How to determine proper wire size to use between Panels and MPPT Controller

rann

Trying to understand how to best/properly determine the correct cable/wire size to use between two 440 Watt Solar Panels and an MPPT Charge Controller.  The distance between the panels and the charge controller will be between 60 and 100 feet. We are installing a 24 V system and figuring out the wire sizes for the load side is not a problem. My confusion comes from trying to understand what voltage the panels are operating at. 

BB.

Welcome to the forum Rann,

Can you tell us more about your solar panels (Vmp/Voc/Imp/Isc) or give us a link?

What brand/model/link of MPPT solar charge controller?

Roughly, where is the system located (looking for high/low temperatures).

We need to know the maximum input voltage of the MPPT charge controller and the working voltage and current of the panels.

I.e., will you be putting 2 panels in series (2x Vmp/Voc) or the two panels in parallel (2x Imp/Isc)?

Solar panels are not a "voltage source" like a battery (i.e., 24 volt battery tries to "hold' 24 volts from zero to full load Amperes).

Solar panels are (for the most part) "current sources". They output their current in propotion to the amount of sun/solar energy hitting the panel. I.e., under full sun, roughly Imp (current maximum power). And Vpanel ranges from Voc (voltage open circuit) to zero volts) depending on the load. The MPPT controller "measures" the solar panel's output current and voltage, and attempts to hold Vpanel to Vmp (figures out the equation Pmp=Vmp*Imp)... Where Vpanel tends to be ~0.85 * Vmp on a warm to hot day. And the MPPT controller controls how much current to draw from the solar array to hold Vmp at that time of day/panel temperature/how clear the day is, etc.
2 v
Here is a thread discussing in detail solar panel electrical properties... 

https://forum.solar-electric.com/discussion/5458/two-strings-in-parallel-with-unequal-string-voltages

And a post with the actual I*V Curves (current vs voltage--Same thread as link above):

https://forum.solar-electric.com/discussion/comment/359660#Comment_359660

The idea for sending solar power longer distances is to have a "high" Vmp-array voltage. Higher voltage, lower current, lower voltage drop.

But you have to match the series/parallel Varray to the Vpanel-max input voltage of the MPPT controller's specifications.

Because Vmp and Voc change with temperature and array loading, you have to make sure that Varray-cold-Voc does not exceed the controller maximum input voltage. And that Varray-hot-Vmp does not drop below that needed for the controller to charge your battery bank.

For example, many smaller controller have a Vinput maximum of 100 - 140 VDC (depending on brand/model). And your 440 Watt panels may have Vmp~40 VDC or Vmp~72 VDC rating (again, depending on brand and model).

2x 40Vmp panels in series in a warm climate may work great. But 2x72Vmp panels will over voltage most MPPT controllers....

-Bill

yachtmoonfleet

Hello Bill, I have read you post with interest, I am in a similar position and need advice. I have 9 JAM60S21-370/MR.
panels Voc 41.30 Vmp 34.23 Pmax 370W panels in 3 groups of 3, wired in Series, all cables to a combiner box. The distance from the panel groupings to the MPPT controller Victron Energy 250/100 MC4. Is 100 metres (320ft)  
 Would I be right to wire the panels in series and my important question what wire would I need to connect from the combiner box to the MPPT controller with minimum loss of current. Grateful thanks Alan

BB.

Welcome to the forum Alan.

I will give a quick answer here, but I ask that you create a new thread  so we can deal with your questions and not confuse original poster's thread with your hardware questions (not that they ever came back).

Just a quick look at the spec sheet... Your controller has two pairs of MC4 connector inputs. Each rated for 30 amps max @ 245/250 VDC max array voltage. I did not find a manual for this controller--As always, make sure to read the manual thoroughly.

Also need Imp (current maximum power). From one spec sheet Imp = 10.81 amps...

• Array working voltage = 3 series * 41.30 Volts = 123.9 volts Vmp-array-std (standard conditions)
• Array working current = 3 parallel * 10.81 amps = 32.43 Amps
• Array distance (single pair of cables) = 320 Feet
• Typical voltage drop (power drop) ~1% to 3%

I am in the US, so I will use US based voltage drop calculator and convert (do you sue mm or mm^2 wire sizes).

First you need to figure out your minimum working temperature and how many panels you can put in series. Guessing you are in London UK, it does not get very cold (i.e., mountain top / arctic temperatures)... You could probably put 4 panels in series (2x4s=8 panels, or 3x4s=12 panels). Higher voltage Vmp-array (and Voc-cold-array), is better as it allows lower current and higher voltage drop to keep losses and wiring costs low.

In your case, you have 9 panels in 3s x 3p. So the calculations.

First, the controller MC4 connectors are rated for 30 amps maximum. With 3 strings in parallel and Imp-array=32.43 amps, that exceeds the 30 amps per MC4 pairs into the controller. You will need to "pair" 3s x 2p for one cable run and 3s * 1p for the second run (or simply run 100 meters with 3 pair of cables and a 2-1 combiner just before the controller... Also because you have 3 parallel strings (or more), each string needs its own series protection fuse/breaker per string rated at 20 amps per string.

2x voltage drop/wire size calculations... 3s*1p Imp, 3s*2p Imp.

• Vmp-array=123.9 volts
• Imp-array-1=10.81 amps
• Imp-array-2=21.62 amps
• distance=320 feet

https://www.calculator.net/voltage-drop-calculator.html?necmaterial=copper&necwiresize=4&necconduit=pvc&necpf=0.99&material=copper&wiresize=0.4066&resistance=1.2&resistanceunit=okm&voltage=123.9&phase=dc&noofconductor=1&distance=320&distanceunit=feet&amperes=10.81&x=41&y=19&ctype=nec

Result 6 AWG for 3s*1p string

Voltage drop: 3.40
Voltage drop percentage: 2.74%
Voltage at the end: 120.5

https://www.calculator.net/voltage-drop-calculator.html?necmaterial=copper&necwiresize=7&necconduit=pvc&necpf=0.99&material=copper&wiresize=0.4066&resistance=1.2&resistanceunit=okm&voltage=123.9&phase=dc&noofconductor=1&distance=320&distanceunit=feet&amperes=21.62&x=59&y=20&ctype=nec

Result 2 AWG for 3s*2p strings

Voltage drop: 2.61
Voltage drop percentage: 2.11%
Voltage at the end: 121.29

Converting AWG to metric:

AWG Number	Ø [Inch]	Ø [mm]	Ø [mm²]
6/0 = 000000	0.580	14.73	170.30
5/0 = 00000	0.517	13.12	135.10
4/0 = 0000	0.460	11.7	107
3/0 = 000	0.410	10.4	85.0
2/0 = 00	0.365	9.26	67.4
1/0 = 0	0.325	8.25	53.5
1	0.289	7.35	42.4
2	0.258	6.54	33.6
3	0.229	5.83	26.7
4	0.204	5.19	21.1
5	0.182	4.62	16.8
6	0.162	4.11	13.3
7	0.144	3.66	10.5
8	0.128	3.26	8.36
9	0.114	2.91	6.63
10	0.102	2.59	5.26

One other option... Given are (maybe?) around London, UK, your maximum cold temperatures (-8C record low?), you would easily be able to put 5s*2p panel array on your controller for 10 panels (if warmer than -8C, you could just about put 6s panels together and just get under the 250 VDC max array voltage limit).

Anyway, just for discussion, look at 5s*2p for 10 panel array.

• Array working voltage = 5 series * 41.30 Volts = 206,5 volts Vmp-array-std (standard conditions)
• Array working current = 2 parallel * 21.62 amps (one pair of cables to controller)
• Array working current = 1 parallel * 10.81 amps (two pair of cables to controller)
• Array distance (single pair of cables) = 320 Feet 
• Typical voltage drop (power drop) ~1% to 3%

First is to parallel 5s*2p at the array, and one pair of cables to controller (note, 2p strings, you do not need fuses/breakers per string--another cost savings):

https://www.calculator.net/voltage-drop-calculator.html?necmaterial=copper&necwiresize=4&necconduit=pvc&necpf=0.99&material=copper&wiresize=0.4066&resistance=1.2&resistanceunit=okm&voltage=206.5&phase=dc&noofconductor=1&distance=320&distanceunit=feet&amperes=21.82&x=43&y=28&ctype=nec

Result 6 AWG one run of cable @ 21.86 amps 5s*2p

Voltage drop: 6.85
Voltage drop percentage: 3.32%
Voltage at the end: 199.65

Second is run each 5s*1p from array to just before controller, and plug into the two pairs of MC4 connectors (again, no series fuses needed):
https://www.calculator.net/voltage-drop-calculator.html?necmaterial=copper&necwiresize=3&necconduit=pvc&necpf=0.99&material=copper&wiresize=0.4066&resistance=1.2&resistanceunit=okm&voltage=206.5&phase=dc&noofconductor=1&distance=320&distanceunit=feet&amperes=10.81&x=65&y=14&ctype=nec

Result 8 AWG two parallel runs of 5s*1p from array all the way to controller

Voltage drop: 5.40
Voltage drop percentage: 2.61%
Voltage at the end: 201.1

A few notes:

3% is "Just a number" picked from the NEC (US National Electric Code) for maximum voltage drop in house wiring. For solar array wiringa  on a properly configured Array/MPPT controller, there is nothing magic about 3%... You could pick 4%, 5%, etc. and everything will work fine. Just more losses (vs cost of copper wire).

What voltage battery bank do you have... More or less, A 3,700 Watt array (10 panels) is about the maximum cost effective array for this controller charging a 24 VDC battery bank. A 12 volt battery bank would require 2x of these charge controllers in parallel to "manage" such a large array... A 48 VDC battery bank would be very happy, and you could even add more panels, if you wanted.

Running the Array Vmp/Voc voltage as high as you can really saves on copper wiring prices--Especially for such long wire runs at this. I have not gone through the cost of wiring here, but you could end up saving money by purchasing a 10th solar panel and wiring out a 5s*2p array (again assuming -8C is your low temperature).

Highly suggest that you do paper calculations (like these and others) before purchasing hardware. Much cheaper to "test" the design on paper than to buy mix & match hardware first.

As always, after doing the "what if" paper calculations, make sure you meet all of the controller requirements. Like 30 amps max per MC4 wire, Max 245/350 VAC in cold weather, Minimum Vmp-array voltage (not an issue in this case).

Wiring cost wise--Most of the cost is the copper itself... So running one large cable with all the current vs splitting the current in 1/2 across two smaller cables--The amount of copper used is the same (for the same voltage drop). So that really does not matter.

You also need to do the same type of voltage drop calculations from controller to battery bank... More or less, suggest 0.05-0.10 volt drop for 12 volt battery; 0.1 to 0.2 volt drop max for 24 volt battery, and 0.2 to 0.4 volt drop max for 48 volt battery.

For higher power systems (larger AH battery banks, large AC inverters, etc.), higher battery bank voltages (i.e., 12->24 or 48 volts) usually is the best choice. Wiring large amounts of current at low battery voltages can end up with very thick and expensive copper cables...

Anyway... It's a start. Any questions/corrections? And I suggest you start your own discussion/thread for further discussions for your system.

-Bill

BB.

Just for a quick cost of copper point... 2 AWG @ 3s*3p array vs 6 AWG @ 5s*2p single string runs:

2 AWG = 33.6 mm^2 for 3s*2p = 9 panel array
6 AWG = 13.3 mm^2 for 5s*2p = 10 panel array
33.6/13.3=2.5x more copper for 100 meter array run for 3s*3p array vs a larger 5p*2s array

In theory, your 3s*3p array 100 meter cable run would cost about 2.5x more than a 5s*2p cable run... And you have a 370 Watt larger array too with the "cheaper" cable run.

In the US, 2 AWG cable from the local hardware store is around $2.37 per foot.

6 AWG is around $0.82 per foot.

640 feet (+ and - cables) * $2.37 per foot = $1,516.80 of 2 AWG cable
640 feet * $0.87 per foot = $556.80 of 6 AWG cable

A 10th 370 Watt panel is usually less thab $370 (perhaps down to $175 excluding shipping and taxes)...

If you can fit a 10 panel array (all panels need to be shade free, and pointing the same direction, no 9 on east side of use, and 1 on west side of house)--And you could save hundreds of USD$ in copper wiring costs.

Anyway... Just some more thoughts.

-Bill