diy solar

diy solar

Feedback on Off Grid Cabin setup

Lookaflyingdonkey

New Member
Joined
Mar 16, 2021
Messages
29
Hey Everyone,

Long time lurker and the info here has been invaluable to planning our new system. Hoping I could get some feedback on component selection.

A couple thoughts:

1. I wanted to build something we can expand over time, 1 or 2 new batteries and expanding the solar array when needed
2. The battery is the only think I have ordered so far
3. Not 100% sure on the distance between solar array and inverter/scc, currently estimated at 200', this is why I went with the 600v SCC vs the 150v as I calculated out #0AWG for the 2s3p setup, which would have been 3x the price for wiring and not expandable
4. I also looked at some of the all-in-one units, but couldn't find anything that really suited the long distance and the future growth/reconfiguration.
5. I picked the Bifacial panels with the hope that when it snows in winter and we may not get up to the cabin immediately, the reflected light below will at least get some charge into the system until the top melts off, not sure if I am dreaming though. Apart from that they are nice panels if not a bit large.

Not 100% on the breaker sizing either, I have selected one of the Midnite Solar EPanels for the Conext 4048, not sure they are 100% required, but sure make it neater I think.

This is all for an off grid cabin in upstate NY, constant loads would be ~1.2kw/day (fridge, internet, camera, etc) and possibly dropping 1kw in off season if we dont have fridge running.

Diagram.png

Thanks!
 
At our cabin I had about 200' between the ground mount array (2S3P, 280W panels) and based on my math (aiming for <3% loss at max power) I used 6AWG wire. Not sure why you would have to use 1/0 AWG.

Anyway, it sounds like your system is sized similar to mine. I went with a 24V system, using the SW4024 and the MPPT60-150.

Edit: I just noticed your panels are a fair bit higher power than mine, but they are both higher voltage and higher current, so I still don't think you would need 1/0 AWG wire.
 
Hi Horsefly,

I was using 49.1 voc and 11.38A Isc per panel, so ~100v @ 12A per string, then 100v @ 36A for the 2s3p setup.

Running 100v @ 36A through the renogy calculator (https://www.renogy.com/calculators#tab_solar-cable) with 2% loss for 200' came out at #0.

Am I not supposed to add the current together for parallel? I also used the open voltage, not the max

Cheers,
Dean
 
You don't want to use the open circuit voltage or the short circuit current for figuring out wire sizes. The max loss in the wire will be at max power, so you should use the max power voltage and current. I didn't use any on-line calculator. I used the wire resistance tables and calculated it myself.
 
Ah k, that'll do it. So if I recalculate at 40.8v @ 10.66A per panel, so 82V @ 10.66A per string, 82V @ 33A for the 2s3p it is #0 for 2% loss and #2 for 3% loss.

Either way #2 or #0 are significantly more expensive right? So I could spend the difference in wire cost and go up to the 80A 600v SCC, not sure if it is a waste though, the price difference is around $850, it does give more headroom though.

Also good to hear you have a similar system and it is working well
 
Ok, so I'm back, but after dinner and 3 glasses of wine. So, excuse any mistakes due to the wine...

Here's my take, assuming your numbers: You've got 82V on the line, and you want no more than 3% total loss on the wire of 400ft (200ft each way). 3% of 82V is about 2.5V, so you want no more than 2.5V voltage drop. The voltage drip is the current times the resistance, so 2.5V = 33A * R, so R = 0.076 ohms. At 400 ft, that's 0.076/400 = 0.00019 ohms/ft. Resistance of stranded wire is usually expressed in ohms per 1000 ft (or per km), so this would be 1000 x 0.00019 = 0.19 ohms / 1000 ft is the maximum resistance wire you would want for no more than 3% drop. If I look at the NEC table 8 "conductor properties" I see that 2AWG wire has 0.194 ohms/1kft. That's pretty close, so I guess you are right. 2AWG would be the best fit.

You didn't mention where your cabin is. Does it get cold? If the temp doesn't get cold, you could possibly go with 3S2P, which would increase the voltage and decrease the current, and make the wire cheaper.
 
Thanks for the run down, good to know the math behind the magic calculator!

The property is upstate NY so gets down pretty low, I think it would push it above the 150v limit of the smaller SCC.

Knowing that would you think it is better to spend the extra money on bigger wire or a bigger SCC?

Back of the envelope given $0.80/ft on the #2 and $1.98/ft on the #6, with 600ft (3 cores), the savings would be around $700, nearly the difference in controllers. With the advantage of having headroom on the #6 for expansion.

Another option would be aluminum wire, but that requires even more upsizing right?

Cheers!
 
Yeah, best not mess with the hyper-Voc at cold temps.

This may sound weird, but I have an off the wall suggestion for your consideration. The price of wire (copper) has been steady or going up for years. Meanwhile, the price of solar panels has been dropping like a rock. So if you are looking at the economical way to get the power you need, one could forget about the 3% or 2% or 4% loss on the wire, and just add lots of panels. You could end up with 3s4p and have losses way over 4%, but it would still give you all the power from the solar array you would need. Just a thought.

One other consideration: The "nameplate rating" of your panels (435W) is a rating at STC (standard test conditions). These conditions are 25°C and 1000 watts / square meter of irradiation. That is an ideal case that almost is never met in the real world. For that reason, the standard recommendation is to derate the panel output to 75%-80% of the nameplate rating. Doing this cuts both ways: It would decrease the losses on the wire (it would mostly be the current that would go down, not the voltage), but it would also decrease the power available to you. In my case our cabin is in the clear dry air at 9000 ft elevation in Colorado. I've witnessed the panels delivery pretty close to the nameplate rating, but that is unlikely in upstate NY.

If the panels could produce nearly the nameplate rating, you are already getting close to the current rating of the smaller charge controller: 60A. That is, 435W / 48V = 54A. Not much room to grow there, so that may argue for the bigger SCC.

Edit: I didn't answer about the bigger wire consideration, or specifically the multiple lines to simulate a bigger wire. I guess that is worth considering, but remember if you run more wires you will be dealing with a bigger conduit for 200 ft. Not a huge deal, but it can be a pain.
 
I’m thinking the higher voltage SCC may be the way to go, with #6 there is plenty of room to expand (higher number in series and additional strings in parallel), with what seems like similar upfront cost given the price of copper.

The panels are undersized for the battery capacity too so may need to add more panels sooner rather than later if we don’t get enough output up that way.

The one downside to the higher voltage DC is the additional costs on breakers/fuses/etc. but not really material.
 
Also wondering @Horsefly what you used for the switchgear? Did you use the official Conext AC/DC enclosures? The Midnite Solar E-Panel? Or just built it out yourself?
The official switchgear looks nice and straight forward, but pretty pricey. Midnite one is good too as it’s all in one, but not sure if mixing the ac/dc wiring in the single enclosure is the best idea.
 
Ah nice! Good to see someone who has one in use and it was worth the extra $$

Looks like a very neat install, I see the lightning arresters there too, will have to add those to the shopping list.

Are those current meters (blue screens between the SCC and Inverter)?

Also what is that coming out the top?

Cheers,
Dean
 
The two meters are a couple of the cheap made-in-China Amps/Volts/Watts/Wh displays you can find on Amazon or EBay. Both are connected to shunts, with one measuring the power from the panels to the CC and one from the CC to the rest of the system. If I was doing it again now I would have not bothered with that, and instead used a Victron Smart Shunt monitoring the energy into / out of the battery.

Coming out of the top is 24VDC, and the little black box is a 24VDC-to-12VDC buck converter. We have a number of 12VDC things running in the cabin, including nightlights, stereo, and phone chargers. We can operate these without taking the inverter out of standby.
 
Was there a reason for not using the monitoring in the Conext system to get those values? Or just looking for a second opinion on them? Conext software should be able to get power in/out of all the components right?

Good idea with the DC circuit too, may be worth trying to run some of the gear direct to avoid the inverter overheads
 
When you say "Conext software" I'm not clear what exactly you mean. I have the SCP (shown in the lower left corner of the E-panel), not the Gateway. But yes, you can maneuver through the menus and get how much power is coming into CC, and then maneuver over to another menu and see what is coming out of CC. But you can't really see both at the same time. I can also maneuver through the one-line CC display and see a 30-day history of Ah harvested, which is at least interesting.

Like I said, I don't think it really was worth it anyway. I came to realize as I was designing the system that the center of an off-grid Solar power system is the battery. Everything begins and ends there, and nothing works without it. So what makes more sense is seeing what is going in and what is going out of the battery. For my AGM batteries (until I can replace them with LFP), there is no specific gravity to check, so it's one of the few ways to see how the batteries are doing. So even if I don't replace the AGM batteries this year, I'll probably put in the smart shunt.

One other thing about those little meters: They are limited to 100VDC. Vmp at STC for my 3S2P configuration is higher than that, but I thought if I derated the panels to real world (75%-80%) I would stay under that. Fact is, solar panels are effectively a current source. The voltage under sunny conditions stays up close to Vmp, but the current will probably be lower than the Imp at STC. So from the get-go the meter from the array was not of much value, as it just flashed "99.9V". :rolleyes:
 
You said something about the SPD's (surge protection devices) and I forgot to point out: I actually have 3. The two shown in the photo are on the input of the inverter from the generator, and on the output of the inverter to the breaker panel. The third one is out in the combiner box at the ground mount for panels. I'm convinced you need all three, unless you don't have a generator, in which case I guess you only need two.
 
I think you do need the combox to get the monitoring, it looks like the attached image, can see flow direction and energy going between all the components, also does datalogging so I assume you can review historical.
1616446651395.png


I suppose I should be able to monitor in/out from battery in the BMS too, as I am building a DIY pack with one of the Overkill Solar "Smart" BMS's.

At least the 99.9v works as an alarm ?, if it is anything less than that you know something is wrong.

So you have the third lightning arrester out on the array, is there any issues with exposure to the elements? I have seen some people put all three in the midnight enclosure, or attached to the MPPT Charger, but not sure if one or the other is better
 
I think it's best practice to put the SPD out at the array. Otherwise a lightning strike could be traveling down your 200' of buried conduit. Even if it does get stopped at your CC, you could have wires fuse in some unknown spot along that 200'. A real pain to find and fix.

I used the exact same Midnite Solar MNPV3 combiner box, and I think there is a punch-out on the side that I used for the SPD (or maybe I drilled a hole - I don't remember):
SAM_3012.jpg

Also, I used breakers instead of fuses, as you can see. I almost always use breakers instead of fuses.
 
Makes sense you want to protect the underground wiring too, that would be an expensive replacement!

I think that combiner has a. knockout on both sides, so makes sense that's there the arrester would go.

I was going to do breakers, but they only have 150 and 300vdc versions and say you need to use fuses for 600v or 1000v, I am sure I could find another brand that has 600vdc breakers, but not sure on warranty with midnite if you aren't using their components.
 
Back
Top