Mixed AWG Solar Install

[warrior 90]

Using a MPPT charge controller allows to wire solar panels in series where by U can run higher voltage and avoid resistance problems using the same wires

[smac999]

keep in mind you want less then 3% for the whole run from panel to battery. you can't just calculate each section less then 3%. if panel to transom is 3%, then transom to controller is 3% ,then controller to battery is 3%. you're up to 9%...


I'm aim for well less then 3% total for solar. it's so expensive per watt. you don't want to lose a single one. the cables are cheap vs the panels and controllers.


sounds like you are at 70+ total feet. that puts you into the whole run in 2awg.


or wire in series to drop that amps down to the controller.


at 13a you are at 6awg. then some in 8 and some in 4 would probably work out.

[NahanniV]

Originally Posted by Paul L

....
100 watt panels are cheap Chinese flex panels.

How are those panels holding up ?

Cheers,
JM.

[Paul L]

Originally Posted by NahanniV

How are those panels holding up ?

Cheers,
JM.

I've used them in the tropics for about a year and half. There output is fine. The cover is somewhat smudgy looking now. Plus the wire connectors have needed to be strengthened. So, OK for the low price, but not outstanding.

[colemj]

Originally Posted by smac999

keep in mind you want less then 3% for the whole run from panel to battery. you can't just calculate each section less then 3%. if panel to transom is 3%, then transom to controller is 3% ,then controller to battery is 3%. you're up to 9%…

I'm aim for well less then 3% total for solar. it's so expensive per watt. you don't want to lose a single one. the cables are cheap vs the panels and controllers.

sounds like you are at 70+ total feet. that puts you into the whole run in 2awg.

or wire in series to drop that amps down to the controller.
at 13a you are at 6awg. then some in 8 and some in 4 would probably work out.

Belizesailor, are your panels wired so that a single run comes down the arch, or do you have them in paired sets with multiple wires?

This post caused me to think that we were under-wired until I remembered that our 4 panels are broken into 2 banks with separate 8AWG coming off each - so each run only has half the current of the total.

If yours are wired likewise, then you only need to be concerned about half your total current on each.

Mark

[belizesailor]

Originally Posted by smac999

keep in mind you want less then 3% for the whole run from panel to battery. you can't just calculate each section less then 3%. if panel to transom is 3%, then transom to controller is 3% ,then controller to battery is 3%. you're up to 9%...


I'm aim for well less then 3% total for solar. it's so expensive per watt. you don't want to lose a single one. the cables are cheap vs the panels and controllers.


sounds like you are at 70+ total feet. that puts you into the whole run in 2awg.


or wire in series to drop that amps down to the controller.


at 13a you are at 6awg. then some in 8 and some in 4 would probably work out.

Interesting, if the loss is indeed additive with mixed gauges then that's a strong case for single gauge all the way. Do you have a reference for that? Easy enough to prove to myself with a few sections of cable and a multimeter too.

Since my original question was about mixed AWG and not sizing the circuit (which most seem to have focussed on) I was not very precise in my numbers (shoulda known better). More details on numbers below.

Distance: Approximate measured one-way distance is about 28', I used 35' for conservative estimating purposes. I expect actual to be more like 30'.

Amps: Max amps for the controller is 30, but the panels (3) are rated at 7.39 max (Impp) and 8.02 short-circuit (Isc). So if we use 8 amps as a reasonable approximation then the max amps will be 24.

Voltage: We've been assuming a nominal voltage of 12, but actual voltage from panels to controller will be higher. Max voltage (Vmpp) for these panles is 17.6, open circuit voltage (Voc) is 21.9. So 12V is going to give conservative sizing.

Target voltage drop: 3%

So if we use a few different approaches to calculate AWG then below is what we get:

Using the table Gordy posted earlier with these numbers then Amps x Round Trip Length = 24 x 60 = 1,440. Using that to enter the 12V table it falls between the numbers for 4 & 2 AWG, much closer to 4, but rounding up as specified in the table instructions gives us 2 AWG (wow!). Pretty significant difference due to rounding up...a fundamemtal problem with using tables.

If I plug the numbers into the calculator at the link below (it takes AWG as input and calculates V drop) then I get a V drop of 3.07% for 4 AWG. I can live with that.

http://www.powerstream.com/Wire_Size.htm

If I plug the same numbers into another calculator (link below) it gives me 3.00%. Close enough.

http://www.calculator.net/voltage-dr...s=24&x=66&y=17

So, my take is that 4 AWG is good.


Flaws in my logic or numbers?

[colemj]

I think the voltage drop may not be as important here - particularly if you have a PWM controller and not an MPPT. The panels will put out 17V, and you need 15V at the controller. So a 10% drop is not going to hurt you as much as you may expect.

This is even more so when you consider that in your cruising grounds, the panel output will be less than rated max do to heating. We get ~15% less when hot.

FWIW, we have 8AWG, running about the same length as you, handling 15A with no problem or noticeable decrease in terminal output. Your last calculator above shows this as 1.87% voltage drop at 15V.

Mark

[barnakiel]

Mixed AWG is at times done too.

Also Victron etc manuals quote cases. E.g when terminals are of set max size. Keep the smaller AWG as short as possible.

But best would be go series and keep the wires small. You want MPPT controller for this.

b.

[belizesailor]

Controller is MPPT, but as stated already is rated for a max of 50V which is less than all 3 panels in series. Also shading for in series and heat here in the tropics I think make MPPT less of an advantage.

[barnakiel]

Yep. 50 V will only, I think, allow for 2. Then you are in parallel, you size AWG for 12 Volts and if you must use finer wire then you want this finer wire be as short as possible.

See if the parallel combo is still within the max Amperage of your regulator.

When mixing up various AWG you still want the smaller size wire as short as possible.

b.

[warrior 90]

To have max flexability MPPT controllers should have a min capacity of 150 V.
This offers a fair choice of strings
Anything below is a waiste of money if U want serious solar power >+500 W/P

Lot´s of mambo jambo talk about MPPT

For orientation use Morningstar string calculator

http://www.morningstarcorp.com/string-calculator"

Shading in strings can only be eliminated with by pass diodes.

"in my opinion" and tested
I got the best results with amophore panels from UniSolar. Unfortunatly they were killed by Chinese panels.
Even so they needed bigger surface per W/P over all performance is superior.....no shade issues with By Pass Diodes, better low light performance, better temperature tolerance, easy instalation...and U can walk on them.
There are still some on EBAY.

It´s all about charging Amp hours to the battery

[belizesailor]

In my case I don't have physical room for > 500W without some re-engineering. These three panels (Kyocera 130) get me to 390 and thats plenty adequate for our usage.

Panels and controller (BlueSky 3000i) were a package deal from a friend who does large land based off-grid installations. These just didn't fit with the scale of projects he does anymore. Good deal for me because they were already in-country (no shipping or customs) and the price was favorable.

If I feel the need to go in series in the future I can just swap out controllers or add a second one (redundancy is good).

[warrior 90]

I know U don´t like to hear that but that controller puts U in a very unfortunate position with max 50 OVDC.
Energy on board is an important feature
Since U are allready limited by size of panels try to make the best of them wich means a MPPT controller with higher OVDC than Ur panels in series. If U have a good instalation shading is no issue and U can expect a conservative increase of 10%
As long as U don´t run big consumers like fridge etc. U should be OK

U don´t want a new controller
Since efficiency of solar systems is very low (+/- 15%) U don´t want to make it worse by not upgrading the wire

[colemj]

Originally Posted by warrior 90

Since efficiency of solar systems is very low (+/- 15%) U don´t want to make it worse by not upgrading the wire

I think you are dicing up the theoretical considerations a bit too fine.

His current 3% voltage drop will mean that instead of 17.3V being available at the controller, he will only have 16.8V. Going through an MPPT conversion, this will amount to a daily amp hour loss of <2Ahr. Probably <1Ahr in real-life conditions.

This is, of course, assuming the controller stays in bulk mode the entire time. If it starts to limit current in absorption, then there is no daily loss at all.

So if he decides to turn off a reading lamp 5 minutes earlier each night, he should be fine…

Mark

[belizesailor]

Originally Posted by warrior 90

I know U don´t like to hear that but that controller puts U in a very unfortunate position with max 50 OVDC.
Energy on board is an important feature
Since U are allready limited by size of panels try to make the best of them wich means a MPPT controller with higher OVDC than Ur panels in series. If U have a good instalation shading is no issue and U can expect a conservative increase of 10%
As long as U don´t run big consumers like fridge etc. U should be OK

U don´t want a new controller
Since efficiency of solar systems is very low (+/- 15%) U don´t want to make it worse by not upgrading the wire

As posted, its is an MPPT controler. Given that, wiring in series vs parallel isn't going to gain anything over what MPPT would already gain (no more than 10% here in the hot tropics), providing the wiring is properly sized either case.