Another shower thought: parallel/serial PV switching

[frater secessus]

The choice of cabling panels up as parallel or serial is, for many of us,  quasi-permanent.  We makes our choices and sticks with it.

Consider this:  a switch that the panels are cabled into and the output is either serialized or parallelized.  Maybe it's mechanical like one of those Battery A/B switches and the user can manually flip it in challenging conditions.  

But if it's electronic (or a solenoid or something), further consider this:
An MPPT controller identifies problems with tracking:  maybe it's low (shade? leaf on panel?) or unstable (clouds) or maybe the algorithm is just struggling.  It flips the switch and runs MPPT on parallel panels for a while to watch for an improvement.  It might find an improvement if there is partial shade on the panels.  

Further, if electronic maybe the switch can watch for overvoltage and swap to parallel (or just cut) if the serial voltage would exceed controller capabilities.  Could provide some protection from cloud edge effect and also maybe increase yield for dark/cloudy areas like high latitudes.  

Ok, that's my crazy idea for the day.

 

[John61CT]

The good quality controllers that accept such high voltages are expensive enough you usually want to push right up to the maximums, which would eliminate that sort of flexibility. Plus you'd have to go with the thicker wiring required by the lower voltages.

If you're that prone to shading, just go one inexpensive controller per smaller sized panel, likely PWM to save $. Not really that much less efficient.

​On this topic (shading), I am looking for US retail suppliers of (or willing to ship in) gear using the new embedded "MPPT on each cell string" ICs from Maxim. Jinko "Eagle MX" is one line, others mentioned on Maxim's site are Trinapeak, and "ET COM"?

 

[djkeev]

Shower???

 

[John61CT]

Where the best ideas arrive, out of the blue.

 

[frater secessus]

The good quality controllers that accept such high voltages are expensive enough you usually want to push right up to the maximums, which would eliminate that sort of flexibility. Plus you'd have to go with the thicker wiring required by the lower voltages.

If you're that prone to shading, just go one inexpensive controller per smaller sized panel, likely PWM to save $. Not really that much less efficient.

Yeah, I think you are right on both counts.