Michael Qulek

My suggestion would be to go with a 24, or 48V AC cabin system.  The single cheapest components right now are solar panels.  Don't buy little 12V panels you see online.  You get far more bang for your buck going with large residential panels in the 250-350W range.  With a good MPPT controller, a series string of several large panels will be transformed down to battery charging voltage, making extra charging amps out of the extra volts.

Some may recommend a 12V RV style system, but I'd say that's a mistake.  12V is just too limiting.  As you scale up to higher system voltages, you gain a LOT of capacity.  My 48V cabin system can run my 240VAC well pump.  It's a 1hp Grunfos.

Where you need to start is with an itemized list of everything you want to power.  Start with the dehumidifier, but also include lights, TV, the computer, a refrigerator, an air-conditioner.  Add up the Watts per day you think you might need, and then plan from there.

Just keeping the lights and TV on, with a refrigerator running in the background, I'd suggest you'd need about 3kWh of power per day.  Remember that the inverter being on consumes power even if nothing is turned on.  Some of the cheaper AllinOne inverters (AiO) can consume more than 100W per hour, which works out to be >2.4kWh of power per day, just leaving the inverter on.

Here is a quick list of things to make a whole-cabin system that will power what I'm mentioning.

Four 6V golf-cart batteries, wired in series for 24V  CostCo has a 210Ah battery for 110$ now.
Four 250W 30V residential panels.  About 50$ each on Craigslist right now, more or less depending on your location.  Don't buy online panels.  Buy locally with cash and carry purchase.
40A MPPT charge controller.  Epever makes a budget model called the Tracer 4210AN for 125$.  For a bit more money, you could get the 5415AN, or Triron4215N, which have significantly higher voltage limits.
Sine-Wave inverter.  Don't be tempted by cheap MSW models.  They will shortly burn out anything running an electric motor.  The Samlex PST2000 is UL listed.  Cheaper inverters are not.  I have a Conext SW4024.  It can make both 120V and 240VAC.

The system I'm describing might be a bit large for only a dehumidifier, but is far more scaleable than a simple 12V system.  Being a component system, it can grow by adding/replacing individual parts one by one.

I use it first for cleaning tools, and/or motor parts, collecting it an open basin.  I pour that back into a gallon jug, and use it again for lighting burn piles during spring fire abatement.

Though I haven't done it myself yet, what I've read about cast iron is that it should be heated almost red hot before stick welding, and then after the weld is completed, to tap the welded joint with a hammer to stress-relieve it as it cools to room temperature.  If you give it a try, let us know how well it turned out?

OK, those numbers appear better.  I think you are good to go.

I have plugged the specs for the panels I plan to use into this and it came out with a max Voc of 120.1 @ -30C. I doubt I'll ever get that cold so I think I'll be safe with a 150V charge controller. Thanks for the link though, it was helpful.


Can you please post the actual specifications of those panels?  I am a bit suspicious, and I think there might be some mistakes here?  With three 72-cell panels in series, I routinely see 120-125VDC on regular summer days, not the winter.  To me that 120V reading at a frigid -30 doesn't seem right?  I want to make sure you are not getting advice that could result in a fire at your place.

I'd suggest that only a 5% overvoltage estimate is too low.  Midnight Solar recommends 12% for 0C (freezing).  To be really safe, I'd suggest you plug your panel specifications into Midnight's string calculator.  That will give you an accurate Voc at whatever your winter low is.

http://www.midnitesolar.com/sizingTool/index.php

For my own arrays, with three panels in series, I went with Midnight's 200V controller.  That gives me a very broad safety margin.  If you haven't gotten your charge controller yet, I'd really recommend you get a higher voltage model, and also higher amperage.  If you plan on going with 48V, which I think is a good idea, you'll most likely want to upgrade your system in the future.  Limiting yourself to only 35A is going to hobble yourself.  In terms of those home-made terminals, I was daily passing 60-65A through them with no issues.

1275W of solar is quite small for a 48V system.  I really think you should plan for upgrading your system in the future.  For my own cabin, I've got a 48V system that can make American split-phase 120/240VAC, which I use to power my well-pump.  That's a monster that needs ~9200W at startup, and 2000W to run.  I understand you don't need that much power, but who's to say what you will want in the future.  Make sure you have a clear path to future upgrades.

One thing you can do now with that idea in mind is going a gauge or two thicker in copper wire than what you need today.  Besides the extra safety margin, you'll save money in the long run because you won't have to rip out thinner wire when you decide you want to upgrade.  Believe me, I know.  I've upgraded my cabin system three times since I installed it in 2017.  I've added an additional 1000W with each upgrade whenever I got good deals on Craigslist panels.  I've got 4 gauge between the charge controller and batteries, and 0000 wire between the batteries and the inverter.

I'd encourage you to get a clamp meter that can read inrush.  I started out with an expensive Fluke meter, but found a much cheaper alternative that I use routinely now, a UniT-216C  I can read both AC and DC amps, AC inrush, and regular volts and ohms.  In the US, it's about 75$ on Ebay right now.

One last thing.  Don't believe the salesman hype about the deep discharges of Li batteries.  Although they do have a deeper discharge curve than lead-acid, very deep discharges shorten their lives almost as much as lead acid.  If you want to have long battery life, plan on never discharging them more than 50-75% on a daily basis.  You would be wise to double your battery capacity right now in anticipation of lower discharges.

Rather than focusing attention on crimping right now, I'd recommend you focus more attention on overall system design, and worry about crimping later.

The grinder or sander will be your biggest loads, and I focus on supplying those first.  Other stuff is trivial compared to the draws those two will need.  If you focus your attention on those, the rest will be easy.

A good rule of thumb I follow is to have about 2X your largest load as for how many solar Watts you need.  So with a 800W load, maybe 1600W of solar panels.  The good news is that high-voltage residential panels are getting dirt-cheap right now, and you can find 30V/250W panels for 40-50$ right now.  So, six of those would be close to perfect in terms of meeting your wattage needs.  That's about 300$ right now in the cash and carry market.  Don't order panels online!  You will pay outrageous shipping fees.  Shop locally on Craigslist, with cash and carry purchases.

The amperage to handle 1600W is a bit too high for a basic 12V system, so I'd recommend going with at least 24V.  That's what I have in my workshop right now.  In terms of a budget charge controller, take a look at Epver's new XTRA Series MPPT Solar Charge Controller.  40A of current at up to 150V, for only 120$.  40A would be good for more than 1000W of panels.  In general, it's OK to exceed the recommended amps of a unit.  Under no circumstances though should you ever exceed the voltage.  Remember though that the voltage of panels goes up as the temperature goes down.  A good safety margin is ~20-25%.  So, for a 150V controller, never surpass ~120Voc.  That should be good for three 30V panels in series.  Their Voc at freezing is going to be ~40Voc.

I have a clamp meter with "inrush current" capability.  What I've found is that the brief starting surge for motors that start under load is ~4X.  So, for that grinder or sander that needs 800W to run, expect them to need up to 3200W for less than 1 second to start.  So, I'd suggest shopping for at least a 3000W 24V inverter.  Get Sine-Wave for anything that will be running an electric motor.  Cheaper modified sine-wave models are well-known for burning out electric motors in short order.

Lastly, you'll need batteries.  You might want to take a look at CostCo's 210Ah 6V golf-cart battery.  At 110$ right now, you'd need four of them in series to make a 24V battery bank.  That would be 440$ plus tax and core charges.  Assuming you don't drain these lead-acid batteries more than 50%, you'd have 2.5kWh of available power.

If you built a rotating array frame like I have, you could reduce your panels needed to down to 1000W.  My frame easily holds four large residential panels in portrait, but can hold as many as six oriented in landscape.  I just sunk a 3.5" schedule 40 pipe in concrete, and slipped a 4" schedule 40 pipe on top of it.  The array frame is welded to the top of the 4" pipe.  This will allow you to rotate (by hand) East in the morning, South around noon, and West in the afternoon.  That will basically double the number of watthours you can make in a day.

BTW, I've made excellent terminals for my solar wire with cut lengths of copper tubing.  For my 000 battery wire, I belled 1/2" copper tubing, slid the wire halfway inside, then flattened the other end.  Once flattened, I soldered the wire to the tubing.  I then drilled 5/16" inch holes in the flattened ends to accommodate that stainless steel mounting bolts I used to attach the batteries to the inverter.

At this point, I would not worry too much about the shading.  With your sized battery bank, you'd have nowhere to put all the watts.  With your house drawing, what's the direction of the drawing?  I'll guess that down is South?    As discussed above, you might want to consider six panel strings of 30V residential panels, 180V in series, making 1500W per string.  Right now, it looks like you could fit a 6-panel string at the lowest point of the roof, and then a second 6-panel string right up at the ridgetop.  Assuming you can get 3.0 sunhours per day in December, that's 9000Wh of power, about double what the battery could accept even at zero charge.

If you want to add another 6-panel string towards the middle section, you could position the panels in two rows of three panels, placed on the far-right side of the roof.  That would at least allow you for some extra power at noon?

Lastly, you can utilize virtual tracking to add even more arrays.  Maybe another 6 panels facing due West, or 6 facing due East, depending on the orientation.  The advantage of virtual tracking is that you can get additional power in the early morning, or the late afternoon when the main South-facing array is not putting out much.   With your roof, I'm guessing that the roof to the far left of the pics is facing West?

By breaking up the panel strings in to shorter 6-panel units, you decrease the very negative effects of shading to maybe only one string at a time.

With a cross-section of 25mm2, it looks like it lays somewhere between 3 gauge, and 4 gauge, following American guidelines.  In the real-world, that looks like it is likely to handle 80A without getting too hot.  That works out to be about 4000W with a battery feeding the inverter at around 50V.  Most likely you are likely to get by as long as you never attempt to run the inverter flat out, at maximum wattage.  But, it's your call as how hot you are willing to push your wiring.

I can understand though that you are in a country at war right now, and you really have to make do with what you can find at the moment.  What I can suggest you do now is just attempt to run the unit supporting loads like the refrigerator, and lights, but not turn on any big-ticket items like an electric heater, or a microwave.  You can put a hand on the insulated part of the wire while running your loads to see just how hot it's actually getting.

You can control that a bit better by fusing it with a lower rating fuse, like 80A, even if you install it in front of a 150A breaker.  Some time in the future, when it becomes practical to upgrade your wiring, the 80A fuse can be dispensed with.

Have you started working on your solar input yet?  You haven't mentioned one word about panels yet?