diy solar

diy solar

Adding Schneider XW Pro

I wonder if having a DC line surge suppressor would have protected the Tigo TS4 box?
 
The warranty replacement for the failed Tigo module arrived today. I'm not sure I want to try another RSD test. I know the firefighters would be safe if they hit the switch, that might be enough. The hardware may not survive, but the firefighters would.

Coincidentally, I had the city inspection today!
Quickest inspection ever. He looked at the 15 amp breaker, looked at the panels from the ground and signed off.

I put the power monitor back together cleaned up some.

PXL_20221028_221756844.jpg
 
Congratulations on the inspection !

Pardon my ignorance but I understand DC wires in the house need to be in metal conduit. Is all that flex you have considered metallic or did your AHJ not care ?
 
The two parallel flex down below are AC (they sort of form a U shape between the pull box and distribution panel.

The only flex with DC is the short vertical run from the charge controller to the PDP. That's 48 volt in there, not higher voltage PV.

I don't know the exact number, but it's probably 60 or 80 volts that is required to be in metallic conduit. I'm not even sure code specifically requires metallic conduit. I seem to have forgotten...


But in short, I doubt they even noticed the flex. He looked for the breaker in the Midnite combiner box, didn't care about the battery side breaker or much of anything else.
 
you all scared me there a moment

I started with a roof 24v array in 2000, then in 2018 upgraded to 48v arrays - all DC 6 awg wires in gray plastic pipe.
HIgher voltage arrays were very rare in 2000.

and congrats on the passing grade
 
This is getting frustrating. When I first added the Schneider inverter and battery, I called PGE to start the interconnect process/update my NEM 2.0 agreement. We talked though the components, they told me the application would be denied as they would show no way to charge the battery. I don't remember the exact reasoning on their end, but something about the previous NEM and AC coupling. In short, I needed to add some panels DC coupled to the solar. I should have argued more, but I wanted to add more panels anyways.

Well, now something like 18 months later. I've added the panels and I'm working though the PGE process. But, now it's entirely online. I've been asked to edit and resubmit the SLD 3 times. Each time editing and adding basically the same stuff. Just more details are added each time.

Ugh, I'm going to perform the 3rd edit this evening. Maybe a call tomorrow will get some assistance, but I'm not hopeful.
 
I've been working on my system logic and modbus commands to integrate actual PV output into figuring out when to send the start charge command in the morning after the sun comes up.

Then the next day after I set this up, the clouds rolled in thick enough to drop PV production down so low it couldn't even cove bas loads.

I looked at node-red and realized it would be pretty easy to stop the charge and let the battery cover loads on these days

@GXMnow
What modbus address/command are you using to switch from charging back to discharging during the day? Do you have this functionality? I feel like you did. When the loads in your backup panel go above PV.
 
Last edited:
What modbus address/command are you using to switch from charging back to discharging during the day? Do you have this functionality? I feel like you did. When the loads in your backup panel go above PV.
When I decide charge should stop, I just send the Charge "No Float" command Send the number 3 to location &H165
As long as the battery voltage is above the "Grid Support Volts" it will just start supplying the loads in the backup panel. If you also set "Grid Sell" it will do that as well, but I have grid sell blocked before about 3 or 4 pm, if I remember correctly. I didn't want to just dump the battery into running the A/C before we get to the high rate time. I may have to work on making it also not run the backup panel until 4 pm as well. Save what little we can get on bad production for the high rate time.

We had horrible production here as well on Monday and Tuesday. Finally a bit on Wed, and then good production today.
Since I don't try to predict, the system used the battery up to totally cover Monday. At 11 pm, the battery hit 51 volts and stopped running the house. I still net exported a tiny bit that day, even with only 3.9 KWHs of solar coming in. That is bad, I should have blocked using the batteries outside of the high rate, at least from the main panel. I will have to think about how to work out that logic.

Then on Tuesday morning, The battery sat at 50.87 volts all the way until the sun came up and it started charging.

But then on Wednesday it appears the tiny load of the DC charge controller pulled the battery down or something. A bit before 2 am (1:51 am), the XW-Pro started a charge cycle. The odd thing is, the voltage never fell below 50.94 volts. So was it Recharge? At least it stayed at just 5%, 7 amps, so it was not a huge power drain. But it happened when we got fair sun back. IT didn't quite top out the battery but it hit 56.90 by 4 pm, and it normally stops at 57.00 so it was very close. The 700 watts for 6 hours during the cheap rate is not that big of a deal, and I am still showing a net export for the month.

So today, the voltage only fell to 53.01 so with the good sun, it fully topped out the battery to 57 volts before noon.
But now it looks like the BougeRV DC charge controller went dumb again. It still has the historical data. The production on 11/6 makes no sense. It says it only made 850 watt hours, but then on the 7th it did 1,724 watt hours, even though we had lousy solar output. I think the Daylight savings time change may have confused it. Production from the Enphase system was still pretty weak on the 8th, but the DC system shows 3,174 watt hours, which is quite good. That is 3.17 Sun Hours for the DC system, vs just 1.96 Sun Hours from the Enphase system. Something is off. Then today, the DC system was way down, because the voltage got over the bulk charge limit so early, the DC was curtailed. I am going to send an e-mail to BougeRV. I don't want it to keep losing it's settings.
 
When I decide charge should stop, I just send the Charge "No Float" command Send the number 3 to location &H165
Is that really all it takes? I almost tried that, but decided "no float" couldn't be it.

Thanks

We had horrible production here as well on Monday and Tuesday. Finally a bit on Wed, and then good production today.
It's been poor production up here too. But, enough to cover my loads for the day, without charging the car.
Since I don't try to predict, the system used the battery up to totally cover Monday.
I'd love to have the skills to program some predictive model based on weather forecast...
But then on Wednesday it appears the tiny load of the DC charge controller pulled the battery down or something. A bit before 2 am (1:51 am), the XW-Pro started a charge cycle. The odd thing is, the voltage never fell below 50.94 volts. So was it Recharge? At least it stayed at just 5%, 7 amps, so it was not a huge power drain. But it happened when we got fair sun back. IT didn't quite top out the battery but it hit 56.90 by 4 pm, and it normally stops at 57.00 so it was very close. The 700 watts for 6 hours during the cheap rate is not that big of a deal, and I am still showing a net export for the month.
That is odd that you're seeing this repeatedly. It's never happened to me.
 
That is odd that you're seeing this repeatedly. It's never happened to me.
It's only the second time it started charging without being told. The first time, I assumed my code did it. Nothing logs if a command was sent or not. I am pretty confident that my PLC did not send the code this time. I think the XW really did see a low enough battery that it triggered the recharge event.

Since I don't need that any more, I will set the recharge volts much lower and see if it makes a change to this. The grid support volts will still stop it from inverting. Right?
 
I'd love to have the skills to program some predictive model based on weather forecast...
I wonder how hard a coarse calculation would be. For my location I doubt adequate information exists to accommodate the microclimates, and there is a big difference between getting rain clouds at 9AM vs noon this time of year.

The math itself is easy enough to do with NodeRed. It would all come down to how granular cloud forecasts you can get. A starting point is to pull the 8760-hour spreadsheet that PVWatts can generate; I think it actually has the formulas in it.
 
I am using Node-Red currently. It's pretty easy/addictive. I believe I have 5 or 6 flows set up currently.

I just added a flow to pull SOC, range, and charge connector state from my Chevy Bolt and I'm working on one to put the Schneider back into inverter mode when there isn't enough PV to cover loads.


But, now that I think more seriously about adding a predictive model, what would I do with it?

I don't have anything I could use for dump loads I could enable when there's going to be excess sun.
For the past couple months, I have been limiting the car to 6 amps @ 240 volts.
In the spring I was having the EVSE ramp up current to follow excess PV and take priority over the home battery.

I suppose I could raise the 6 amp limit when the model predicts excess PV generation past what the home battery would need.
 
Well, no progress on the software.

But I did add 4 more west facing panels to my Solar Edge AC coupled system.
I picked up 4x Hyundai 365 watt panels that just barely fit by rearranging my west facing array.

Basically, I've got 12 panels where there used to be 8.
Took my total AC coupled array to 7960 watts on the 6,000 watt inverter.
I really should have gotten pictures...


I realize it's a waste of time/money, but this is a hobby and the 11 year old was interested in building a solar tracker. So, we're off down that rabbit hole. Wish me luck, hopefully it isn't too expensive.
We're looking at building something to hold 2 or 3 of the same big 480 watt panels I just added with the Midnite charge controller.


I think the car is going to charge a bit easier this winter. As long as these stupid clouds blow over.
 
I shouldn't have said anything.
Roughly 9kW of panels and it doesn't even hit 1,000 watts at peak production today.
2.5kWh total production for the day.

Screenshot_20221201-181721.png
 
You still beat me. My 4,800 watts of Enphase panels only did 1.6 KWHs yesterday. And the 1,000 watts of DC panels did 479 watt hours. I am surprised the DC panels actually did that well. But it was not enough to do anything. The house basically ran on grid power the whole day. I put 2 KWH into the battery from grid over night.
 
Finally made some progress towards adding the remaining 16 - LG LG425QAK-A6 (425W) to the shed roof.
Started in Sept, but hurricanes Ian and Nicole diverted my time and energy to cleanup and repairs. (Irritated chainsaw noises...)

The Schneider MPPT 100/600 CC for the LG panels is all wired up, waiting in "standby" mode.
Finished installation of the rails for the 8 LG panels on the West facing roof.
Daughter says she will help me lift the 45lb, 75" things up there, on her day off from work.
I was a little concerned about adding 1200 lbs to the shed roof, but as it is distributed over all the trusses, should be OK.
(400lbs for the existing REC365AA & mounts, 400 more for the LG, x2)

8 more LG panels are going on the East roof, so as to pick up more sunrise energy.
One tree is blocking East insolation until about 10AM.
The tree is being terminated w/ extreme prejudice. (Happy chainsaw noises...)

No hurry, as I am trying my best be safe.
It is still hot weather here.
Taking breaks and staying hydrated.

The added 6800 watts are not going to be needed until May-2023, as the pool pumps are off for the season, so the Schneider equipment is only running the heat pump, well pump & misc small loads.
LG_rails_west_roof.jpg
tree_east_panels.jpg
 
Last edited:
Nice progress!
I think you're right on the extra weight, 50 lbs per truss isn't much. But, might be worth it to double up on the hurricane ties so the roof and panels together don't leave!

I think your label on the tree would be more accurate "tree, soon to meet it's doom" lol

What is with the antenna/mast/guy wire situation?
It looks like the top is bent over under tension to one specific guy wire and that some of the guy wires are paralleled with spacers between them. I've never seen anything like it.
 
It looks like the top is bent over under tension to one specific guy wire and that some of the guy wires are paralleled with spacers between them. I've never seen anything like it.
Hurricane Nicole broke some of my antennas.
Have not fixed that one yet!
The spacers are part of the antenna feed system - "ladder line"
 
Oh, sorry to hear that. It didn't quite look broken at first glance. But now that you say that I can totally see it.
 
Back
Top