Making nested series of strings of a series, does this get around shading issues?

[ddxv]

Woke up in the middle of the night thinking about this, had trouble searching for my idea and ChatGPT seemed to think it was ok/possible (whatever that is worth).

My original plan was 2 strings of 8 panels, which equates to 12a and 278v. In mornings there is a passing tree shadow that falls across the rectangular array.

Idea
Why not make 4 strings of 4 panels in series, then take the 4 strings and wire those in two series. This seems like it gets back to the original 2x 270v and 12a but with the benefits of the shadow effecting fewer panels at a time.

The idea gets even better... or maybe worse?
This idea seems like it never ends, what if I do 3 times nested! 8 strings of 2 panels in series!? Series the strings into 4 strings, series those 4 strings back into the final 2 again for the 270v/12a.

So the math seems to work out, but feels like I'm probably missing something for why I've never heard of anyone doing this or suggesting the idea.

 

[SeaGal]

This seems like it gets back to the original 2x 270v and 12a but with the benefits of the shadow effecting fewer panels at a time.

Not it won't - voltage will drop as you have less panels in series.

edited because I didn't read the question correctly!!

 

[timselectric]

Placing any number of series strings in series, just makes a large single series string.
Unless I'm misunderstanding the description.

 

[ricardocello]

4s2p2s?

   |— x — x — x — x —|         |— x — x — x — x —|
———+— x — x — x — x —+—————————+— x — x — x — x —+—————

<Aside: What a pain that was to get the ascii art to show up properly!>

This will give some redundancy on the parallel strings to handle shading, and still give 8s voltages.
I think.

Current will be limited by whichever pair of 4s strings has the lowest current.
So a shaded string still affects the total output.

 

[ricardocello]

what if I do 3 times nested! 8 strings of 2 panels in series!? Series the strings into 4 strings, series those 4 strings back into the final 2 again for the 270v/12a.

2s4p2s

 

[SeaGal]

Ok - I misunderstood the question.

@ricardocello love your 1970's graphics and the 2s4p2s.... YYURYYUBICURYY4ME

 

[ddxv]

Ok, finally woke up enough and drew out what I had originally meant:

Dream becomes reality:



So making 'nested' or hierarchical sets of strings and series just collapses down to a single string, so that was just some dream nonsense.

 

[SeaGal]

So, ChatGPT was dreaming nonsense too! No surprises there

 

[ddxv]

4s2p2s?

   |— x — x — x — x —|         |— x — x — x — x —|
———+— x — x — x — x —+—————————+— x — x — x — x —+—————

<Aside: What a pain that was to get the ascii art to show up properly!>

This will give some redundancy on the parallel strings to handle shading, and still give 8s voltages.
I think.

Current will be limited by whichever pair of 4s strings has the lowest current.
So a shaded string still affects the total output.

I had trouble following that, so I tried again and came up with this, which does allow some type of nesting, but I'm not sure if it is right?




I guess it's still different from your idea since it ends up at 140v instead of the 270v/280v (no worries about the rounding difference).

 

[Quattrohead]

Ok - I misunderstood the question.

@ricardocello love your 1970's graphics and the 2s4p2s.... YYURYYUBICURYY4ME

You bury big curvy me
You buy Curry for me
I can't figure out which one it is

 

[ricardocello]

Let’s examine the 4s2p2s case.

Current in and out and between the pairs of strings all must be equal (by basic electrical theory).
Let’s say that unshaded panels are producing 10A at the max power point, and that a shaded panel is only giving 5A.
I’m intentionally ignoring any change in tracking voltage.

       |— x — x — x — x — 10A |            |— x — x — x — x — 10A |
20A ———+— x — x — x — x — 10A +—— 20A —————+— x — x — x — x — 10A +—————  20A    270V

Now make one of the strings shaded (shown with S), see how the current drops in the second pair of strings as well as the first:

       |— S — S — S — S — 5A  |            |— x — x — x — x — 7.5A |
15A ———+— x — x — x — x — 10A +—— 15A —————+— x — x — x — x — 7.5A +—————  15A    270V

Now do the standard 8s2p that you were originally going to do, with shading on only 4 of the panels (shown with S):

       |— S — S — S — S — x — x — x — x — 5A —|
15A ———+— x — x — x — x — x — x — x — x — 10A +—————  15A    270V

The result is the same, 15A. So there is no point in doing 4s2p2s. EDIT: For this particular shading scenario.

 

[ddxv]

Let’s examine the 4s2p2s case.

Current in and out and between the pairs of strings all must be equal (by basic electrical theory).
Let’s say that unshaded panels are producing 10A at the max power point, and that a shaded panel is only giving 5A.
I’m intentionally ignoring any change in tracking voltage.

       |— x — x — x — x — 10A |            |— x — x — x — x — 10A |
20A ———+— x — x — x — x — 10A +—— 20A —————+— x — x — x — x — 10A +—————  20A    270V

Now make one of the strings shaded (shown with S), see how the current drops in the second pair of strings as well as the first:

       |— S — S — S — S — 5A  |            |— x — x — x — x — 7.5A |
15A ———+— x — x — x — x — 10A +—— 15A —————+— x — x — x — x — 7.5A +—————  15A    270V

Now do the standard 8s2p that you were originally going to do, with shading on only 4 of the panels (shown with S):

       |— S — S — S — S — x — x — x — x — 5A —|
15A ———+— x — x — x — x — x — x — x — x — 10A +—————  15A    270V

The result is the same, 15A. So there is no point in doing 4s2p2s.

Thanks for adding in the Amps and Voltages there, that helped a lot!

 

[SeaGal]

Now make one of the strings shaded (shown with S), see how the current drops in the second pair of strings as well as the first:

What about shading like this though?

|— S — S — S — S — 5A | |— x — x — x — x — 10A |
15A ———+— x — x — x — x — 10A +—— 15A —————+— S — S — S — S — 5A -+————— 15A 270V

Hmm, same result

 

[ricardocello]

What about shading like this though?

|— S — S — S — S — 5A | |— x — x — x — x — 10A |
15A ———+— x — x — x — x — 10A +—— 15A —————+— S — S — S — S — 5A -+————— 15A 270V

YYURYYUBICURYY4ME!

Yes, in that shading arrangement, the 4s2p2s wiring is superior, because you get 15A instead of 10A!

       |— S — S — S — S — x — x — x — x — 5A —|
10A ———+— x — x — x — x — S — S — S — S — 5A  +—————  10A    270V

I guess there is an optimal wiring for a specific shading scenario.
But who wants to optimize to that level? Seasons change and shadows move anyway.

I wonder if there are counter shading examples where normal 8s2p has better performance?

 

[featherlite]

You bury big curvy me
You buy Curry for me
I can't figure out which one it is

I dunno ... Google says: "Too wise you are, too wise you be; I see you are too wise for me."

 

[SeaGal]

.... is the correct answer

 

[HRTKD]

Move the array or trim/cut the tree.

Modern PV modules do as good a job as they can to handle shade. I have PV on the roof of my RV. I operate on the assumption that if one panel is shaded, they're all shaded - a worst case scenario. I have a separate set of ground deployed panels and a long extension code to work around shading on the roof.

The solar charge controller(s) you use may dictate the configuration. If the solar charge controller requires 120 volts minimum then 2s (no matter what the P) isn't going to cut it.

 

[ddxv]

Couldn't help coming back to this one more time. My brain keeps thinking about this while I'm sleeping, so I had to draw out the version that would get to the desired 280v:



I also found this Reddit thread discussing it as well:

https://www.reddit.com/r/diySolar/comments/13pa23p/can_i_connect_2_panels_in_parallel_then_connect

A comment from there, though the person cautions they are guessing:

In some situations they are not self-balancing with current and the hotter semiconductors may take more current, causing a runaway. Yes it's plausible you could push 20A through a 10A panel.

Has anyone ever done this? Is there danger of damaging the panels if I decide to try this? So tempted to try it, seems so easy, but can't risk breaking any panels.

 

[timselectric]

It's taking a risk, without any gains.
Not something that I would ever consider doing.

 

[underdog5004]

It's taking a risk, without any gains.
Not something that I would ever consider doing.

What's the risk? I've got an array that 2s2p2p, general consensus here was that was a fine configuration....

 

[ddxv]

What's the risk? I've got an array that 2s2p2p, general consensus here was that was a fine configuration....

Sorry, I keep not understanding those abbreviations. Does that mean 2 panels in series in parallel for 4 panels total?

 

[underdog5004]

Sorry, I keep not understanding those abbreviations. Does that mean 2 panels in series in parallel for 4 panels total?

I have 8 panels total. Lowest level, 2 panels in series (2s). I have two sets of series panels in parallel (2p), then I've paralleled two sets of these 4 panels, leading to 2s2p2p.

I did this configuration for shade tolerance while still getting array voltage above battery voltage.

 

[ddxv]

I have 8 panels total. Lowest level, 2 panels in series (2s). I have two sets of series panels in parallel (2p), then I've paralleled two sets of these 4 panels, leading to 2s2p2p.

I did this configuration for shade tolerance while still getting array voltage above battery voltage.

Thank you, appreciate that, the double possibilities for s (series / strings) and p (panels / parallel) was what was tripping me up.

 

[ddxv]

I have 8 panels total. Lowest level, 2 panels in series (2s). I have two sets of series panels in parallel (2p), then I've paralleled two sets of these 4 panels, leading to 2s2p2p.

I did this configuration for shade tolerance while still getting array voltage above battery voltage.

Thanks for your help, one more question:

Wouldn't your final "parallel (2p)" have only increased the amps and not the voltage? Unless you mean that you wired them in series, negative to positive which would increase the voltage?

 

[robbob2112]

I have 8 panels total. Lowest level, 2 panels in series (2s). I have two sets of series panels in parallel (2p), then I've paralleled two sets of these 4 panels, leading to 2s2p2p.

I did this configuration for shade tolerance while still getting array voltage above battery voltage.

I need a picture to visualize that.

Panels in series the voltage adds. Panels in parallel the current adds.

if you put panels in series that have different amounts of current possible they will be limited to the lowest current. If you put panels in parallel that are not the same voltage they will be limited to the lowest voltage.

You can have multiple strings independant of each other combined in parallel if the voltage is the same. If you want to have multiple strings with different charateristics then you need multiple MPPT or multiple channels on a single MPPT.