Adding storage to my Enphase system

[Shimmy]

Once you add DC charging, I think you can make the XW-Pro do what you need without the PLC or Pi to control it. But you will need to size the DC array to provide the power you want from the batteries. The AC coupled solar won't provide charging to the batteries.

And that is where everything breaks down for me.

I would need 3.5kW on the DC side to cover battery charging, but I only need another 1-1.6kW of PV for my other needs. I guess my cheapest route would be a couple 20A AC chargers tied into the batteries... which would need some automation control but nothing too complicated.

 

[GXMnow]

Shimmy makes a very good point about trying to sell the PLC setup.

That is a big reason I have not looked into selling it as a package. There is no easy user interface to even set the ip addresses, let alone the sell and charge threshold levels and times. I did start to move a lot of the setup variable to a non volatile table area, and the PLC does support a web UI that could make it easier to change them, but I am not at a comfort level to put all that code together yet. If someone wanted the experiment with the code, I would just give them a copy of what I have, with some notes about what variable are doing what.

Shimmy said...
"I would need 3.5kW on the DC side to cover battery charging, but I only need another 1-1.6kW of PV for my other needs. I guess my cheapest route would be a couple 20A AC chargers tied into the batteries... which would need some automation control but nothing too complicated."

Before I bought the Schneider XW-Pro, I looked into the idea of using three 600 watt 48 volt battery chargers. Then have a simple power meter kick each one on as the power from the grid tied solar inverters became high enough to cover the load. So that would give the charging. Then I was looking at those cheap Ali Express grid tie inverters that can limit to zero export with a grid side current sensor. The more I thought about it though, I didn't like the idea of 4,000 watts going through those cheap boxes with no safety certs. And the XW-Pro really did not cost that much more when you figure in that is can charge and invert at up to 6,800 watts with very fine control and UL and California grid code legal support.

When the stupid firmware refused to star an AC coupled charge cycle, I almost implemented my idea of the three or even four 600 watt battery chargers. I actually have one that I use to charge my E-Bike batteries. And for a month, I did run it off of a Christmas light timer to kick it on for 8 hours while the sun was shining. It worked ok until the timer failed. 8 hours at 600 watts was putting 4.8 KWHs into the battery bank each morning. And then the XW-Pro would push that back to the grid from 4 pm to 9 pm during the peak rate. It was not ideal, but it worked.

 

[silverlexg]

Shimmy makes a very good point about trying to sell the PLC setup.

That is a big reason I have not looked into selling it as a package. There is no easy user interface to even set the ip addresses, let alone the sell and charge threshold levels and times. I did start to move a lot of the setup variable to a non volatile table area, and the PLC does support a web UI that could make it easier to change them, but I am not at a comfort level to put all that code together yet. If someone wanted the experiment with the code, I would just give them a copy of what I have, with some notes about what variable are doing what.

Shimmy said...
"I would need 3.5kW on the DC side to cover battery charging, but I only need another 1-1.6kW of PV for my other needs. I guess my cheapest route would be a couple 20A AC chargers tied into the batteries... which would need some automation control but nothing too complicated."

Before I bought the Schneider XW-Pro, I looked into the idea of using three 600 watt 48 volt battery chargers. Then have a simple power meter kick each one on as the power from the grid tied solar inverters became high enough to cover the load. So that would give the charging. Then I was looking at those cheap Ali Express grid tie inverters that can limit to zero export with a grid side current sensor. The more I thought about it though, I didn't like the idea of 4,000 watts going through those cheap boxes with no safety certs. And the XW-Pro really did not cost that much more when you figure in that is can charge and invert at up to 6,800 watts with very fine control and UL and California grid code legal support.

When the stupid firmware refused to star an AC coupled charge cycle, I almost implemented my idea of the three or even four 600 watt battery chargers. I actually have one that I use to charge my E-Bike batteries. And for a month, I did run it off of a Christmas light timer to kick it on for 8 hours while the sun was shining. It worked ok until the timer failed. 8 hours at 600 watts was putting 4.8 KWHs into the battery bank each morning. And then the XW-Pro would push that back to the grid from 4 pm to 9 pm during the peak rate. It was not ideal, but it worked.

I appreciate the detailed responses, I’m struggling with a very reliable grid (Kansas), very easy install with micro inverters but also wanting to time shift. The XW pro looks fantastic but my panel location isn’t ideal (no physical location for the XW pro or a critical loads panel). I really love the ease of adding 48v batter banks. The emphase battery option is very expensive, and requires basically everything the XW pro needs (panel, wiring, critical loads panel, and wall space), and then I’m tied to their battery. I’ve thought of using a few spare microinverters attached to whatever flavor of 48v battery and charge when excess power exists (I have an emporia vue for monitoring, I could use a Shelly em or something similar as well, turn on chargers to push power into the battery and then flip grid onto the micros at night and let them provide power as needed. It’s basically what enphase is doing but obviously lacks the grid form or grid failure capabilities. Much much cheaper which Is appealing (and allows whatever battery option i want). Im still getting solar panel and capacity worked out atm. I’d like to do it myself, especially with some of the hack installs I’ve seen from pros. I’ll be sure to follow up once I make a decision. The XW Pro may have to be for the next house when I have more say in the home layout.

 

[GXMnow]

My system had an odd glitch overnight.

I found what caused it, but can't explain why it did what it did as a response.
Late last night, the XW-Pro logged an error "AI Over Frequency" so it disconnected from the grid briefly. I have no idea how long, but it appears to have been very short. The XW only logs the disconnect, not the re-connect. There was no sun, so the Enphase solar did nothing and there is no log entry in the Envoy from 6:45 PM to 6:45 AM.

So the odd thing is, when it re-connected to grid, it started a charge cycle. It does not look like the battery voltage fell to recharge volts, so I don't think the XW did it on it's own. And the energy flow on the XW output should never have gone reverse, so the PLC should not have commanded a charge cycle. I think the Gateway may have glitched and sent bad data back to the PLC. Maybe it reversed a sign bit or something. It ended up charging at about 35 amps into the battery from grid power. Had it been a single glitch, it should have just been a fixed charge current. But here is the odd part. Looking at the battery summary graph, the PLC was obviously adjusting charge current to track load as the current went up and down several times, even making single 1% steps as well as bigger jumps. For example, I see it dip similar to when the refrigerator cycles, but the pattern is a little different. In the early morning before the sun comes up, I am used to seeing the battery discharging, with the steps going to 3 amps more discharge current when it is running and then back when it cycles off. This morning the system is charging instead of discharging, and the steps are 6 amps instead of 3. It is dropping from 36 amps of charge current to 30 amps, and then back.

The battery became full at about 5 am, and it went back into grid support and started running the house off of the battery. Everything seems normal after that. I did a reset on my PLC, but not the Gateway yet. Running the house for 3 hours pulled the battery down almost a volt, so when the solar started producing, the PLC saw the energy coming in and properly triggered a charge cycle to start again. It won't take very long, and it will be back full, and I will be exporting my extra solar today. It will offset a lot of the cheap energy I used overnight charging it up.

I need to think through what triggered the charge cycle, and why it stayed charging using grid power like that. At the least, the PLC should have dropped the charge current when it saw energy coming in from grid like that. I don't understand why it decided on the 36 amp charge current, and what was it tracking to make the adjustments in that range. Too bad my PLC does not log anything.





Here is the battery graph from midnight up until the sun started coming up.

 

[Shimmy]

Could a race condition have started the charge-- low voltage battery due to discharge and then a return of utility power? That is a common problem with using modbus for "real-time" controls.

 

[400bird]

Wow, that is odd.
I've got my program set up so it checks against measured PV output and won't go above that, just in case something odd happened like you had last night.

 

[GXMnow]

Wow, that is odd.
I've got my program set up so it checks against measured PV output and won't go above that, just in case something odd happened like you had last night.

That is a good idea. I don't have the PLC reading the true solar production, it only knows the difference between solar production and the loads in the backup panel. I just get that from the output power reading in the XW. My routine should see when no power is coming in it should not set it to charge, so this is still odd. I could get one more of the cheap power meters and directly read the solar as well.

 

[GXMnow]

I have been wanting to add a bit more solar for a while and I have been watching prices and thinking about pulling to trigger. The problem is the shipping. I don't have a vehicle that will hold full sized panels. Renting a truck is about the same as shipping costs. For just 2,000 watts of panels, the cost of renting a truck ads a lot to the panel cost per watt.

Up to a few days ago, I never would have thought about using smaller panels. But something changed. There are now a few companies that sell on Amazon that have gone to the latest mono cells in a smaller form factor. Typically these panels were not very cost effective. Some are going for over $2 per watt. I can get brand new 60(120 half cut), 66(132 half cut), or 72(144 half cut) panels all for around the $1 a watt mark, with the cheapest dropping under $0.80 per watt. But then there is the shipping issue. Real Goods are offering a free shipping deal, and I almost did it, BUT.... it is a 9 panel minimum order, and the only panels they have under $1.00 a watt at 360 watt and larger panels. 360 x 9 = 3,240 watt of panels. That is 50% more panel than I need. And even with free shipping, it is $2,600 in panels. Not out of the question, but I really can't fit that many panels.

Then one ad on Amazon caught my eye. A small 100 watt panel. Most of them are junk, but I clicked on it. I took a look at the data sheet and it seems legit. I even found a few reviews for it, when it was selling for $120. And it makes it's ratings. It is using the newer mono perc cells and even has a 25 year production warranty. And it went on sale for $84 for the 100 watt panel. So $0.84 per watt. And free shipping. And only 15 pounds per panel. They are more efficient than my current 300 watt panels. 3 of them is still less area than one of my panels, and they even have a better temp coefficient. The only down side I am seeing is having to mount up 3x as panel panels. That might seem like an issue with the RSD stuff too. But since these are just 32 cell panels, 2 in series is still under the voltage limits for shut down, and they are also only 5.5 amps panels, so they can also be wired in parallel pairs. So every 4 panels can be treated as a single 400 watt 128 half cut cell panel. 5 of these groups hits my 2,000 watt goal. I just drew out a diagram, and I can get away with just 3 of the Tigo dual panel boxes and be well within spec. 2 of the boxes will each handle 8 panels, wired as two 400 watt series parallel groups. The Tigo box is rated at 500 watts per channel, 90 volts per channel, and 15 amps per channel. The quad panel group is only 44 volts VOC at 11.1 amps ISC. The third Tigo box will only have 2 panels in series on each channel, so half the current. Then the outputs from the 3 Tigo boxes will feed into the charge controller. 88 volts VOC and 27.75 amps ISC. VMP is then 76.24 volts, and IMP is 26.3 amps. That is my 2,000 watt target.

But then I still need to deal with Arc Fault.... Still no "cheap" option. The Midnite Classic charge controller still looks like the lowest cost option. But the versions with the Arc Fault protection are all over $600 so my budget is still stretched. The value of the cheap panels becomes less and less.

 

[dougbert]

this one?

https://amazon.com/Newpowa-Monocrystalline-Compact-High-Efficiency-Off-Grid/dp/B09D7BMBWP/ref=sr_1_8?crid=34K1Q2VUNISGD&keywords=100%2Bwatt%2Bsolar%2Bpanel&qid=1664914081&qu=eyJxc2MiOiI1LjgzIiwicXNhIjoiNS41MSIsInFzcCI6IjQuODMifQ%3D%3D&sprefix=100%2Bwatt%2Bsolar%2Bpanel%2Caps%2C124&sr=8-8&ufe=app_do%3Aamzn1.fos.006c50ae-5d4c-4777-9bc0-4513d670b6bc&th=1

 

[400bird]

It never occurred to me to use 20 tiny panels. Interesting idea. I went absolutely the opposite direction and have 4 giant 480 watt panels.

 

[Ampster]

I was intrigued with the idea until I thought about the racking. It would not work with my one DC string so the cost of the micros has to be figured into the overall equation..

 

[GXMnow]

Yes, the ones I am looking at are the NewPowa panels. I almost got the Bouge RV 100W ones, at $83 each, but their price went back to $107

If I had a way to go pickup 4 big 480 watt panels, semi locally, I would also go that way. I just have not found a reasonable way to do that.

The racking of 20 small panels is not ideal. Since the panels are so small and light, I am thinking about pop riveting then into groups of 4 to make virtual 400 watt panels. Then those could mount nicely on normal racking rails. The series parallel group of 4 panels should work perfectly with an Enphase iQ7A inverter. But for this experiment, I am going to go the DC charge controller route.

I may also try the new Bouge RV 40 amp charge controller. It will take 2,400 watts of panels on a 48 volt system and it has bluetooth monitoring with history graphs. It looks like a quality unit. And, if/when I upgrade my system to fully legal, I would then use it for my portable/shed power power system. I already have the 120 amp hour 3S battery left over from my Bolt cells. I have been using it as a power bank. I could leave that in the shed on float and the little inverter in there could run the chargers for my cordless power tools and run 12 volt LED lighting out in my yard.

If I do this, it will be a "portable" battery charging rig, that I just set up on my garage roof for a while. 20 of those panels comes out to about 9 foot wide by 12 foot long. Wired as either 4S 5P (76.24 volts at 26.3 amps) or 5S 4P (95.3 volts at 21.04 amps) it is well in range of the Bouge controller. It has a max recommended input voltage of 145 volts, the 5S VOC is only 109.1 volts, so even dead cold should be fine as 20% up is still just 131 volts.

So yeah... I thought about this a little. If I try it with 10 panels as a "portable" system charging my battery bank, I can try it out with just the panels charge controller, cables, and fuses. It won't technically be legal, but it's not over my living space. If I mount then as an awning along the side of the house, it skirts that rule. If I like it, I can add the Tigo boxes and upgrade to Midnite charge controller for the arc fault protection when I see one go on sale. Or just maybe someone will offer a reasonably price combiner box with arc fault detection in the near future. What I am proposing is every bit as safe if not more so than many of the things Will has in his videos. If I rack the panels on the garage roof, I will at least add the Tigo RSD setup.

My cart at Amazon is now at $1,100 for 10 panels, cables, MC4 connectors and crimper, and the Bouge charge controller. That's not bad for a 1,000 watt charging setup. And just add 10 more panels for 2,000 watts total. I may even be able to fit a 6th row of 4 panels and go to the full 2,400 watts. 20 panels would be 9 feet wide by 12 feet long. To go to 24, I would turn the panels the other way, the 4 wide becomes 10 feet across, and the 6 panels long goes to 14 feet. I think either setup will sill give plenty of walk way around the array on my garage roof.

 

[GXMnow]

So now the title should be "Adding some DC charging to my Enphase/Schneider system".
I placed the order for just 10 of the 100 watt NewPowa panels and the Bouge RV 40 amp blue tooth charge controller.
Under good conditions, the panels should top out close to 800 watts. At 50 volts, that works out to 16 amps. It's not a whole lot, but it will be a good test to see if it is worth adding more of these, or if I want to get bigger panels.

If nothing else, I can use this stuff as a portable power station. It will work at 12 volts, but the 1,000 watts of panels is a bit overkill then. For the first test, I am going to wire the 10 panels 5S 2P. I also got #10 solar cable for the run to the charge controller. With only 1,000 watts of panels, the current won't top 12 amps, so it is way over kill, but if I go to the full 24 panels 6P 4S, it could hit over 30 amps from the solar, so the #10 is on the edge. It's high temp cable, so it is rated for more current, but that is out of NEC spec. With the full 24, I will probably do 2 drops.

I'll post some pics once the gear is here. I should be able to get a quick temp connection up and running fairly quick to get an idea how well this is going to work. If the numbers look good, I'll get the TIgo RSD stuff and be at least semi legal. And I will keep looking for a reasonable Arc Fault detection unit. If I run the panels in 3S, it would be below the 80 volt limit so Arc Fault would not be required, but, the also makes the VMP just a tick over 54 volts. That is a little low to put much charging power into my battery bank. It charges above 54 volts by 10 or 11 am.

The A/C ran for almost 2 hours yesterday, so for the first time in a full week, I ran out of power again and had to use some grid power. The 1,000 watts of panels probably would have covered it.

 

[Ampster]

I do not remember if you already had some DC coupled solar?

 

[GXMnow]

No, I don't have any DC panels yet. I have talked about doing it for a year though. This will be the first DC solar charging.

I just did a little math. If I run the panels in 3S, the VOC is just 65.46 volts, so low enough that Arc Fault is not required. Then I have it charge a 36 volt battery bank. The Bouge charge controller can do that.

That also makes it a completely separate off grid portable system. But to make use of it, I then use a DC to DC boost converter to push the energy into my 48 volt bank after 4 pm. I use about 1,000 watts pretty much steady around then, and it goes to 4,000 watts when the A/C is running. If I just have the PLC turn on the boost converter after 4 pm until the 36 volt battery drops to 31 volts or so, I would use all the power the off grid system makes each day. Obviously, the Boost converter would be set to a max output voltage of 57.5 volts, so the current would be about 17 amps to push 1,000 watts max. Sure, it is another efficiency dip, but it is a way to be legal until I get the Arc Fault safety shut off.

I just found a 30 amp step up CC CV boost converter for $42. It's rated at 97% efficient. 10 to 60 volt input works fine on the 10S 36 volt battery. And the output can go from 12 to 97 volts, so my 57.5 volt is right in the middle. At the 30 volt minimum input voltage, the 30 amp limit makes it just 900 watts. Close enough. 15.5 amp charge rate into the 48 volt battery bank will work just fine. Now I just need 10S of batteries at 60 amp hours or so. Oh crap... that costs more than the Arc Fault protection. Hmmm... I can use a much smaller capacity, if I have it transferring the power at the same time it is being produced. It will still need to handle the currents though. Worst case would be 2000 watts going in. The charge controller is limited to only 40 amps. If I use RC car batteries, I can run them up to 3C on the charge side without worry. So I would need 3P of my little 5 amp hour 30C rated packs. They would have no trouble putting out the power. They sell them in 5S at Hobby King. So 6 packs and a BMS. Any time the voltage goes above 38 volts, turn on the boost converter, and if the voltage goes below 36 volts, turn off the boost converter. It's not ideal, but it will work. 15 amp hours at 36 volts adds only 540 watt hours of capacity. That is $420 worth of LiPo batteries. Keeping the LiPo batteries in that narrow range means they are not cycling, they just float near the ideal 50% charged state. So the limited 300 charge cycle life should not be a big problem. But yes, the cost savings pretty much goes away.

Oh.... Battery Hookup has a 48 volt max bank of Ultra Capacitors. They can handle over 100 amps in and out. They can discharge to zero, so running it at 36 volts is no issue at all. It's 18 caps in series, so they would sit around 2 volt each and they are rated to 2.7 volts. Do I still need a BMS/balancer to keep the caps at the same voltage? The charge controller could be limiting to 40 amps going into the cap bank, and the DC to DC boost converter could be pulling 30 amps out. They will be coasting along. The only issue would be near solar noon, the caps might top up and the charge controller would have to go to absorb mode. The whole bank of caps only stores 54 watt hours. Any time the solar power is over 1,000 watts, the caps will be going up. With just 9 panels connected 3S 3P it won't be an issue. At 18 panels, it might be a problem. And 1,800 watts just happens to be the limit of the Bouge charge controller at 36 volts.

And the cap bank is going for $325 so cheaper than the LiPo's and a lot safer.

Yeah, this is a totally crazy idea.

 

[Ampster]

Yeah, this is a totally crazy idea.

I still have four used panels and four micro inverters from my other home. I was planning on putting them on a west facing roof to increase my production. Now that seems like a crazy idea since I just got my True Up and was a Net generator for the year. However, I am still a Net Consumer for the first four or five months and that makes me a little nervous so having some excess production might be worthwhile since I already have the equipment. The attic run will be real easy and now that the weather is shifting, getting up on the roof won't be as hot. I also am going to be adding more devices like a HPWH, Another minisplit for two more zones and possibly an induction cooktop. There is always discretionary EV charging that can suck up any of that Net generation as I approach next year's True Up. An of course there is always the uncertainty of the weather, especially in winter.

 

[GXMnow]

The parts are starting to show up. I got the charge controller. It actually has a vey high quality look and feel. The die cast aluminum housing is well made and certainly has some mass to it. I powered it up off of my 3S battery and connected to the BlueTooth ap. Here are a few pictures.



I also got the red and black #10 awg 50 foot solar extension cables to get the panels up on the roof. And I got some #8 Windy Nation cable for the run to the battery bank, since it could hit 40 amps if I max this out. The last pic shows the #8 cable will easily fit in the cable clamps on the charge controller.

All of the MC-4 connectors I have here are from Bouge RV. I also got their crimper and MC-4 connector tools. I am a bit nervous about how lose the fit on the contact pins feels. The are very consistent, with no combination of pins feeling any looser. The one loose pair of pins that came with the 50 foot cables do fit a little tighter into each other, but is seems it is the female barrel is just a little smaller, the pin matches all the others. Since these panels are only 6 amps max each, I don't think these connectors will have any issues, even after the 2 parallel strings, I am only dealing with less than 12 amps. There is a lot of contact area in an MC4, but I just expected to need a bit more force to push the pins together. I guess I expected something like an XT-60 connector. These need far less force to push together.

And the "Y" cables just showed up. Their MC-4 connectors actually fit tighter, but I am a bit disappointer they are made with just 4 mm^2 wire instead of the 6 mm^2 wire I got for the home runs. But yes, it can take the current of 12 amps and it's only a foot of it.

The last thing to come will be the actual solar panels. They are now showing getting here tomorrow (Tuesday 10/11/22) by 10 pm.

 

[Risky Rob]

The parts are starting to show up. I got the charge controller. It actually has a vey high quality look and feel. The die cast aluminum housing is well made and certainly has some mass to it. I powered it up off of my 3S battery and connected to the BlueTooth ap. Here are a few pictures.

View attachment 115834View attachment 115833View attachment 115832

I also got the red and black #10 awg 50 foot solar extension cables to get the panels up on the roof. And I got some #8 Windy Nation cable for the run to the battery bank, since it could hit 40 amps if I max this out. The last pic shows the #8 cable will easily fit in the cable clamps on the charge controller.

All of the MC-4 connectors I have here are from Bouge RV. I also got their crimper and MC-4 connector tools. I am a bit nervous about how lose the fit on the contact pins feels. The are very consistent, with no combination of pins feeling any looser. The one loose pair of pins that came with the 50 foot cables do fit a little tighter into each other, but is seems it is the female barrel is just a little smaller, the pin matches all the others. Since these panels are only 6 amps max each, I don't think these connectors will have any issues, even after the 2 parallel strings, I am only dealing with less than 12 amps. There is a lot of contact area in an MC4, but I just expected to need a bit more force to push the pins together. I guess I expected something like an XT-60 connector. These need far less force to push together.

And the "Y" cables just showed up. Their MC-4 connectors actually fit tighter, but I am a bit disappointer they are made with just 4 mm^2 wire instead of the 6 mm^2 wire I got for the home runs. But yes, it can take the current of 12 amps and it's only a foot of it.

The last thing to come will be the actual solar panels. They are now showing getting here tomorrow (Tuesday 10/11/22) by 10 pm.

There's a sweet deal on Ebay for 200W panels.

If you are in the market for 200w panels next 12hrs (10/10/22)

Just happened to notice over on Ebay that Eco-worthy has their 200w RV panels on sale for the rest of today. Free shipping. Not as cheap as used panels but I have no access to used in my neck of the woods. Price about $175.99 I ordered 2 of the 200w for myself for a RASP (Redundant Array of...

diysolarforum.com

 

[GXMnow]

I looked at a few different ads on Ebay. That is good pricing for 200 watt panels. The ones that look for real all seem to come in at the 80 to 90 cents per watt range. I got the NewPowa ones off Amazon for 82.25 cents per watt. So if they produce as claimed, they are a decent deal, especially when working with smaller panels. I have had a little bad luck with Ebay in the past, so I am a little skeptical. I have gotten some bad crap on Amazon as well, but it is usually easier to deal with returns and refunds. I'll report on the panel quality once I get them connected. The delivery notice says before 8 pm tomorrow, so there might not be any sun when they get here.

 

[GXMnow]

My SilFab/Enphase panels made just over 4 sun hours today. Just over 20 KWHs on 4.8 KW of panels. Assuming the NewPowa panels are honestly rated, the 1,000 watts of panels should then make 4 KWH's towards my system. My existing system, was able to push 11.6 KWHs into the battery bank from the energy beyond what my house was using all day. So over half of my solar production went into the battery bank. This small 1,000 watt array of DC coupled panels is going to make another 34% energy into the battery bank. The battery did make it over night yesterday on just the 12 KWH's from the Enphase system, but it never did top it up. It only got it up to 55.3 volts or 3.95 volts per cell. That works out to just 85% charged. 4 KWHs is 11% of my battery bank capacity, so today, it could have pulled the battery up to 96%. I don't really want to pull them that high, but what that means, is it is the difference of topping up the battery, and stopping short at 4 pm each day.

Obviously, if it was hot summer with the A/C running, I would still not be topping up the system, but it goes to show that just 1,000 watts of solar panels is a lot of energy. I really hope the panels and this charge controller are for real. My total investment to get this temp portable rig up an running is going to be around $1,300 The grounding for the panels is going to be a chunk of that since I need a WEEB ground lug for 10 separate panels and a lot of #6 awg solid copper grounding wire. Grounding a single rail in a "normal" array is a lot less pieces and labor. That is the one big drawback to using a bunch of small panels. And they are not really that small. About 28 inches square and 15 pounds each. For the metric people that is about 0.5 square meters per panel at 7 Kg. 5 of these panels are needed to match the big panels 400bird just got. His weight 65 pounds each. 5 of these is 10 pounds more at 75 pounds. That is probably due to more frame to cell ratio.

 

[400bird]

Yeah, I've just got the one grounding lug for the new panels, easy as pie. I hadn't considered grounding all those tiny panels. Are you not going to put them on racking?

Also, does it feel like the pack isn't making it full like it did in the spring? This feels like similar weather, but I lost my data to back that up.

 

[GXMnow]

Yeah, I've just got the one grounding lug for the new panels, easy as pie. I hadn't considered grounding all those tiny panels. Are you not going to put them on racking?

Also, does it feel like the pack isn't making it full like it did in the spring? This feels like similar weather, but I lost my data to back that up.

I have not decided on how I am going to mount all the little panels just yet. I want to see how well they perform before I dump more money into this setup. But for safety sake, if they are going to end up sitting on my garage roof, they will be grounded down to the ground rod next to my garage. WEEB grounding lugs look like they are going to hit me for about $5 a panel. I found some cheaper WEEB washers I could put between the panel and a cheaper lug, or to a rail of aluminum to tie a row together to a single ground lug. Bolting ground wires onto aluminum has to be done right.

Basically, with the lowering sun angle and shortening days, I am now down to just 4.1 sun hours coming in from the Enphase system. That is just over 20 KWH's and I am using that much without the A/C coming on. And since we are still into the 90's every day, the A/C still runs once in a while. So I am just falling short of coving the loads. I still have it exporting a tiny bit while the sun is up because the 1% charge rate steps are just too big. If I try to cut it any closer, I could end up buying power from time to time. It averages out to exporting just 60 watts or so, so not much at all. After 4 pm, I am a bit torn. I have it stop charging. The power is worth more as export after 4pm at 49 cents than it is if I charge the battery and use it again after 9 pm. If I run the battery out before 9 pm, then storing it might be more cost effective, but that was only happening when the A/C was running several hours including the 4 to 9 block. Last night, it just made it to the sun coming up at 8 am. The previous 2 days, it gave up around 4 am, so I bought cheap rid power from 4 am to 8 am. 4 hours at 700 watts = 2.8 KWHs. The 10 x 100 watt panels should cover all of that and a bit more. The dead of winter might get interesting. Gong back to January, I was having to buy about 3 KWHs a day on average. The Enphase system was doing 17 KWHs a day. That is 3.5 sun hours. The 10 panels could then make 3.5 KWHs. So it could even zero out my January energy usage. And this is just the 10 panels. I could go all the way to 24 x 100 watt panels without over powering this charge controller.

Once I see what I get for a few weeks with the 10 panels, I will have a much better idea if this is going to do what I expect. If I keep topping out the batteries, my options are discharge the batteries deeper each night, or just know, the charge controller is throwing away some power. In which case, I can se the PLC to reduce the charging from the AC coupled Enphase and have it export a bit more for credit. Even credit at the cheap rate is better than just curtailing production. But honestly, that is still not going to happen in summer with the A/C running. Maybe with the full 2,400 watts of DC panels??

 

[GXMnow]

The panels are here.

They are packaged very well. One of the 5 boxes (2 per box) showed a little damage, so I opened that one first. The panels are packed fact to face with some nice foam between them. Then the outer edge is wrapped in cardboard with a layer of foam inside as well. There was no hint of visible damage anywhere. Build quality look pretty good. The corners are a bit sharp. And the frames don't have drain holes like most larger panels. Since I will have them near flat, I don't think its too big of a deal, but I will keep an eye on that. We don't get much rain, but when we do, it can be bad for a day or two.

Of course the sun is really low in the sky. My entire 4,800 watt roof array was down to just under 300 watts total by the time I got one of the new panels connected up to my 12 volt test rig. Tilting the panel right at the sun, barely above the horizon, through some clouds, I saw it hit up to 22 watts from the single 100 watt panel. I wanted to see more, but with how low the sun was, I am not disappointed. Before I got it on the charge controller, I did do a quick open volts/shorted amps test. I got just over 20 volts, rated at 22. And I got 3.5 amps, rated at almost 6. So yeah, the sun was low and the sky is hazy. It took my another 20 minutes to get the charge controller wired to the battery, and to plug in the panel and get it aimed.

I may be able to get a few laid out for the morning sun to get some charge.

 

[GXMnow]

I got the charge controller wired into my battery bank, and I made up cables to run outside. In the morning, I will connect up some panels and see what it does.

The charge controller has a few limitations. The voltage settings are fairly big steps. Do I want the absorb volts to be 56.8 volts or 57.2 volts? And it has no setting anywhere to limit the current. This is not an issue for me as I want everything the panels can make. And the 40 amp limit of the charge controller is really nothing for my batteries. But that does explain why they say to not over panel past 40 amps on the input or 2400 watts total. 40 amps into 57.2 volts would be 2,288 watts. I know my 10 panels won't even come close. I am hoping for 80% or 800 watts peak out of the 10 x 100 watt panels. At my minimum voltage of 51.0 that would be 15.69 amps. I have it connected to the buss bar on a 30 amp MBRF fuse. With the 5S2P panel configuration, it should be about 85 volts at 9.4 amps from the panels. The #10 awg cables are complete overkill. I have to unbox the rest of the panels.

One reason for the price is that the panels are CE rated, but not UL Listed. And "Made in Vietnam".

 

[Hedges]

So that's what the box meant by "off grid".