connecting solar panels

[100kwh-hunter]

I think this will be the right way to connect my solar array?
Every paneel wil give (death south) ~40v 10 A.
3 panels combined will give ~120v at 10A
I would like to connect two sets of 120v 10a to one to make 120v 20A, see drawing.
this times two would give me 120v at 40a? charging at 2 times a 150v 30/40A victron charger.
But i have a couple of meters of 4mm2 wire/awg 11, 200 meters red and black combined 400 meters/1312 foot
So i can connect it what ever is best for cables with minimum of los.
I think 120v 20A aka 2 sets of panels, one with 16 meters of cable and the other set with 14 meters of cable will be enough to get to the solar/battery charger(will be Victron!) or??? you fill in the length or is it better to do one set of 3 panels 160v 10a with a separate wire or even chase 120v at 40A thru a 4mm2/11awg cable?
Please give me your thoughts, if my jibber jabber chat make any sense(sorry English is not my native language, and i refuse to use google translate)
or even 160v at 10A x 3 160v at 30A?

For finishing my roof atm (did entirely new btw) my easiest way would be connect 3 panels to 120v 10a or ~40v at 30 A? what are your thoughts?
I would like to connect them from top to botten all 3 and not from right to left all 4(160v and 10a or 40v and 40A) or there must be a lot of benefits.

I would really appreciatie your thoughts and the why it is the best.
Many thanks in advance with best regards Igor K

 

[OffGridInTheCity]

It doesn't matter PV power wise = e.g. 4s3p (160v@30a) or 3s4p (120v@40a) are both 4800watts.

The *key* thing that matters is what the Victron charge controller can accept. You must be below the max voltage and its good to be <= max overall wattage. Can you share the Victron model / specifications? Does it have one PV connection or 2 separate ones?

Secondarily the wiring / combiner box placement may matter as long as it does not violate charge controller specifications. For example, it's OK to have 2 separate 'sets' - e.g. 3s2p and 3s2p and then combine those OR 3s4p and combine all 4 strings. In my case I ran extra wire so that all 5 of my strings came down to the combiner box - took extra wire - so I wasn't dependent on strings being combined on the roof where I can't reach them. This let's me verify each separate string at the combiner box.

 

[italianuser]

Uhm, correct me if I'm wrong, these are the formulas I use.

The longer the cable the higher the voltage drop, that is a loss between the panel and the inverter/charger. Plus, the higher the amperage the higher the loss.

So, supposing a 10A current on a 32mt 4mm2 cable (16mt x 2, there and back), plus considering a 0.02ohm/mt resistance for a copper cable, gives a voltage drop of:

a) Calculate cable resistivity
R = K*L/S = 0.02ohm * 32mt / 4mm2 = 0.16ohm

b) Calculate voltage drop
V = I*R = 10A * 0.16ohm = 1.6V voltage drop.

If amperes running in the cable are 30A it becomes:
V = I*R = 30A * 0.16ohm = 4.8V voltage drop.

In any case I'd consider using at least a 10mm2 cable, reduces voltage drop by a 2.5x factor.

 

[OffGridInTheCity]

Its common to use 10awg PV wire - with a cross section of 5.26mm2 per my chart. The OP says he's doing 11awg (4mm2). 11awg is in range of 'normal' for 10-20a. So individual 10a per series/set of wires or 20a for a combined series.

My arrays of 3s5p have 100ft to 50ft of 10awg to the combiner boxes at 9a @ 110v on each set of wires. Then 45a @ 110v from combiner to charge controllers is 150ft of 6awg in conduit. Later portions of the wire in the conduit reach 50C - all pretty normal from what I can tell. An electrician actually selected this wiring.

I see posts about voltage drop but it seems to me the key question is power loss. Its it 50w or 200w or.? For me, this is not a key issue for 150ft runs of reasonable sized wire. But hey! I'm a layman on this for sure. What keeps me up is the 15% losses at my inverters!!!

 

[italianuser]

For finishing my roof atm (did entirely new btw) my easiest way would be connect 3 panels to 120v 10a or ~40v at 30 A? what are your thoughts?

Well, yes, I agree with you @OffGridInTheCity. If cable length is big it makes sense to use a bigger cable. For your 150ft, using my worst case calculated above, you'd be losing nearly 15V from your panels to the controller, which is more than 10% on a 120V panel setup. Reducing losses, where you can (not on the inverter maybe!) does make sense.

To answer @100kwh-hunter's question: "connect 3 panels to 120v 10a or ~40v at 30 A? what are your thoughts?" -> I'd go for a lower amperage and higher voltage to reduce loss to the minimum (the loss will be on the voltage MPPT controller sees). And, again, if possible use bigger cables (I get 10mt/~30ft of 10mm2 for 20€ excl.shipment).

 

[ajw22]

The *key* thing that matters is what the Victron charge controller can accept. You must be below the max voltage [...] Can you share the Victron model / specifications?

Also, the model or spec sheet of your PV panels. Key issue is whether the "~40v" is Vmp (Voltage for max power) or Voc. If Vmp, the total Voc could potentially go above 150V and damage the charge controller. Especially if you live in a cold climate, as PV panels produce ~0.3% more voltage for every degree C below 25C.

If you haven't bought the charger yet, Victron also makes 250V models.
If you have surplus cables anyway, definitely run 3 (4?) separate sets of cables to reduce losses in the cables.
Here's an online calculator for cable voltage loss: https://www.calculator.net/voltage-drop-calculator.html

 

[OffGridInTheCity]

Well, yes, I agree with you @OffGridInTheCity. If cable length is big it makes sense to use a bigger cable. For your 150ft, using my worst case calculated above, you'd be losing nearly 15V from your panels to the controller, which is more than 10% on a 120V panel setup. Reducing losses, where you can (not on the inverter maybe!) does make sense.

To take this a step further - let's say 15v of a 3s 120v series of panels is lost. What is the actual power loss? What is the formula?

For example, if the formula is 15v loss * 9a (panel series amps) it would be 135w per string. In my case I have 15 strings (45 panels) so it would be 15 * 135w = 2025w loss on a 12,875w array = 15.7%. In theory I'd only be able to achieve 10,850w out of the array (e.g. 12,875w - 2025w).

But that's not what I see in practice as I do hit 12,000+ watts and I wonder if MPPT mitigates this somehow? BUT without proper measuring equipment and the shear variability of panels/angles/sun/temps/time-of-day its like pushing two ends of a string to verify efficiency from panels -> charge controller output and anecdotal feelings are no match for actual science.

So maybe I'm just whistling past the graveyard?!?

 

[Redpacket]

For the MPPT choice, you must use Voc open circuit panel voltage at the coldest daytime temp you will get.
Eg: number of panels in series X Voc X panel temp rating factor for your min temp = total Voc for MPPT min voltage.

Vnoct (normal op temp, which is less than Voc) must be above min startup voltage for the MPPT.
Eg if you had a "48V" battery bank system & only 1 panel in series, the Voc & Vnoct would be below the battery voltage & the MPPT would not startup.

Re wiring, I ran multiple cables (panels to MPPT gear) to get the least losses.
Specs here in Australia require only a few % loss in cables to keep overall efficiency high.

OffGrid if I was really loosing ~15.7% power just in cables, I wouldn't sleep at night!!!
You must have more cables ie so you're not getting a 15v cable drop loss?

 

[100kwh-hunter]

Wow, thank you all for your replays, more info than i expected, enough detailed info to proceed, i will try to give answers and explanations on what i am about to do with your info, thanks again, this takes a lot of painfully thinking away.

It doesn't matter PV power wise = e.g. 4s3p (160v@30a) or 3s4p (120v@40a) are both 4800watts.

Somehow i ended up with just 1200W? becouse i thought 4.8kw was to much. So with 2 hours of sun we are good for a whole day.

an you share the Victron model / specifications? Does it have one PV connection or 2 separate ones?

I still must go to the store for this, i am still open for suggestions, i think a dc/ac converter and one or two solar chargers with a transfer switch for mains and battery, battery empty switch to main, and vica versa.
(280A lifepo4 at 48V nom, atm, will be 4 times before end of the year)
Or a all in one go like the multiplus 5000/48v/70-50

In my case I ran extra wire so that all 5 of my strings came down to the combiner box - took extra wire - so I wasn't dependent on strings being combined on the roof where I can't reach them. This let's me verify each separate string at the combiner box.

I think i am going to do the same, give every string 120v/10a his own set of wires to a combine box that is also a dc fuse box, into the main breaker cabinet, (with dc fuses) before leading them into the charger.

The benefits for me i think: i can sleep better with knowing that: #1 i dont have connectors under my roofing tiles(cables direct from panels to inside to the fuses)
#2 The cables wont warm up and #3 the losses would be minimal? if i understand it all correct.
So i can put a bundle of cables into a pvc squire "pipe" under my gutter with special separation clips, i think you know what i mean

b) Calculate voltage drop
V = I*R = 10A * 0.16ohm = 1.6V voltage drop.

I downloaded the v drop calculator and he is stating that in my longest cables i will have a v drop off 0.42v in awg 12 if i fill in awg 10 i will have just 0.25v drop both with 10a btw.

I think this is negligible? my scholing for this was over 35 years ago btw.... and i changed course sinds then.

I'd go for a lower amperage and higher voltage to reduce loss to the minimum

done....i set on 120v at 10a.

And, again, if possible use bigger cables (I get 10mt/~30ft of 10mm2 for 20€ excl.shipment).

With the buy of the solar panels i also bought +/- 150 meter/450 ft of black and +/- 150 meter/450 ft red, both 4mm2/11 awg
I only must buy a thick grounding wire and pound a ground rod into the ground(water is pretty high here) , i dont think i will get away with a wire in a bag filled with dirt in the gutter

Also, the model or spec sheet of your PV panels. Key issue is whether the "~40v" is Vmp (Voltage for max power) or Voc. If Vmp, the total Voc could potentially go above 150V and damage the charge controller. Especially if you live in a cold climate, as PV panels produce ~0.3% more voltage for every degree C below 25C.

Adding some more difficulties into the mix.
The printed sticker is barely readable, i must go with what my meter is telling me.
This is what i could read under a 5000k led light:

vmp 34.4v
imp 2.9 a
isc 3.0 a
voc 43.2
pmin 95.1 w
maximum system opp v 600v
bypass diode 6 a
series fuse rate 5 a

I think those are good numbers with what my dmm was telling me?

If you haven't bought the charger yet, Victron also makes 250V models.

I am open for suggestions, please feel free.

If you have surplus cables anyway, definitely run 3 (4?) separate sets of cables to reduce losses in the cables.

Cables are not surplus and i have even bought to much, but in the future it will come handy, i am almost sure
It was really a lucky buy, before the prices went bunkes...

But that's not what I see in practice as I do hit 12,000+ watts and I wonder if MPPT mitigates this somehow? BUT without proper measuring equipment and the shear variability of panels/angles/sun/temps/time-of-day its like pushing two ends of a string to verify efficiency from panels -> charge controller output and anecdotal feelings are no match for actual science.

different example, for my train hobby in the garden i use 3.6v li ion cells to power a turnout on my longest cable~20 meters(40 meters back and toward) on a 30awg cable, i put in 3.8v while the engine is using 0.8a i got no significant v drop.

Eg: number of panels in series X Voc X panel temp rating factor for your min temp = total Voc for MPPT min voltage.

I am not sure if i understand this one.

Eg if you had a "48V" battery bank system & only 1 panel in series, the Voc & Vnoct would be below the battery voltage & the MPPT would not startup.

This i get.

OffGrid if I was really loosing ~15.7% power just in cables, I wouldn't sleep at night!!!

Why, the cables are getting hot?
At a certain point the cables must go thru the wall with 20 cm of pir foam sheets, i would like to maintain the insulation value and no condese so i will make tiny holes and foam them close, or stuf the hole with rockwoll? regarding warming up.
The pir sheets can have a max temp of 110C combustion at ~170/180C but self extinguishing.

Folks thank you all very much for your answers and in advance for answers that will come.
I am going with 120v 10a full lenght cables with 1 connector to the panel.
The longest cable will be 16 meters max.

With best regards Igor K

 

[italianuser]

To take this a step further - let's say 15v of a 3s 120v series of panels is lost. What is the actual power loss? What is the formula?

The formula is the one I shared, there's only a small offset in real systems due to material difference and measurement devices accuracy/precision. It should be quite easy to calculate power loss having the voltage loss.
Disclaimer: I did not invent the formulas, that's physics guys LOL

For example, if the formula is 15v loss * 9a (panel series amps) it would be 135w per string. In my case I have 15 strings (45 panels) so it would be 15 * 135w = 2025w loss on a 12,875w array = 15.7%. In theory I'd only be able to achieve 10,850w out of the array (e.g. 12,875w - 2025w).

Uhm, nope.
I calculated the losses for you, in your two main segments; they're in an acceptable loss range (very good ). Your cable choice was excellent.

First segment:
I considered 75ft (you say 100ft to 50ft) 10AWG (5.26mm2) 9A@110V

This segment loses 0.78V, very good result.

Second segment:
150ft 6AWG (13.3mm2) 45A@110V

This segment loses 3.09V, still an acceptable value, below 4% loss (4.4V would be a 4% loss).

But that's not what I see in practice as I do hit 12,000+ watts and I wonder if MPPT mitigates this somehow?

I don't think MPPT can do anything about losses, a part from doing it's normal job.

BUT without proper measuring equipment and the shear variability of panels/angles/sun/temps/time-of-day its like pushing two ends of a string to verify efficiency from panels -> charge controller output and anecdotal feelings are no match for actual science.

I don't agree with this. Physics laws can be used effectively to calculate everything in your electrical system, physics it's an actual science.
If our instruments are not very accurate/precise we can take offsets into account.

So maybe I'm just whistling past the graveyard?!?

No LOL

In the moment your solar system, in that exact moment, is making a lovely 12875W you're losing:

• 0.71% in the first segment;
• 2.81% in the second segment.

I'll suppose, not 100% sure, there's an inverse proportion between the voltage loss and the amperes flowing.

 

[OffGridInTheCity]

Thanks for the detailed (and positive) post!

In the moment your solar system, in that exact moment, is making a lovely 12875W you're losing:

• 0.71% in the first segment;
• 2.81% in the second segment.

These numbers make sense - thank you for calculating them.

I'll suppose, not 100% sure, there's a inverse proportion between the voltage loss and the amperes flowing.

But if I can double-down a bit because I see these threads/discussions often.... what I don't understand is how to translate voltage loss to actual power loss. By loss I mean - lost kwh(s) of PV compared to 0v drop.

 

[OffGridInTheCity]

Why, the cables are getting hot?
At a certain point the cables must go thru the wall with 20 cm of pir foam sheets, i would like to maintain the insulation value and no condese so i will make tiny holes and foam them close, or stuf the hole with rockwoll? regarding warming up.
The pir sheets can have a max temp of 110C combustion at ~170/180C but self extinguishing.

One comment here - in the US, anything that penetrates a wall must be in conduit. For US, all DC wiring > x Volts within the wall or house must be in conduit (e.g. AC does not have this requirement). Maybe check you're building codes?

Practical reasons....
I understand the reason for conduit for wall penetration is to stop fire from going thru the wall. But I'm no expert

 

[italianuser]

But if I can double-down a bit because I see these threads/discussions often.... what I don't understand is how to translate voltage loss to actual power loss. By loss I mean - lost kwh(s) of PV compared to 0v drop.

I found the complete formulas online, it also accounts for cable temperature. And there's a nice online calculator (link is below).

The answer to your question is: percentage of power (energy) lost in watts is equal to the percentage lost in volts, wow; well, it was quite easy, actually; so watts lost are 0.71% in the first segment; and from the "remaining" watts subtract another 2.81% for the second segment.

The formulas and the online calculator are on this website: https://photovoltaic-software.com/solar-tools/voltage-drop-calculator-dc-ac

Notes on the online calculator

• It doesn't let you insert cable size in AWG, only in mm2;
• The cable length you insert is one way (the distance between source and destination); and will then multiply the value by 2 to get the right value (factor b in the formula); I directly used the cable length x2;
• Resistivity for copper cable is set to 0.017ohm/mt (I use 0.02);
• The difference in results between the formula I wrote in the previous posts and the online calculator (temp set at 40°C/104°F) is ~8%.

Sorry Igor! I don't want to hijack your thread!!

 

[Redpacket]

I think i am going to do the same, give every string 120v/10a his own set of wires to a combine box that is also a dc fuse box, into the main breaker cabinet, (with dc fuses) before leading them into the charger.

This is a good idea.

The benefits for me i think: i can sleep better with knowing that: #1 i dont have connectors under my roofing tiles(cables direct from panels to inside to the fuses)
#2 The cables wont warm up and #3 the losses would be minimal? if i understand it all correct.
So i can put a bundle of cables into a pvc squire "pipe" under my gutter with special separation clips, i think you know what i mean

Yes, good reasons.
Any wires outside should have conduit to protect from the sun.

I downloaded the v drop calculator and he is stating that in my longest cables i will have a v drop off 0.42v in awg 12 if i fill in awg 10 i will have just 0.25v drop both with 10a btw.

I think this is negligible? my scholing for this was over 35 years ago btw.... and i changed course sinds then.

done....i set on 120v at 10a.

This will be good.

The printed sticker is barely readable, i must go with what my meter is telling me.
This is what i could read under a 5000k led light:

vmp 34.4v
imp 2.9 a
isc 3.0 a
voc 43.2
pmin 95.1 w
maximum system opp v 600v
bypass diode 6 a
series fuse rate 5 a

I think those are good numbers with what my dmm was telling me?

Looks good.
So the max voltage for the MPPT will be 3 panels x 43.2 x (1 - 0.32%/°C)
I got the - 0.32%/°C from my panels but yours will be similar.
At your place what is your cold winter day temp? Eg -5 degC?
43.2 is for 25 degC so if -5 at your place, it's = -0.0032 x (-5 -25) = -0.0032 x -30 = 0.096 so 9.6% higher volts on cold day
Vmax = 3 x 43.2 x (1+0.096) = 142V max.
So you will need a 150V MPPT.

I am not sure if i understand this one.

Why, the cables are getting hot?

You don't want cables to get hot only a tiny bit warm.
Hot = wasted power + dangerous

 

[100kwh-hunter]

Those panels are not 10a blush blush...the other two are
I am checking at the moment the smaller panels and they are at best 3A at there short circuit.
so times 12 in that array i wil get at best 1200w per hour
The two bigger ones are each 315w at also 43voc...plus one small one100w

So the first field will be 3 x 120v at 3a.
The red cable is 4mm2 and the black cable is 6mm2, i dont think there will be a problem what so ever.
Tomorrow i will make some conduits thru the roof and feed the cable to the fuse box.

Thanks again for your help and support!
With best regards Igor

 

[100kwh-hunter]

Sorry Igor! I don't want to hijack your thread!!

You did not, what you did is starting a war between my brain and my memory.
My brain: come on you know this....my memory: nope...

Anyway i would like to go to a lower system voltage, cous the solar charger controller for 150v is 380 euro and for 100v max 1160watts is just 140 euros...
Benefit i can connect those two bigger ones to a other solar charger and keep 100 euros in my pocket....
One bummer...i pre cut and prepared all the cables for 120v....

So back to the drawing board and making new wires...or is there anything wrong that i ga back to ~80v solar array?
Thanks in advance!
With best wishes Igor

 

[Korishan]

No problem with having bigger gauge wire. If there are sized for 150V, and you are dropping down 100V, this will give lower voltage drop on the wires. No since in taking the savings of the smaller controllers and spending it on new wire.
As long as the runs are the same length, and connect the same way, no point in redoing the wiring.

 

[Redpacket]

No problem with having bigger gauge wire. If there are sized for 150V, and you are dropping down 100V, this will give lower voltage drop on the wires.

Voltage drop depends on current not voltage, so the voltage drop on the cables would be the same if you had two panels in series ie 2s, or 3s or other, as long as it's no more in parallel.

As long as the runs are the same length, and connect the same way, no point in redoing the wiring.

+1 !!

 

[100kwh-hunter]

Oke here are the stats:
Some cables are shorter for about 2 to 6 meters!
I still use 4mm2 for the plus and 6mm2 for the negative side.
Per string of solar panels i have 77-80V and a 5.5A
Longest pos cable is just 13 meters, neg is 13,5 ish
Except for the far away roof...22 meters...80v and 10A...i think i will get some other cable for them

This was really awkward btw: the sun was behind some clouds and i got 80V, the sun was coming full again and the V dropt to 77V.
Oke temperature wise, it was warmer in the sun than 28C, out the sun(clouds) with a cool breeze 23C
So there will be more power when its cooler. Funny, i just experienced it in real life.

But the first string( just a 400 watts,(2,9A with 42ish volt each panel)) is inside the house, ready for a solar charger.

Yes indeed i am going to buy some cheaper solar charger, for 100V max and max 1160W.
They are cheaper and easier to connect, and way simpler to add more different solar panels in the future.
I think for the minimum voltige start up it would be the same with a 250V system?
Regarding losses?
Sorry i dont really know how to question this, the difference between more expensive (high voltage)solar chargers and the cheap ones (for low system voltage.)

And the are really easy to expend the system.

Thanks for reading i hope you understand this jibberchat.

No since in taking the savings of the smaller controllers and spending it on new wire.

I bought two times 250meters of each, and i just shortend the existing wires and skip a set, aka from 4 to 3 sets of panels.
From 120V to 80V from 3A to 6A per set

Voltage drop depends on current not voltage, so the voltage drop on the cables would be the same if you had two panels in series ie 2s, or 3s or other, as long as it's no more in parallel.

Each panel is a ~40v and 2.9A, I connected them as ~80v and ~5.5A so 2p2s?

With best regards Igor

 

[Redpacket]

So your cables are about 13m long. One is 4mm, one is 6mm.
Using 2s2p is good, approx 80V & 5.5A will work nicely with your cables and not have too much loss.

For the 2nd roof with 22m it will be better to have both cables = 6mm (or put 2x 4mm together)

In my system, I have my solar panel array (24 x 325W panels) wired as multiple 2s1p & see approx 7A best current.
My cables are all 6mm and the run lengths are between 11m & 26m, = approx average 2% loss & I'm happy with that.
I didn't want any combiners, etc on my roof.

For MPPT, yes, use 100V ones with the 2s2p panel arrangement. It will be close to 100V Voc on very cold days but should be OK (approx 47V each panel).
Re startup, if your batteries are approx 50V, for Victron MPPTs the panel voltage just needs to be 5V higher than the battery & it will startup.
So again the 2s panels will work well.