Panels with more bypass diodes

[s/v Jedi]

Originally Posted by noelex 77

It is accurate to note that multiple controllers, or parallel connection, needs a larger (or multiple ) wire sizes than series connection. This makes installation more difficult and more costly. This same wire gauge could also be used in in a series connection, but the gain would be small and the results would poorer overall than using multiple solar controllers.

The bottom line is that multiple solar controllers results in the highest solar yield, but is also the most difficult and costly installation. Series connection is easier and less expensive but the solar yield will be less.

There is no perfect answer.

Why do you leave out that dedicated controllers result in the highest cable losses? because that's simply what it is.

[rslifkin]

Originally Posted by s/v Jedi

Why do you leave out that dedicated controllers result in the highest cable losses? because that's simply what it is.

What would cause the higher cable losses if you use the same cable sizing criteria for each setup?

[noelex 77]

Originally Posted by s/v Jedi

Why do you leave out that dedicated controllers result in the highest cable losses? because that's simply what it is.

I see your point, but I don’t think this is correct way of looking at the problem. Multiple controllers require multiple wires of a similar size to that required for series connection. This increases the cost and complexity of the installation. This is a drawback that needs to be considered, but if this work is done it will result in a higher solar yield than is possible with series connection.

With multiple controllers you need multiple wires to have the same cable losses as a series connection.

The benefits of multiple controllers is that the Vmp of each panel is perfectly optimised for each panel rather than accepting a compromise voltage. This is far more significant than any minor cable losses, although I would still prefer the cable size ( or the number of wires) to be designed optimally for the system.

The bottom line is that multiple solar controllers will result in highest solar yield, but at the expense of a more complex installation. A side benefit is the redundancy of the multiple controllers.

[s/v Jedi]

Originally Posted by rslifkin

What would cause the higher cable losses if you use the same cable sizing criteria for each setup?

Originally Posted by noelex 77

I see your point, but I don’t think this is correct way of looking at the problem. Multiple controllers require multiple wires of a similar size to that required for series connection. This increases the cost and complexity of the installation. This is a drawback that needs to be considered, but if this work is done it will result in a higher solar yield than is possible with series connection.

With multiple controllers you need multiple wires to have the same cable losses as a series connection.

The benefits of multiple controllers is that the Vmp of each panel is perfectly optimised for each panel rather than accepting a compromise voltage. This is far more significant than any minor cable losses, although I would still prefer the cable size ( or the number of wires) to be designed optimally for the system.

The bottom line is that multiple solar controllers will result in highest solar yield, but at the expense of a more complex installation. A side benefit is the redundancy of the multiple controllers.

Here’s why you have more losses: assume we have three panels of each 36V 10A. Those panels come with short 12AWG cables that need extension cables. Those are only available in a maximum of 10AWG. So we install the MPPT controllers as close to the batteries as possible to reduce cable losses as much as possible. Now you have three times the cable cost as well as three times the losses of a 10A circuit compared to a series string even though both deliver the same amount of power.

What I will admit is that the losses are relatively small when we use such large panels.

By now it is clear that none of you have data to back up your claims and I need to recollect it with the new array. I bought some extra gear to do meaningful tests incl. extra MPPT controllers (I now have three) so good data and comparisons will be available soon

Edit: I did the math to get a feel for these losses: for the example above, there is 15W cable loss for dedicated controllers and only 5W cable loss for the series string. So let’s say these are 300W panels then you have 15W vs 5W loss for 900W installed. 1.66% vs 0.55%

Small, but significant.

[rslifkin]

If the cables are sized for, say, a 3% voltage drop at full output, I don't see the difference. A panel that puts out 40 volts at 10 amps would lose 1.2 volts and you'd get 38.8 volts, 10 amps. So for 3 of those panels you'd be looking at 12 watts lost per panel times 3 panels, so 36w of loss in the cables for a total output of 1164 watts (out of a theoretical 1200).

Put the panels in series, now you have 120V, 10A. 3% loss there brings the voltage down to 116.4 volts. 116.4 volts at 10 amps is the same 1164 watts as the previous example.

There will be 3 of the same size cables with the individual controllers, but the actual loss is the same.

[s/v Jedi]

Originally Posted by rslifkin

If the cables are sized for, say, a 3% voltage drop at full output, I don't see the difference. A panel that puts out 40 volts at 10 amps would lose 1.2 volts and you'd get 38.8 volts, 10 amps. So for 3 of those panels you'd be looking at 12 watts lost per panel times 3 panels, so 36w of loss in the cables for a total output of 1164 watts (out of a theoretical 1200).

Put the panels in series, now you have 120V, 10A. 3% loss there brings the voltage down to 116.4 volts. 116.4 volts at 10 amps is the same 1164 watts as the previous example.

There will be 3 of the same size cables with the individual controllers, but the actual loss is the same.

Even if you would use 6AWG extension cables, the loss would still be 20% higher and those cables don’t exist. With three cables of 4AWG you would beat the single 10AWG cable of a series setup. This is what I call it getting crazy and nobody actually doing that.

And remember, almost all users buy the 10AWG extension cables so they simply take the hit.

[rslifkin]

Originally Posted by s/v Jedi

Even if you would use 6AWG extension cables, the loss would still be 20% higher and those cables don’t exist. With three cables of 4AWG you would beat the single 10AWG cable of a series setup. This is what I call it getting crazy and nobody actually doing that.

And remember, almost all users buy the 10AWG extension cables so they simply take the hit.

Can you show the math that says the loss would be higher? 3x 10 AWG for 3 panels should have the same loss as 1x 10 AWG for the panels in series. It's only if you parallel panels that you start needing to go huge on the cables to keep losses down.

[s/v Jedi]

Originally Posted by rslifkin

Can you show the math that says the loss would be higher? 3x 10 AWG for 3 panels should have the same loss as 1x 10 AWG for the panels in series. It's only if you parallel panels that you start needing to go huge on the cables to keep losses down.

10AWG has a resistance of 1 Ohm per 1,000 feet so for the example with three panels you need to reduce that to 0.33 Ohm. 6AWG is 0.4 Ohm so 1.2 in total, 20% more than the 1%. 4AWG is 0.24 Ohm so 0.72 Ohm total, beating the series setup with a single 10AWG cable run.

For amount of power, current is squared then multiplied with resistance. So that 10A becomes 100 as multiplication factor for each run. The 7 meter distance is doubled to 14 meters for the circuit as we have positive and negative conductors and when we round it to 50’ for easy math then 10AWG comes to 0.05 Ohms times 100 is 5W cable loss.

So no, three 10AWG cables do not equal one 10AWG for series. The reason is that the current is the same for each run.

[rslifkin]

Originally Posted by s/v Jedi

10AWG has a resistance of 1 Ohm per 1,000 feet so for the example with three panels you need to reduce that to 0.33 Ohm. 6AWG is 0.4 Ohm so 1.2 in total, 20% more than the 1%. 4AWG is 0.24 Ohm so 0.72 Ohm total, beating the series setup with a single 10AWG cable run.

For amount of power, current is squared then multiplied with resistance. So that 10A becomes 100 as multiplication factor for each run. The 7 meter distance is doubled to 14 meters for the circuit as we have positive and negative conductors and when we round it to 50’ for easy math then 10AWG comes to 0.05 Ohms times 100 is 5W cable loss.

So no, three 10AWG cables do not equal one 10AWG for series. The reason is that the current is the same for each run.

Is your calculation accounting for a higher voltage on the series setup when converting to watts?

[noelex 77]

I think we all agree that using series connection has the advantage of saving on wiring costs. Using a single controller per panel you need multiple wires and they need to be slightly thicker if you want to keep the wire losses the same. Therefore, the cost is higher and the installation is more difficult.

You should always calculate the losses in each system and choose the wire size accordingly, but even if we ignore this advice and use the same wire size for series connection and multiple controllers, the practical difference is tiny in most systems. A quick calculation on my 1000w system shows a difference of around 0.14%. in the worst case, at full power.

The gains of multiple controllers are much greater when panels in the array are subject to different conditions. With multiple controllers each panel is individually controlled to optimise the output. The output is not affected by other panels in array. Unfortunately, in series connection the current has to be identical for each solar panel in the array and this requirement can have a significant detrimental effect in some circumstances.

[s/v Jedi]

Originally Posted by rslifkin

Is your calculation accounting for a higher voltage on the series setup when converting to watts?

That is exactly the whole trick of why series is superior: you get extra power from the hugher voltage instead of a higher amperage… and losses are only caused by amperage, not by voltage.

The formula is: Power losses = Amperage squared x resistance.

No voltage in the equation

[s/v Jedi]

Originally Posted by noelex 77

I think we all agree that using series connection has the advantage of saving on wiring costs. Using a single controller per panel you need multiple wires and they need to be slightly thicker if you want to keep the wire losses the same. Therefore, the cost is higher and the installation is more difficult.

You should always calculate the losses in each system and choose the wire size accordingly, but even if we ignore this advice and use the same wire size for series connection and multiple controllers, the practical difference is tiny in most systems. A quick calculation on my 1000w system shows a difference of around 0.14%. in the worst case, at full power.

The gains of multiple controllers are much greater when panels in the array are subject to different conditions. With multiple controllers each panel is individually controlled to optimise the output. The output is not affected by other panels in array. Unfortunately, in series connection the current has to be identical for each solar panel in the array and this requirement can have a significant detrimental effect in some circumstances.

Again, 4AWG is a huge lot more than 10AWG and also it is not available as solar extension cord.

[rslifkin]

Originally Posted by s/v Jedi

That is exactly the whole trick of why series is superior: you get extra power from the hugher voltage instead of a higher amperage… and losses are only caused by amperage, not by voltage.

The formula is: Power losses = Amperage squared x resistance.

No voltage in the equation

As soon as you converted to watts to figure out the true power loss then voltage is a factor.

[sailingharry]

Originally Posted by rslifkin

As soon as you converted to watts to figure out the true power loss then voltage is a factor.

No, system voltage does not factor in to calculating power loss. Power loss in Watts (what you call true power loss) is the current in the wire squared times the resistance of the wire. System voltage is not part of that calculation.
If you have three 10 amp panels of any voltage in series, the power loss is 10 times the resistance of the wire.
If you have three 10 amp panels with separate controllers, the resistance on each panel is 10 * the resistance of The wire -- in that resistance is the same resistance as the series wire. But because you have three wires, it is three times the loss.
In the real world, as opposed to the theoretical world, the loss is so small that other factors begin to have more control. But that shifting set of compromises does not change the fact that for a given wire size, the power loss with three separate controllers is three times the loss of a series system.

[FlyingScot]

Here is what the series people claim to be benefits:
Higher voltage kicks of charging sooner and lasts longer. In reality this is minutes per day. The higher voltage means smaller wires for the same voltage drop. Some claim that you could lose charging from clouds but I don't recall seeing the panels ever drop to zero even during heavy rain and clouds, so this is a non issue for our installtion and previous installation. One or few controllers means less losses from self consumption. Makes sense but I've not seen real numbers, so probably true, but maybe not. Likely a very small difference. Higher voltage means less loss in the wires, but again this is small but real.

1 Panel per MPPT people like me and others:
Requires more wires. It also means more MPPTs, but MPPTs are priced on power output and voltage input. Smaller controllers usually add up about the same price, and when considering redundacy, you come out ahead. Maybe enought to pay for more and larger wires. The wiring is simpler. No need for anything between panels and MPPts. No worry of burning panels from heavy shading.

But 1 MPPT per panel means more output from each panel all day long. We have 7 panels and 7 controllers and I can tell you that they are almost always at slightly different output and sometimes drastically different. 5 of them are identical with identical wiring and controllers and positioned side by side off the back of the boat. I think this is the biggest factor. Maximizing each panel overrides every other argument against one per panel. I think on a land based installation with almost no clouds series would win. On a boat, no.

All of my comments are based upon large panels (300w+). We use 8 awg and calculated under 1% voltage drop.