# Which array configuration for long cable to MPPT?

vtmaps
Solar Expert Posts:

**3,738**✭✭✭✭
Currently I have four samsung 235 watt panels, 2 parallel strings of two panels in series.

The panels have Vmp = 30 volts and Imp = 7.83. Controller is Outback FM60. Battery voltage is 24 volt nominal.

My cable from combiner to charge controller is about 120 ft.

I would like to upgrade my array from four panels to six (from 940 watts to 1410 watts).

I can configure the six panels as Three Strings of two (Vmp = 60 volts, Imp = 23.5 amps) or Two Strings of three (Vmp = 90 volts, Imp = 15.7 amps).

I am OK with cold weather Voc in either configuration.

My combiner box is a Midnite MNPV3 with two breakers. If I configure the 6 panels as 3 strings I will add a third breaker.

I hope I don't need to upgrade the cable.

Here's the dilemma:

The charge controller is more efficient converting 60 volts to 24 volts than converting 90 volts to 24 volts, but the power loss in the combiner-controller cable will be less with 90 volts than with 60 volts.

Searching through the forum I have seen this issue discussed in generalities, but without specific numbers. Here come some numbers!

I have calculated the cable loss for different voltages and current (given that I know the round trip cable resistance is 0.048372 ohms).

The controller losses are calculated from the efficiency of the controller. The efficiency of the controller is determined by interpolation of the Outback efficiency curves (which I have attached to this post. If the graph appear as a thumbnail, click on it).

Attachment not found.

.

.

My question is: Considering the numbers, Which configuration should I use?

My instinct is to go with the lower voltage and produce less heat in the controller. If I do go with the lower voltage, should I upgrade the cable to a lower resistance?

tia, --vtMaps

The panels have Vmp = 30 volts and Imp = 7.83. Controller is Outback FM60. Battery voltage is 24 volt nominal.

My cable from combiner to charge controller is about 120 ft.

I would like to upgrade my array from four panels to six (from 940 watts to 1410 watts).

I can configure the six panels as Three Strings of two (Vmp = 60 volts, Imp = 23.5 amps) or Two Strings of three (Vmp = 90 volts, Imp = 15.7 amps).

I am OK with cold weather Voc in either configuration.

My combiner box is a Midnite MNPV3 with two breakers. If I configure the 6 panels as 3 strings I will add a third breaker.

I hope I don't need to upgrade the cable.

Here's the dilemma:

The charge controller is more efficient converting 60 volts to 24 volts than converting 90 volts to 24 volts, but the power loss in the combiner-controller cable will be less with 90 volts than with 60 volts.

Searching through the forum I have seen this issue discussed in generalities, but without specific numbers. Here come some numbers!

I have calculated the cable loss for different voltages and current (given that I know the round trip cable resistance is 0.048372 ohms).

The controller losses are calculated from the efficiency of the controller. The efficiency of the controller is determined by interpolation of the Outback efficiency curves (which I have attached to this post. If the graph appear as a thumbnail, click on it).

Attachment not found.

.

.

AT FULL NAMEPLATE POWER (1410 watts):AT FULL NAMEPLATE POWER (1410 watts):

Three strings (60 volts)

..... will have a voltage drop in the cable of 1.89% and a power loss in the cable of 26.7 watts.

..... will result in controller efficiency of 96.7% and power loss in the controller of 46.5 watts.

Two strings (90 volts):

..... will have a voltage drop in the cable of 0.84% and a power loss in the cable of 11.9 watts.

..... will result in controller efficiency of 95.7% and power loss in the controller of 60.6 watts.

**AT A MORE TYPICAL POWER (1250 watts):**Three strings (60 volts)

..... will have a voltage drop in the cable of 1.68% and a power loss in the cable of 21.0 watts.

..... will result in controller efficiency of 97.0% and power loss in the controller of 37.5 watts.

Two strings (90 volts)

..... will have a voltage drop in the cable of 0.75% and a power loss in the cable of 9.3 watts.

..... will result in controller efficiency of 95.8% and power loss in the controller of 52.5 watts.

**AT LOWER POWER (500 watts)**Three strings (60 volts)

..... will have a voltage drop in the cable of 0.67% and a power loss in the cable of 3.4 watts.

..... will result in controller efficiency of 97.1% and power loss in the controller of 14.5 watts.

Two strings (90 volts)

..... will have a voltage drop in the cable of 0.30% and a power loss in the cable of 1.5 watts.

..... will result in controller efficiency of 95.3% and power loss in the controller of 23.5 watts.

My question is: Considering the numbers, Which configuration should I use?

My instinct is to go with the lower voltage and produce less heat in the controller. If I do go with the lower voltage, should I upgrade the cable to a lower resistance?

tia, --vtMaps

4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i

#### Categories

- 26K All Categories
- 201 Forum & Website
- 109 Solar Forum News and Announcements
- 1.3K Solar News, Reviews, & Product Announcements
- 154 Solar Information links & sources, event announcements
- 857 Solar Product Reviews & Opinions
- 246 Solar Skeptics, Hype, & Scams Corner
- 20.7K Solar Electric Power, Wind Power & Balance of System
- 3.4K General Solar Power Topics
- 6.3K Solar Beginners Corner
- 922 PV Installers Forum - NEC, Wiring, Installation
- 1.9K Advanced Solar Electric Technical Forum
- 4.9K Off Grid Solar & Battery Systems
- 356 Caravan, Recreational Vehicle, and Marine Power Systems
- 989 Grid Tie and Grid Interactive Systems
- 605 Solar Water Pumping
- 778 Wind Power Generation
- 612 Energy Use & Conservation
- 478 Discussion Forums/Café
- 221 In the Weeds--Member's Choice
- 74 Construction
- 96 New Battery Technologies
- 87 Old Battery Tech Discussions
- 3.8K Solar News - Automatic Feed
- 3.8K Solar Energy News RSS Feed

## Comments

17,615✭✭What is the size of the wire from the combiner to the controller?

And remember that this calculation is for maximum power, which is not what you see/need most of the time. The conversion efficiency difference isn't that devastating in real life. The V-drop tends to be a bigger power-loss problem.

3,738✭✭✭✭I sort of knew you were going to ask, but I was hoping you wouldn't. (I specified the cable resistance, which is all you need). Since you did ask, you get the whole long story.

When I designed my system I specified #4 cable for the 120 ft (one way) run between combiner and controller. The installer had lots of #6 in inventory and that's what he showed up with. He offered to double up the #6 for the quoted price.

Two strands of #6 does have lower resistance than a single strand of #4. The double #6 is protected by 60 amp breaker which is safe for a single strand of #6. I didn't want to get into a discussion of this double strand for fear of derailing the thread.

--vtMaps

17,615✭✭Okay, I understand. But I will say that those parallel wires needs circuit protection on each to prevent an over-current condition should one of them go high in resistance. You can't properly protect the two with one breaker.

Otherwise let's be sure we're on the same page as I see only about 1.5% difference in conversion efficiency between the two Voltage choices.

The power going in to the controller will be the panel current times (Vmp - V-drop). Like:

58.5 * 23.5 = 1374 Watts @ 96.7% = 1328 Watts

vs.

88.6 * 15.7 = 1391 Watts @ 95.8% = 1332 Watts

Use whatever V-drop calculator you like and see what the input and output looks like that way. That's how I'd do it. Usually the higher array Voltage wins out when long wire runs are involved.

3,738✭✭✭✭If one of the parallel wires drops out (high resistance) the other wire can handle more power than the panels can produce. And either wire alone is adequately protected by the 60 amp breaker.

I read some months ago on this forum that one should never double up wires as a way to increase ampacity, but it is OK to double them up to reduce voltage drop. Provided, of course, that the thinnest strand is protected from exceeding its ampacity.

Only at the very highest power (seldom achieved) is the system (cable + controller loss) more efficient at the higher voltage. At lower powers the lower voltage seems to be more efficient. At highest power the difference (cable + controller loss) is quite small (just a watt), so the question is: is it better to dissipate the heat in the cable or in the controller?

--vtMaps

17,615✭✭I agree about the wiring that in this case it doesn't make much difference: one 6 AWG can handle 60 Amps and you're no place near that. I just had to mention the parallel fusing in case someone else was reading this and got the wrong idea.

The problem with the other is that since you're dealing with an MPPT controller the Voltage point of he array is not necessarily the Vmp of the panels. The other thing is that under less than maximum power the whole thing becomes less of an issue as the Voltage and current both will change and it will not be a linear function on either. Pretty much a guessing game.

The good part is that you could easily change between the two configurations, couldn't you? Depending on how much is involved with getting the additional panels to the combiner.

In practical terms you're looking at a percentage of power that is so small that it's insignificant compared to the differences you'll see on a day to day basis due to weather changes. I'd go with whichever you fee is easiest to do. If we start obsessing about every little percentage point of power ... we'll be running mono panes on dual-axis trackers with MidNite controllers and nitrogen-cooled wires and ... And we

stillwon't be able to change the weather!3,738✭✭✭✭very true. I don't have logging software, but when I go to the power shed and look at the controller, the input voltage is usually in the mid to upper 50's.

Well, for me its a bit of a guessing game, but I'm still a bit of a newbie. I was hoping that for you it would be less of a guessing game :-)

--vtMaps

17,615✭✭I wish it were something more precise, but there are so many variables involved.

From what I can see of your system the wire size will mean a minimal Voltage drop either way. And there's a very small difference in the controller power loss at either Voltage level.

If you look at the figures from your own calculations (at high power) you are trading 1.89% wire loss for 0.84% loss (a difference of 1.05%) against a controller efficiency difference of 96.7% vs. 95.7% (a difference of 1%). This is almost dead even.

In the lowest power example you've got a 0.37% difference in wiring vs. a 1.8% difference in controller efficiency. Arguably the lower Voltage scheme is more efficient for the lower power designation. However, that isn't really the point where you need maximum efficiency, is it? You're looking at operating around 1/3 the total capacity of the array; there's plenty to spare at that point.

In practical terms you will not see these minute efficiency differences. A bit of extra resistance in a wiring connection will be more noticeable. Apply the percentage difference to the whole array and you have a worst-case of about 21 Watts less output between the two configurations. Dirt on the panels can have more of an effect.

5,183✭✭✭✭Vt, any thoughts of adding more PV sooner/later? If so, I would go with bigger wire now. Reconfigure later.

KID #51B 4s 140W to 24V 900Ah C&D AGM

CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM

Cotek ST1500W 24V Inverter,OmniCharge 3024,

2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,

Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep

West Chilcotin, BC, Canada

10,300✭✭✭✭i agree with coot that it is 6 of one and a half dozen of the other. this is an expansion of your present system and i contend that it won't be the last and you may want to factor in how another expansion might go to influence how this one is to be set up. do you want to go in multiples of 3 per string or 2?. in any case a further expansion would involve more wiring, but how will the capacity go for the controller. will you need 2 or 1 controller with another expansion? lay out future possibilities for another expansion to see how it may work out with either layout of 2 in each string or 3.

3,738✭✭✭✭Westbranch, Niel,

If I expand beyond 6 panels that will be a pretty big deal... I would need to upgrade my controller if I stay at 24 volts or increase my system volts to 48. I would need more battery.

My motivation for adding more panels is to reduce my generator usage. One way to reduce usage is to buy more battery capacity so I have more days of autonomy. But batteries are so expensive and panels are so cheap now. Rather than trying to

storeenergy for an overcast day, I would ratherproduceenergy on an overcast day.There have been some recent threads about 'oversized' arrays with 'undersized' batteries. Many controllers (including mine) allow you to set a limit on the battery charge current. If I had enough panels to push 60 amps into my batteries, I would set the limit to 35 or 40 amps. If I did have that many panels I could probably get a full charge on some pretty cloudy days.

--vtMaps

10,300✭✭✭✭you are quite right that the 6 pvs will pretty much max you out by stc unless you wish to for real put the max to the cc by adding 2 more in parallel. 8 x 235w = 1880w. 1880w x .77 = 1448w. this would be pushing things, but might be doable. you will lose a tad of power in the wire resistance too putting it right on the borderline. it's not a good idea to do this with strings of 3 was my point as that would be an additional pv (3x3) at 235w over the 8 (4x2) pvs that probably would be wasted most of the time.

not thinking along these lines pretty much makes it a close wash except i think i would prefer the wires heating a tad more rather than the cc.

5,183✭✭✭✭I concur, no good comes from trying to run anything 'flat out' before things start to break...

but it does look like you are at a real juncture re your total system design and size..

KID #51B 4s 140W to 24V 900Ah C&D AGM

CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM

Cotek ST1500W 24V Inverter,OmniCharge 3024,

2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,

Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep

West Chilcotin, BC, Canada

407✭✭✭✭✭✭I just went through this with an off-grid system. The client had a limited budget for now but will be able to invest more later this summer.

It is a 24 volt system.

For now it has 4 L16 batteries. later it will have 4 more.

For now it has 4, 240 watt PV modules in 2 strings of 2 into an 80 amp CC. Later we will add 5 more and have 3 strings of 3. (a bit of a stretch at 2160w but still in the mfgr's specs. Outback)

The neat thing is that 2 strings of 2 has exactly the same line loss as 3 strings of 3 with more than 2 x the power.

-Alex Aragon

8,972✭✭✭✭✭That sounds really odd. Wire resistance stays the same, if you push more amps through it, there is more loss. If you are only increasing voltage (2 strings of

) that would fly.n|| Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||

|| VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

solar: http://tinyurl.com/LMR-Solar

gen: http://tinyurl.com/LMR-Lister ,

407✭✭✭✭✭✭The % line loss;

-increases proportionally to the increase in current.

-decreases proportionally to the increase in voltage.

Check it out with a VD calculator.

240 STC, 30.6 vpm, 7.84 Imp

2 strings of 2 = 61.2 vmp, 15.68 Imp .9 ratio

3 strings of 3 = 91.8 vmp, 23.52 Imp .9 ratio

The increase in voltage is in proportion to the increase in current.

91.8v / 61.2v = 1.5

23.52a / 15.68a = 1.5

This is may be the only situation I've had where it made sense to add 5 modules to an existing system.;)

-Alex Aragon

3,123✭✭✭✭OK, if you count the loss as the percentage loss, you are exactly right. But there would also be ways to reduce the percentage loss and keep the total (watts) lost power the same.

Starting with four panels in 2 strings of 2, you could add only four and get two strings of four. The total loss would be identical (for the same length wire) and the percentage loss would be half of what was with either 4 or 9 panels. So by adding one more panel, you make the situation worse, right? It all depends on what your criterion is.

407✭✭✭✭✭✭Splitting hairs here;

For an accurate ROI calculation I suppose you would have to figure in the complete costs of the Charge Controller, the wiring and other BOS costs.

9 x 240 x 98%= 2116.8w

8 x 240 x 99%= 1908w

The only difference in cost would be the cost of the extra module and the racking for it. Being that 3 of these panels in landscape can fit on a 10' rail the difference in racking and installation cost could be quite small it would probably have a similar cost per watt for either rack. Usually the labor for adding an additional panel could be relative less due to factors like travel and set up/clean up time.

Many assumptions here:

If you figgure $1200 for the 80 amp CC, breakers, enclosure, wiring, etc (no difference between 8-9 modules)

Figgure $ 1.50/watt for modules and racking

Labor not included.

For 9 (2160w STC) the total materials cost could be $ 3240 + $1200 = $ 4440

For 8 (1920w STC) the total materials cost could be $ 2880 + $1200 = $ 4080

Cost difference ratio: 4080 : 4440 = 91.8%

Power difference ratio:1908 : 2116.8 = 90.1%

1.7% greater ROI for the 9 modules.

$360 more for 8% more charging current.

A tight budget would go for 8 while wanting to get the maximum power for $ invested would go for 9.

-Alex Aragon

BTW: Many locations, including mine, would not allow strings of 4 into a 150v max CC. (37.4 Voc x modules per string x cold temp correction factor)