wiring panels with combiner box

Re: wiring panels with combiner box

larry,
if i read this rightly, after the series it will have the combiner right there combining all strings? you can't combine all strings when 2 or more controllers will be used as no pv or series string of pvs should connect to more than 1 controller when being recombined back to a common battery bank. you could send 2 home runs back over that 50ft distance to go to their appropriate controller. do not have the controllers downconvert and then send over the 50ft wire run, but have the higher voltages as long as possible to reduce v drop losses.

now you did not specify what the battery bank voltage will be for this, but i'll go through it up to the controllers. you will need to account for the wires to the batteries and the battery interconnecting wires as well and these v drop percents get added to the v drop percents i will give as an example here. you also need a total on the wire length for tying the series strings together up to the combiner. yes, this includes the ones built into the pvs. say as an example the run to the combiner from the pvs is 10ft each pv has 3 foot pigtails then it would be the 10ft negative run wire from the combiner to the -pv pigtail that is 3ft. next each interconnect will have 6ft of wire with 4 interconnects so that's 24ft of wire. lastly the positive 3ft pv pigtail to the 10ft + combiner run wire. that's 50ft of (#10 i believe you said) wire you need to figure for a voltage drop at 11.48a just for the string to go to the combiner. if all of the string lengths are identical then you only need to figure 1 such string before the combiner otherwise pick the longest or worst case series run. anyway this is now about a v drop of .73v and at 91.5v vmp this is about .8%. this is quite a bit so you can see the importance of keeping wire resistance down. the -10ft and the +10ft combiner runs may be of a bigger wire as you don't have to stay with #10 just as long as you don't cut the mc4 connector off of the pv. use a small #10 mc4 wire that has been cut in half so as to connect that with the run wire, which will now be a bigger wire to match the 50ft run wires to the controllers.

taking the original lengths of all of the pvs' pigtails and adding another 3ft for an adaption. 2 3ft pigtails each pv for 5 pvs is 30ft, + the 3ft #10 adapter is 33ft of #10. this is .5273% v drop. adding say 8.5ft extensions fron the adaption wire will add another 17ft of wire, but let's use #4 here. this will add .0676% v drop. we are now up to .5949%.

so far so good. now it is recommended for the total % from the pvs to the batteries to be under 2% as recommended by some and others have said 3%, but the lower this number is the better so i'll show for up to 2% and not account for the battery interconnections. let's allow .3% for the cc to battery run, but this may be higher depending on the battery voltage you are opting for. remember this area is now operating at the battery voltage and the v drop % will be referenced to that and not the higher pv voltages. the .3% would be good for a 6.5ft run of #4 wire at 34.44a (3 strings x 11.48a) and a 48v battery bank. (mppt action could bump up the current and the v drop %) for lower battery voltages the length may need trimmed back and the use of a thicker wire to keep things within reason. with the .3% added we will see around .9% v drop (rounding up). if being allowed to go to 2% total then the 50ft run can go to 1.1%. as it turns out keeping with the #4 would put the v drop % to about 1.2%. pretty close, but over. this would be at or below 1.1% if you reduced the run length to 46ft or 92ft total for both the + and - run wires. this could also be made up in other areas like as in the cc to batteries for the length might be shortened and/or thicker wires used.

this gives you a ball park example of what you could be looking at. of course the run with only 2 strings will not have as bad of a v drop loss because the 2 strings will only be outputting in the neighborhood of 23a, but keeping the wires as if it had 3 strings would allow for a future expansion of 1 string and still being good on the v drop %.

you can try to follow this or do it yourself with the v drop calculator in my signature line.

Re: wiring panels with combiner box

larry,
i'm just trying to be sure that you do not combine all 5 strings together and then off to the 2 fm80s. yes, combine 3 strings to go to 1 cc and combine the other 2 strings to the 2nd cc through their own 50ft homeruns.

just so we are understanding each other correctly, you will have the combiner, then the 50ft wire runs, and finally the 2 fm80s? i agree with that arrangement as it keeps a higher voltage across the long wire run. i also do not know how closely i laid out the whole actual scenario as that was just an example of how it may be. i don't know some aspects yet, but you can plug values into the v drop calculator as well to confirm it will work to your satisfaction. if you are looking for something under 3% then the #4 would probably suffice quite well being closer to the 2% area, but still higher than that.

if it were me i'd be happy with the #4, but you can do with it as you please. you can make it higher in % or lower as the amount of losses will be your choice to make. now i didn't run calculations for #6 or #2, but #6 will have higher losses and cheaper in costs compared to the #4 wire. the #2 would be more expensive and lower losses than the #4 wire. any changes in the wire lengths compared to my example will also change accordingly with longer being worse. exact figures for all wire sections you would need to plug into the calculator as there are many variables.

Re: wiring panels with combiner box

Are you certain midnight green lighted two controllers from the same combiner?

Midnite says you can do this with the -6 model in their MNPV manual:

The MNPV6 enclosure can be split into two sections. This is sometimes done in 12 and 24V systems where
more controllers are required for additional power. For instance in a 24 volt system using the MX60 charge
controller, you are limited to about 1600 watts of PV per controller. If using Kyocera KC130 modules, you can
make three strings of 4 modules in series. This adds up to 1560 watts per controller. That is a good match of
PV vs. controller capability. The MNPV3 can accommodate this arrangement directly, but the MNPV6 can
accommodate two of these systems, thus saving wiring, space and money. See the following figure on splitting
the busbar into two systems.