diy solar

diy solar

Inverter with Limiter

If your inverter is relatively small, like 3kW, then the allowed increase to 4kW would be significant.
Space for more panels, and mounting them on the roof, is a bigger deal. For those of us who put in less efficient panels years ago, 20% efficient panels available now can produce 50% more power in the same space.

To play the kWh vs. kW game, would want additional angles because more of same angle would clip. I had aimed my panels at 2:00 PM sun to maximize production during the Noon to 6:00 PM peak rates of the time. Now that it's 4:00 to 9:00 PM, can't produce much peak. And I read that will change to even later?
What were once small bushes have grown to shade one of my (ground mounted) arrays in the afternoon. By reorienting it toward morning sun I expect to have the 50% increased kWh stay within Wpeak of the inverters.
 
If your inverter is relatively small, like 3kW, then the allowed increase to 4kW would be significant.
Space for more panels, and mounting them on the roof, is a bigger deal. For those of us who put in less efficient panels years ago, 20% efficient panels available now can produce 50% more power in the same space.

To play the kWh vs. kW game, would want additional angles because more of same angle would clip. I had aimed my panels at 2:00 PM sun to maximize production during the Noon to 6:00 PM peak rates of the time. Now that it's 4:00 to 9:00 PM, can't produce much peak. And I read that will change to even later?
What were once small bushes have grown to shade one of my (ground mounted) arrays in the afternoon. By reorienting it toward morning sun I expect to have the 50% increased kWh stay within Wpeak of the inverters.
Current Microinverter-based AC-coupled array is 4kW with 3kW max inverter power.

Covering the entire SW-facing roof so nothing more to do there.

Going to add another 2kW of DC-coupled panels on another roof face and will likely use a 60A / 1500W MPPT charge controller since my current rig rarely saturates.

When I installed the original AC-coupled system, it was planned for a 1kW to 2kW increase but now that the time has come, leaving the existing grid-tied system as-is and using the new panels for a DC-coupled system with a battery for self-consumption just makes more sense.

And it is getting worse - by 2022, peak window starts at 6pm and ends at 11pm.
Solar power you can save up and self-consume during the peak window will be worth 2X solar power you send to the grid for credit...
 
Solar power you can save up and self-consume during the peak window will be worth 2X solar power you send to the grid for credit...
Those are the economics that will drive more and more battery installations.
In addition there will be a mechanism for battery connected inverters to sell back to the grid at peak rates or perhaps even higher market rates. Unfortunately the new NEC requirements for UL certified battery storage may mean that the mechanism will be out of reach for DIYers.
 
Those are the economics that will drive more and more battery installations.
In addition there will be a mechanism for battery connected inverters to sell back to the grid at peak rates or perhaps even higher market rates. Unfortunately the new NEC requirements for UL certified battery storage may mean that the mechanism will be out of reach for DIYers.
That’s part of what scared me off of the ‘official’ path. Additional $$$s, headaches, and lack of any kind of ‘simple plan’ (as I used to do my own official grid-tied solar install) pushed it out of reach.

During an extended power outage, I can rig up my generator to power my fridges no fuss no muss.

I can also throw up some temporary panels on my deck to charge a battery powering an inverter for the same purpose.

After what my utility has done to me over the past 4 years, I want nothing to do with them (other than having the grid to power from over winter and on cloudy days).

$10 a month for that peace of mind I can live with.

I have absolutely no interest in selling power from my battery to the grid - no worth the headache and cost (at least to me).
 
fafrd, This might do what you are looking for.


This is a 1000 watt that will accept 22 to 65 volts DC and drives 1120 volt out. So you just put one on each leg of the 120/240 split phase. No Autotransformer needed. Each one will just limit it's leg to zero export. It is meant to work direct from a solar panel, but reading the reviews, there are some people running it from a battery bank. I have not used these, so I have no idea how good they might be. You may need to add some kind of timer control to keep it from running down the batteries overnight.

I finally a manual for these Sun GTIL inverters and I’m starting to get more comfortable with the 120V model.

The maximum output current of the 120V model is explicitly specified at 2X the maximum output current of the corresponding 240V model, so that goes a long way to alleviating my concern that output power is reduced by half at 120 versus 240V (though I still have my doubts about the ‘autodetect’ feature).

Stacking is complicated for non-grid-tied inverters because they need to be synced, but making a split-phase GTIL by stacking two out-of-phase 120V GTILs is straightforward because both units are already being synced to the grid.

Explicit mention is made of battery power in the manual. It seems as though the 24V model will just drive it’s MPPT voltage into a battery as though it were a solar panel. And unfortunately, the manual also makes clear that maximum output is reduced under battery power (to 700W for 1000W model driven by 24V battery).

So two of these 1000W/24VDC/120VAC, each with it’s own limit sensor tied to one phase/leg and tied to that corresponding phase/leg with a common neutral should allow me to offset up to 1.4kW of peak-period consumption without needing the two Autotransformers.

The only thing I’m slightly worried about is that the documentation shows stacking with seperate batteries (following the architecture of stacking with seperate PV arrays, which is necessary with seperate MPPTs).

Thinking through how MPPTs will function with a battery, however, I’m pretty certain it should be possible to run 2 or more of these Sun GTIL inverters off of a common battery, as long as each has it’s own high-capacity cables connecting to the common battery leads.

At this point, I’m down to deciding to purchase through Amazon for $300 each with increased chances of a refund if the product does not work off of a 24V battery as specified, or deciding to purchase through AliExpress for $180 each with shipping and no recourse if I get junk.

I don’t need to worry about a timer since I’ll have plenty of solar current to recharge every morning. In fact, the self-regulating aspect of the system is one of the attractive features to me. Here are the different ‘built-in’ phases:

Off-Peak Empty Dark: running house off grid until sun comes up.

Off-Peak Empty Bright: running house off of grid-coupled AC solar while battery recharged off of off-grid DC solar (house consumption always negative in this phase, so GTILs dormant).

Off-Peak Full Bright: MPPT charge controllers shut down off-grid DC solar production.

Peak Full Bright: as long as AC-coupled solar production exceeds house consumption, GTILs will remain dormant. If/when house consumption exceeds AC-coupled solar production, GTILs will kick-in to compensate net house consumption from battery energy (and MPPT charge controllers may reactivate DC-coupled solar charging depending on available solar energy and battery State of Charge).

Peak Full Dark: after PV production has ended, GTILs will offset up to 1.4kW of house consumption from battery energy.

Off-Peak End-of-Charge Dark: when the battery voltage drops below GTIL minimum voltage requirement of 22V, the GTILs will shut down and house consumption will revert to being supplied by the grid (so back in Off-Peak Empty Dark mode). This will be true whether End of Charge occurs during peak window or during early morning hours after that. In the case consumption was so low that the battery never drains before the sun comes up, nothing really changes except Off Peak Full Bright is reached more quickly.

I like how it’s self-regulating, redundant (parallel to grid), and doesn’t need lots of little additional components like timers or switches to make it work - it will just naturally start using battery energy to compensate for net consumption once net house consumption switches from negative to positive as the sun goes down...
 
Completed a bit more research on this and here are some more details:

-the older models were either 120VAC or 240VAC while the new 2nd-generation model appears to only be the Autodetect model. I’m skeptical that output power @ 120V will match output power @ 240V (Current will double) but it is easy enough to verify and the Gen1 120VAC model clearly put out twice the maximum current of the Gen1 240VAC model, so I think I’m ready to give it a try.

On AliExpress, you can find clear battery-driven specifications indicating that the battery voltage must be at least 26V for the 24VDC model to’start up’ but then that once started, it will run as long as the battery voltage exceeds 22V,

On Amazon, all the vendors state that the 24V model needs a 48V battery to operate off of battery power (more conservative with specifications to avoid battery-related returns).

So you will not be able to return an Amazon purchase because you cannot get it to work with a 24V battery.

The Y&H model is $275 on Amazon while the least-expensive no-name vendor I can find on AliExpress for the same Gen2 product comes to $200+$75 shipping, so purchasing a single Y&H unit from Amazon seems like the prudent thing to do.

If it works with my 24V LiFePO4 battery as I suspect, I am good to go and can get a second.

If if only works with a 36V or 48V battery but otherwise performs as expected in all respects, I can’t return it without covering shipping cost, but at least know what I am dealing with and can consider upgrading my battery to 36V or 48V.

And if it doesn’t perform to specification even when powered with a 48V battery, Amazon will make sure I get my money back.

Also appears that the Gen2 model is spec’d to deliver 900W continuous, even when running off of a 48V battery, so if it functions with a 24V battery but delivers less continuous power of as little as 700W continuous, at least I’ll know what I’m dealing with and I’ll be pretty happy (no worse than Gen1 @ 24V battery power).
 
Just got off of the phone with Amazon and confirmed that since the Y&H vendor ships from US stock and qualifies for Prime, I have 60 days to test the product during which time I can return it at no cost to me for any reason.

This was too good of an offer to pass up and I just placed an order for 2: https://www.amazon.com/gp/aw/d/B07GC53QBD?ref=ppx_pt2_mob_b_prod_image

They get here just in time for Christma ;).

I will report back on my experience with these products but would appreciate guidance as to whether I should start a new thread or just keep posting here.
 
Just got off of the phone with Amazon and confirmed that since the Y&H vendor ships from US stock and qualifies for Prime, I have 60 days to test the product during which time I can return it at no cost to me for any reason.

This was too good of an offer to pass up and I just placed an order for 2: https://www.amazon.com/gp/aw/d/B07GC53QBD?ref=ppx_pt2_mob_b_prod_image

They get here just in time for Christma ;).

I will report back on my experience with these products but would appreciate guidance as to whether I should start a new thread or just keep posting here.
So any updates on the system? Looking at buying some of these units and as I read through your thread you have asked many of the questions I have been asking and scouring the web for answers.

The units you linked to on Amazon appear to do split phase? What did you end up doing for the CT clamps? 2 in series? I would love to hear more about how it is going!

Thanks!
 
So any updates on the system? Looking at buying some of these units and as I read through your thread you have asked many of the questions I have been asking and scouring the web for answers.

The units you linked to on Amazon appear to do split phase? What did you end up doing for the CT clamps? 2 in series? I would love to hear more about how it is going!

Thanks!
Still a work in progress but here is a bit of an update on equipment so far:

GTILs - work like a charm so far. I have one 1000W on L1 and one 1000W on L2, both powered by the same 580Ah 24V LiFePO4 battery. Each GTIL has it’s own CT clamp on it’s appropriate main leg in from grid and it works as it should (<10W output per leg).

Only 2 complaints are that efficiency is pretty poor (~80%) and when outputting max power of ~850W, fans kick on any it’s pretty loud. But overall, I’ve had no issues running off of an 8S LiFePO4 battery (I’m using dual 2/0 cables to minimize voltage drop when pulling a full 80+ amps).

I’m essentially zeroing out consumption on each leg individually (so zero risk of pulling in from one leg and pushing out through the other).

This means if I’m pulling over 850W on one leg, I can’t zero out even though there is another 850W available, but it’s simple, it works, and my only real goal is to zero out consumption of fridges that burn under 850W each.

When the electric oven element turns on and 3000W if 240V energy are being consumed, both GTILs crank up to 850W, fans make it sound like a jet is landing in my basement, and my meter reads 1.3kW being consumed.
 
I’m essentially zeroing out consumption on each leg individually (so zero risk of pulling in from one leg and pushing out through the other).

This means if I’m pulling over 850W on one leg, I can’t zero out even though there is another 850W available, but it’s simple, it works, and my only real goal is to zero out consumption of fridges that burn under 850W each.

Don't think it would be a problem if you pushed power out one leg and pulled back in the other, for net zero power export.
Utility meter will measure power on each leg and record the sum. Current crop of meters, so far as I know, won't tattle on you.
This assumes of course the meter does correctly calculate power, not just volts x amps, on each phase. (Otherwise you would be charged not credited for export on the leg you drove past net zero.)
 
Still a work in progress but here is a bit of an update on equipment so far:

GTILs - work like a charm so far. I have one 1000W on L1 and one 1000W on L2, both powered by the same 580Ah 24V LiFePO4 battery. Each GTIL has it’s own CT clamp on it’s appropriate main leg in from grid and it works as it should (<10W output per leg).

Only 2 complaints are that efficiency is pretty poor (~80%) and when outputting max power of ~850W, fans kick on any it’s pretty loud. But overall, I’ve had no issues running off of an 8S LiFePO4 battery (I’m using dual 2/0 cables to minimize voltage drop when pulling a full 80+ amps).

I’m essentially zeroing out consumption on each leg individually (so zero risk of pulling in from one leg and pushing out through the other).

This means if I’m pulling over 850W on one leg, I can’t zero out even though there is another 850W available, but it’s simple, it works, and my only real goal is to zero out consumption of fridges that burn under 850W each.

When the electric oven element turns on and 3000W if 240V energy are being consumed, both GTILs crank up to 850W, fans make it sound like a jet is landing in my basement, and my meter reads 1.3kW being consumed.
Thanks for the update.

I am looking at the 240V ones. Not sure there is much benefit other then I can start with 1 string of panels. Maybe i Need to rethink a bit.
 
Thanks for the update.

I am looking at the 240V ones. Not sure there is much benefit other then I can start with 1 string of panels. Maybe i Need to rethink a bit.
So no battery?

For me anyway, evening / night consumption is far higher than daytime consumption (when PV array is producing), so using a 24V battery to store up solar energy generated during the day to be consumed offsetting nighttime consumption after the sun has gone down was central to the system making sense…
 
So no battery?

For me anyway, evening / night consumption is far higher than daytime consumption (when PV array is producing), so using a 24V battery to store up solar energy generated during the day to be consumed offsetting nighttime consumption after the sun has gone down was central to the system making sense…
Good point. I just spend a few minutes reviewing my Emporia data. By all accounts my min usage through peak sunny days is 1kWh. Averaging closer to 1.7kWh.

If I get one of these inverters and a few panels a battery will be on the short list. I just don't have any 48V batteries right now. I could cobble some lead acid together but its kind of wasted money. I will order some cells before to long and build a battery regardless!
 
Back
Top