Panel Array Wiring

[williampines]

Hi there,

I'm working on a 3.5kw 48v system for my off grid cabin I am building. The trouble I'm running into now is how to best wire the panels.

I was fortunate enough to receive 10 First Solar series six panels for free, the problem is they are very high voltage (168.2v mpp, 212.4 Voc).

My current plan is to have 5 strings of 2 panels wired in series, and then combining the 5 strings in a box before the inverter. If I'm not mistaken this would still leave me with ~425voc running to the charge controller/inverter, which seems like a helluva a lot more than most can handle.

Besides microinverters, are there any other options? Or am I just stuck spending the extra money on multiple mppts and a higher voltage inverter? Panel installation guide says modules cannot be wired in parallel.

Thanks!

 

[DThames]

Growatt makes a 5000 or 6000 watt all in one that will take 450v input. Note, you don't need to supply nor use 5000w.

 

[boondox]

Sounds to me like you have found the reason why they were free. I think I'd cut my losses and look for something else. What wattage are they?

 

[williampines]

Growatt makes a 5000 or 6000 watt all in one that will take 450v input. Note, you don't need to supply nor use 5000w.

Yeah I actually just started looking at this brand today. I do see the 5kw with the 450v input now. I think I may have skipped over it since the inverter is 240v out of the box, but now I see you can add transformer to accommodate 120/240v. I haven't looked into adding transformers at all though, not sure how tricky that would be... Thanks!

 

[williampines]

Sounds to me like you have found the reason why they were free. I think I'd cut my losses and look for something else. What wattage are they?

Hah, maybe so. My brother in law runs a big electronic recycling company and they get tons of solar panels in that manufacturers deem unsellable for whatever reason, so he was able to spare me a few. Apparently these didn't provide ideal output, but there are stickers on them showing what their output was when tested and I've more or less factored that in to system. They are rated at 365 watts each, the test stickers showing Pmax from about 300w to 350w. Plan is to just pair up in series the ones with closest outputs to reduce any further production loss. If that makes sense...

 

[Hedges]

Midnight has 200V and 250V SCC. One of those should handle the ten panels all in parallel.
Trace makes a 600V SCC.
Any of those is close to $1000.

What I use are 600V GT string inverters, cheap enough used or new-old-stock (e.g. $500). For off-grid that would require a grid forming battery inverter, maybe higher price than what you're looking for (e.g. $2000 to $3500).

SolArk is 500V (550V absolute max). Price about $4500 for 5kW model.

We may pay $0.20 to $0.50 for PV panels, so $700 to $1700 to buy same wattage of panels in your choice of ratings.

Have you figured out what you need for inverter capability? That figures into the choices too.
Battery is what could cost the most. Inverter also relates to that.

If you want to stay real cheap, could be designing around low-cost hybrid and buying compatible panels.

 

[boondox]

Midnight has 200V and 250V SCC. One of those should handle the ten panels all in parallel.
Trace makes a 600V SCC.
Any of those is close to $1000.


If you want to stay real cheap, could be designing around low-cost hybrid and buying compatible panels.

OP says the manufacturer states they can't be put in parallel. Not sure why that would be.

And yeah, might be time to look a gift horse in the mouth.

 

[williampines]

Midnight has 200V and 250V SCC. One of those should handle the ten panels all in parallel.
Trace makes a 600V SCC.
Any of those is close to $1000.

What I use are 600V GT string inverters, cheap enough used or new-old-stock (e.g. $500). For off-grid that would require a grid forming battery inverter, maybe higher price than what you're looking for (e.g. $2000 to $3500).

SolArk is 500V (550V absolute max). Price about $4500 for 5kW model.

We may pay $0.20 to $0.50 for PV panels, so $700 to $1700 to buy same wattage of panels in your choice of ratings.

Have you figured out what you need for inverter capability? That figures into the choices too.
Battery is what could cost the most. Inverter also relates to that.

If you want to stay real cheap, could be designing around low-cost hybrid and buying compatible panels.

Thanks for the info.

Yeah for whatever reason these panels aren't supposed to be wired in parallel. From the installation guide: The shorter string size and low string current of Series 6 Modules make it possible to connect multiple strings together in parallel and return a single pair of DC cables to the combiner box or string inverter.... Do not exceed system design voltage and inverter design specifications when connecting modules in a series string. This is typically ensured by limiting series strings to four modules for 1000 VDC applications or six modules for 1500 VDC applications. Do not connect modules in parallel. Strings of modules can be installed in parallel for DC collection purposes.


I've looked into the sol-ark 5kw but was hoping to find something a little less pricy than that, while also sticking with the simplicity that the all in one inverters offer.

I think a 5kw inverter will be plenty of power, with some room for expansion. I just need to find one that can handle the high voltage.

The Growatt 5k or 6k inverters look like they might be a good option, as mentioned above.

 

[MisterSandals]

They are rated at 365 watts each,

high voltage (168.2v mpp, 212.4 Voc).

365W / 212V = 1.7A

10 in parallel would only be 17A. You would need to fuse these and combine somehow, but that makes voltage reasonable.

A Victron 250 would be a good fit. I am sure there are a few others that would work.

 

[Hedges]

Yeah for whatever reason these panels aren't supposed to be wired in parallel. From the installation guide: The shorter string size and low string current of Series 6 Modules make it possible to connect multiple strings together in parallel and return a single pair of DC cables to the combiner box or string inverter.... Do not exceed system design voltage and inverter design specifications when connecting modules in a series string. This is typically ensured by limiting series strings to four modules for 1000 VDC applications or six modules for 1500 VDC applications. Do not connect modules in parallel. Strings of modules can be installed in parallel for DC collection purposes.

It says you can connect multiple strings in parallel. You can do 6s2p. You can do 6s6p.
I say you can do 1s6p, that's multiple strings in parallel, just happens to be strings of "1"

What you're not supposed to do is 2p6s. That would be connecting pairs of panels in parallel first, then connecting 6 sets of those in series.

I've pondered such things before, and realized it could put 2x Isc through one panel, or its diodes. With semiconductor junctions in parallel, you can't count on current splitting evenly. It works for resistors where V = IR. It doesn't work well for diodes, because current is exponential function of voltage, and also a function of temperature. The more current they carry, the hotter they get. The hotter they get, the more current they carry. One ends up shouldering all the current.

 

[Hedges]

365W / 212V = 1.7A

10 in parallel would only be 17A. You would need to fuse these and combine somehow, but that makes voltage reasonable.

A Victron 250 would be a good fit. I am sure there are a few others that would work.

But what is the maximum fuse rating on panel label? There is a chance these could have more than two in parallel without fuses.

 

[williampines]

It says you can connect multiple strings in parallel. You can do 6s2p. You can do 6s6p.
I say you can do 1s6p, that's multiple strings in parallel, just happens to be strings of "1"

What you're not supposed to do is 2p6s. That would be connecting pairs of panels in parallel first, then connecting 6 sets of those in series.

I've pondered such things before, and realized it could put 2x Isc through one panel, or its diodes. With semiconductor junctions in parallel, you can't count on current splitting evenly. It works for resistors where V = IR. It doesn't work well for diodes, because current is exponential function of voltage, and also a function of temperature. The more current they carry, the hotter they get. The hotter they get, the more current they carry. One ends up shouldering all the current.

Sorry I’m just getting back to this.

How exactly would it looks to do say 1s10p?

Right now I have it drawn out as 2s5p with a positive “tree” of Y branch connectors and a negative “tree” of Y branch connectors, ending in just one positive and negative cable running to inverter.

For 10 paralleled strings of “1 series” panels would I just run positive and negative from each individual panel to branch connectors for a total of 9 branch connectors in each tree? I’m confused as to how this would be any different than simply paralleling modules, but perhaps I’m looking at it wrong.

 

[Hedges]

If you build a tree of "Y" connectors, sum of current will probably exceed ampacity of wire and rating of MC connector.
With 3 or more in parallel, each string needs a fuse.
You can bring MC pigtails into a box, join on busbar or similar device (after a fuse typically for positive side.)

 

[Struc]

I would do 2s5p as you originally mentioned, utilizing a combiner box with proper fusing, etc. The combiner box will add some cost, but for free panels and the benefits of it, it would be a no brainer.

I saw some of these (in the 435w variety) on eBay recently and thought they would be a perfect match for a Growatt (I have two SPF5000ES waiting to be installed in my plethora of parts I'm putting together).

Something like this:

6-1 Solar Panel Combiner Box – 6 circuits to 1 (1 x 80A output) – Watts247 Wholesale

watts247.com

(I have not used / know nothing about this particular unit - just showing an example)

 

[Hedges]

365W / 212V = 1.7A

10 in parallel would only be 17A. You would need to fuse these and combine somehow, but that makes voltage reasonable.

I forgot these were high voltage, low current.
17A probably doesn't exceed what MC connector can handle.
17A x 1.56 multiplier = 26A, so at least last "Y" cable combining down to 1 wire should be 10 awg and rated 30A. The other "Y" could be same or lower rating.

Do the panels have a fuse rating? Probably that is low enough that it needs fuse per panel to protect against other 9 panels backfeeding into a short, but possibly not.

 

[Zwy]

Hah, maybe so. My brother in law runs a big electronic recycling company and they get tons of solar panels in that manufacturers deem unsellable for whatever reason, so he was able to spare me a few. Apparently these didn't provide ideal output, but there are stickers on them showing what their output was when tested and I've more or less factored that in to system. They are rated at 365 watts each, the test stickers showing Pmax from about 300w to 350w. Plan is to just pair up in series the ones with closest outputs to reduce any further production loss. If that makes sense...

Put them out in the bright sun and measure the VOC. A 365w panel with a 212VOC?

You could have all panels in parallel. Growatt and MPP could handle up to 250VOC with many of the SCC's. If amperage was an issue, something like the LV6548 with dual PV inputs allows you to parallel 5 panels on each input.