they have some limitations.

Good MPPT controllers are also very expensive and in general the cheap versions are not worth bothering with.

 

[DannyB]

I may be way off base here, but if you use two or more controllers, wouldn't the output from one controller affect the output from the others. I guess what I'm thinking is that the regulator would sense the voltage from the other regulator rather than the voltage of the battery. Or am I overthinking this?