So a parallel string facing AM sun and a parallel string facing PM sun will perform better on different controllers as opposed to on the same controller? I was under the impression this was a recommended configuration, esepcially for over-panelling, and that any negative effects were negligible. Partial shading of a series string will affect its performance, but I am also under the impression that with bypass diodes, that's even negliable anymore.
Happy to admit my ignorance. Would love to learn more if you can explain it.
If you are putting parallel panels on a common charge controller, it does work and the loss is not bad, but when the sun is near straight south but still hitting both east and west facing panels, it can't optimize both sets. One string will be over loaded a bit, and the other string will be under loaded a bit. If both strings have their own MPPT controllers, you can harvest a bit more as both strings can be optimally loaded.
With a single MPPT, if one of the strings does have a shadow on one panel, and the other string does not, then you will basically only get the power from the string without the shadow. It is like the other string has one less panel in series, so the ideal MPPT voltage is lower by one panel. With a separate MPPT controller, the bypass diodes will allow the MPPT to find the maximum power point by loading down to the voltage with the one panel being basically zero. But if there is another parallel string without a shadow, that can't happen. The MPPT will find the maximum power point of the string without a shadow. To get power from the shaded string, it would need to pull the voltage down to match the string with the shaded panel. Some "shade tolerant" MPPT system will test for this and pull the extra current to try and find this other maximum power point, but it is still sacrificing power from the non shaded string as it is dragging the voltage so much lower to find more power. So it does work with a single MPPT, it is not as good as two. Without shadows, with all of the panels seeing light, the two parallel strings works pretty well, and you can over panel with different orientations by using that trick, but it is not as good at dealing with shadows. On the good side, you will get most of the power from any parallel strings without shadows, but once a panel is shaded, you are basically losing that string.
So with 12 panels, and charge controllers that can take 3 panels in series, I would think you would do the best with two parallel strings of 3 on each of the 2 charge controllers. And if you have a case where a shadow from a pole or tree could pass across the panels, you could go a step further by having the panels that are likely to be shaded at the same time to be in each series group. So the first three panels to see a shadow on charge controller #1, then the next three on controller #2, then the 3rd group back on controller #1, and the last group back on controller #2. So as the shadow first hits, you lose the first 3 panels, but the second 3 on controller #1 are still well out in good sun. Controller #2 is getting good sun from all 6 of it's panels. But then the shadow advances onto the second group of 3 panels. So controller #2 drops to 3 good panels. Soon after, controller #1 get's the first 3 panels back again. And this continues across the rest of the array. It makes the wiring a little more complex, but I think it could be a big benefit if you can't eliminate shadows. Just the shadow from a power line can be enough to take out a string as it passes by.