Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Building my own pseudo(Powerwall)

This site may earn commission on affiliate links.
I wasn't aware of there being an ability to "curtail" solar inverters using the frequency. Where is this specified and does the Sunny Boy 7000 implement this capability? I was aware of the ability to shut off any UL1741 inverter by raising or lowering the frequency above or below the limit. It seems almost abusive to use that capability to cycle the inverters every 5+ minutes plus charge time. I was a system designer and shutting down system power that frequently just seems like a rough way to limit charging.

I'm not sure if the 7k has it but here;s the official SMA doc with models that support it, page 7 has info on curtailment
https://files.sma.de/dl/7910/SB-OffGrid-TI-US-en-20.pdf

Here's the tech specs on Enphase (although only the IQ line can curtail) page 6
https://enphase.com/sites/default/f...Considerations-AC-Coupling-Micros-Battery.pdf

Another thing that bothers me is both my solar inverters produce 240V and not true split phase 120V. That suggests that home 120V appliances cannot really use solar inverter power since they draw power between neutral and one 120V phase which my inverters do not support. That suggests solar power cannot be used by the home except real 240V appliances. The home would only be able to use true split phase 120V/120V to neutral and the system inverter must support split phase with flexible loading on either phase. Is my thinking correct? Tesla's PW2 inverters claim to produce flexible split phase with the load on either 120V phase.

Wait, your inverters output single phase 240? are you sure it's not split phase 240? I don't know anywhere in the US where single phase 240 is used, that's mostly overseas.
Page 38 in the link below shows the 2 configurations for the 7k, you have the bottom on with only N, L1 and Ground(PE)?
https://www.cedgreentecheast.com/files/SB5678 INSTALLATION(3).pdf


The charging problem: I don't see a charger power supply in your list and the one you specified in an earlier post is no longer available. Solar inverters are current sources. They won't naturally provide less than 100% of the available solar power less inefficiencies. They will increase their forcing voltage to try and force current into a grid load. If they cannot Force 100% of their current onto a load, they will increase their voltage until they shut down due to overvoltage. That is not good when one is trying to limit both the maximum charge current into the batteries and stop charging one the battery is full.

The inverter I'm using is made by Sigineer
12000 Watt Inverter | Pure Sine Wave 48V 12KW Inverter Charger
It's an inverter charger so it has a charger built in that lets you charge the batteries off the grid. However, in an AC coupled system the built in charger isn't used as much, the reason for this is because the grid tied inverters are connected on the output side of the battery inverter. When the battery inverter switches its internal ATS from grid to battery backup it controls the frequency since its internals are generating the sinewave. After 300 seconds (per UL) the grid tie inverters will have synced to the sinewave of the battery inverter and will start to produce power. This power will start to power the loads and start to lighten the load on the battery based inverter, if you are producing more power than the loads require the power will flow backwards through the H bridge on the inverter, will be converted to 48v and it'll be stored in the batteries, if the batteries are nearing full the battery inverter will start to increase the frequency causing the grid tied inverters (which at this point are synced to the battery inverter not the grid) to start to lower their output, when the batteries fill the battery inverter will simply shut off the GT inverters by rising the frequency to 62.5hz or above. The GT inverters will stay off until the frequency drops again.

On my inverter (Sigineer) where there is no gradual shift of frequency only trip, I have my controller configured to listen to the EVTV controller which is sending out messages on the battery state, when a cell reaches 4.1v I tell the Sigineer to trip and turn off the GT inverters. Every 2.5 minutes it continues to trip for 200ms (UL requires a 5 minute timeout between trips before GT inverters can continue to output) this ensures that the GT stay off until the battery drops by 10% (threshold configured by me) at which point my controller stops triping the frequency and the GT inverters start to produce, when a cell reaches 4.1v again the cycle repeats, in reality the time window depending on what loads I have running is about 1.5-2 hours.


So I'm not sure how your actual charging system works. An off the shelf power supply doesn't provide the constant current needed for Lithium batteries plus must be stopped at the maximum voltage and switched to a voltage source and then stopped. What capabilities do the Tesla BMS's offer? Does Tesla provide specs on that? I see "BMS $2,495.00 EVTV 5/17/29" but I don't understand how one goes from raw AC to a proper Lithium charger with the appropriate current limit, overvoltage limiting, undervoltage cutoff, low temp cutoff etc being done by a coordination of a powersupply and the Tesla BMS, EVTV and solar inverter curtail or 60HZ 5 minute shutdown cycles. I need to look at it closer.

The EVTV BMS controller has contactors built into it. It also communicates with the BMB boards inside each module, it gets voltages of every cell plus the temperature for each thermistor. I can configure OV, UV, OT all within the EVTV controller. When one of those limits is reached the contactors inside the controller open.

upload_2019-12-29_18-57-22.png


Doesn't it mean the maximum power output of my solar inverters at say 11KW must not EVER exceed the maximum charger current of the battery pack ( or there must be a way of throttling the inverters such as the frequency control you mentioned). But my Sunny Boy solar inverters were producing 6.6KW yesterday afternoon which was a sunny day just past the lowest sun day of the year. So wouldn't the battery pack size be set roughly at my 11KW or 11KW+4KW= 15KW per day full output since that's the most KWHRs I can charge in a day. That suggests my battery size is about the size of 60KHWHrs of a full Tesla battery pack since I can produce about 70KWHrs on a sunny summer day.

You're partially correct on the charging side of things, if you remember the table I posted on my last post it had my calculations on it, broken down to this:

14,500kW of solar / 240v = 60.41A, that's the amperage that will be sent to the inverter at no load and full solar (worst case scenario), if we step that down to the 48v side of things (14,500kW / 48) 302.08A will be sent to the pack to charge the batteries. Each module can take a peak of 8kW (360A) and a continuous 5kW (225A) of charging amps. If i had a single module that would present an issue as a spike in solar productiong lasting longer than 10 min (8kW peak) would cause damage, but I have 14 modules, this means that the 302A is actually 320/14 = 21.57A, or in other words, I'm charging each module 52% below the max charging amps. In other words the battery output/input capacity is a lot higher than what my solar can produce so I don't need any kind of charge controlers to lower the amperage being sent to the battery.

Now that being said, near the top end 4v+ side of things throwing that much power will not get you a nicely balanced pack, this is why I have my controller set to disable the GT inverters when a cell reaches 4.1v so the pack can balance itself without reaching its limits.

The second part, you don't always have to match your battery size to your dialy avg output. For around spring I have days when I produce +100kWh but my pack is only 60kWh. when that occurs my controller knows if the grid is available or not if it is, instead of triping the frequency and stopping solar production it tells the sigineer inverter to switch to grid mode (essentially stand by mode in a traditional ATS) this directly ties my GT inverters to my power company grid and sells any excess back after the pack is full.

I need to measure how much power I can get away with over night to see what my lower limit is. The PG&E power shut offs this year were 3 days each and we rarely have cloudy days in August thru November. That would set my lower battery size that will "get me by". Since I can only produce about 6-7KW peak in November, that would suggest a battery size in the 30-40KWHrs size since a larger battery would have little value in November or during winter storms when the power outages are most likely. One can buy 5KW Tesla modules for about $1000 on EBAY which would require about 6 or so for $6000. Two Tesla PW2's would be 28KWHrs with 10KW continuous power output.

Yes, having a good idea of power usage is a must here. Oversizing is always fine, under is a bad idea.

The Tesla PW2 design moved my 2 Sunny Boy's that are line side fed to a 200A panel on the house side of the Gateway which acts as an ATS. My thinking was to move it inside my ATS also. That would eliminate your concern. 200A panels are cheap.

That's correct, otherwise AC coupling would be impossible. You could even use the same Siemens panel I used that has an ATS built into it.

The quote I have from Tesla pairs my 2 Sunny Boy systems with each of one a 200A panels and 1 Powerwall 2 each. They said this was done to avoid the 120% panel current limit rule. They left my Enphase M215 system outside the Gateway so that it isn't used for charging the PW2's. It only acts as a Grid Tie system and backfeeds Grid power while PG&E power is present only.

You could always de-rate your main to get around 705.12(D)(2) (120% rule), just be careful that you don't de-rate below what you would normally use when no solar is available.

I have separate diagrams for my 2 systems. I have a hand done diagram of my entire system I sent to Tesla, I'll post next. I used to use a SMA RS485 monitor for my system which no longer works after 11 years. I only had one Sunny Boy failure in 11 years. So I stopped watching my system production years ago. Before my solar I had a $900+ per month PG&E bill. Now I get about $500/year check. So my system saved nearly $10k per year over 11 years it paid off during the third year. I occasionally walk out and look at the power production on the Sunny Boy panel. I checked it a few times last summer around 1PM or peak sun and say power in the 5.3-5.5KW range. Its possible it exceeds 11KW on the perfect day after I clean the panels but not by much.

I'll take a look at it in a few, my GF is telling me I need to get off the computer and spend time with her, so ...

If I choose to not use my Enphase based system and use less batteries such as 30KWHrs, I could use the curtail method if Sunny Boy 7000's support it. That would seem less "gross" than the shut it off for 5 minutes or more approach. I'm getting lost in details and need to get my diagram and go back and study yours.

Just FYI, when nearing full charge even the PWs use the same shut off and resume method.
Take a look at this video, He logged the status of the PW using the API

I left and looked at your diagram:

I see the inverter charger is connected to grid only power on your service entrance panel. How would the charger get solar power for a power outage longer than one day. It appears not possible for your inverter charger to recharge off solar production each day while in backup mode. Its AC input is grid only and not connected to your house loads panel where the solar power would appear each day while the grid power was down. It would appear the Sigineer Inverter acts as both an inverter with frequency control and as your charger supply. I will look at its capability. I see Sigineer appears to have a number of products that look like possible fits to this problem. How does the"BMS $2,495.00 EVTV 5/17/29" coordinate between the Sigineer Inverter/charger and the Tesla custom BMS's on each module?

I believe I answered some of the charging questions in a couple questions above, as for how to charge in a power our scenario, if you look at the diagram you'll see that the ATS upstream of the house and the sigineer defines what source is powering the home (grid or inverter) the backfeed breaker is on the same side as the house loads so solar will also move back and forth with the house.

Gotta go, I was just told a second time to get off the PC. let me know if that clarifies things.
 
I wasn't aware of there being an ability to "curtail" solar inverters using the frequency. Where is this specified and does the Sunny Boy 7000 implement this capability? I was aware of the ability to shut off any UL1741 inverter by raising or lowering the frequency above or below the limit. It seems almost abusive to use that capability to cycle the inverters every 5+ minutes plus charge time. I was a system designer and shutting down system power that frequently just seems like a rough way to limit charging.

I'm not sure if the 7k has it but here;s the official SMA doc with models that support it, page 7 has info on curtailment
https://files.sma.de/dl/7910/SB-OffGrid-TI-US-en-20.pdf

Here's the tech specs on Enphase (although only the IQ line can curtail) page 6
https://enphase.com/sites/default/f...Considerations-AC-Coupling-Micros-Battery.pdf

Another thing that bothers me is both my solar inverters produce 240V and not true split phase 120V. That suggests that home 120V appliances cannot really use solar inverter power since they draw power between neutral and one 120V phase which my inverters do not support. That suggests solar power cannot be used by the home except real 240V appliances. The home would only be able to use true split phase 120V/120V to neutral and the system inverter must support split phase with flexible loading on either phase. Is my thinking correct? Tesla's PW2 inverters claim to produce flexible split phase with the load on either 120V phase.

Wait, your inverters output single phase 240? are you sure it's not split phase 240? I don't know anywhere in the US where single phase 240 is used, that's mostly overseas.
Page 38 in the link below shows the 2 configurations for the 7k, you have the bottom on with only N, L1 and Ground(PE)?
https://www.cedgreentecheast.com/files/SB5678 INSTALLATION(3).pdf


The charging problem: I don't see a charger power supply in your list and the one you specified in an earlier post is no longer available. Solar inverters are current sources. They won't naturally provide less than 100% of the available solar power less inefficiencies. They will increase their forcing voltage to try and force current into a grid load. If they cannot Force 100% of their current onto a load, they will increase their voltage until they shut down due to overvoltage. That is not good when one is trying to limit both the maximum charge current into the batteries and stop charging one the battery is full.

The inverter I'm using is made by Sigineer
12000 Watt Inverter | Pure Sine Wave 48V 12KW Inverter Charger
It's an inverter charger so it has a charger built in that lets you charge the batteries off the grid. However, in an AC coupled system the built in charger isn't used as much, the reason for this is because the grid tied inverters are connected on the output side of the battery inverter. When the battery inverter switches its internal ATS from grid to battery backup it controls the frequency since its internals are generating the sinewave. After 300 seconds (per UL) the grid tie inverters will have synced to the sinewave of the battery inverter and will start to produce power. This power will start to power the loads and start to lighten the load on the battery based inverter, if you are producing more power than the loads require the power will flow backwards through the H bridge on the inverter, will be converted to 48v and it'll be stored in the batteries, if the batteries are nearing full the battery inverter will start to increase the frequency causing the grid tied inverters (which at this point are synced to the battery inverter not the grid) to start to lower their output, when the batteries fill the battery inverter will simply shut off the GT inverters by rising the frequency to 62.5hz or above. The GT inverters will stay off until the frequency drops again.

On my inverter (Sigineer) where there is no gradual shift of frequency only trip, I have my controller configured to listen to the EVTV controller which is sending out messages on the battery state, when a cell reaches 4.1v I tell the Sigineer to trip and turn off the GT inverters. Every 2.5 minutes it continues to trip for 200ms (UL requires a 5 minute timeout between trips before GT inverters can continue to output) this ensures that the GT stay off until the battery drops by 10% (threshold configured by me) at which point my controller stops triping the frequency and the GT inverters start to produce, when a cell reaches 4.1v again the cycle repeats, in reality the time window depending on what loads I have running is about 1.5-2 hours.


So I'm not sure how your actual charging system works. An off the shelf power supply doesn't provide the constant current needed for Lithium batteries plus must be stopped at the maximum voltage and switched to a voltage source and then stopped. What capabilities do the Tesla BMS's offer? Does Tesla provide specs on that? I see "BMS $2,495.00 EVTV 5/17/29" but I don't understand how one goes from raw AC to a proper Lithium charger with the appropriate current limit, overvoltage limiting, undervoltage cutoff, low temp cutoff etc being done by a coordination of a powersupply and the Tesla BMS, EVTV and solar inverter curtail or 60HZ 5 minute shutdown cycles. I need to look at it closer.

The EVTV BMS controller has contactors built into it. It also communicates with the BMB boards inside each module, it gets voltages of every cell plus the temperature for each thermistor. I can configure OV, UV, OT all within the EVTV controller. When one of those limits is reached the contactors inside the controller open.

View attachment 494215


Doesn't it mean the maximum power output of my solar inverters at say 11KW must not EVER exceed the maximum charger current of the battery pack ( or there must be a way of throttling the inverters such as the frequency control you mentioned). But my Sunny Boy solar inverters were producing 6.6KW yesterday afternoon which was a sunny day just past the lowest sun day of the year. So wouldn't the battery pack size be set roughly at my 11KW or 11KW+4KW= 15KW per day full output since that's the most KWHRs I can charge in a day. That suggests my battery size is about the size of 60KHWHrs of a full Tesla battery pack since I can produce about 70KWHrs on a sunny summer day.

You're partially correct on the charging side of things, if you remember the table I posted on my last post it had my calculations on it, broken down to this:

14,500kW of solar / 240v = 60.41A, that's the amperage that will be sent to the inverter at no load and full solar (worst case scenario), if we step that down to the 48v side of things (14,500kW / 48) 302.08A will be sent to the pack to charge the batteries. Each module can take a peak of 8kW (360A) and a continuous 5kW (225A) of charging amps. If i had a single module that would present an issue as a spike in solar productiong lasting longer than 10 min (8kW peak) would cause damage, but I have 14 modules, this means that the 302A is actually 320/14 = 21.57A, or in other words, I'm charging each module 52% below the max charging amps. In other words the battery output/input capacity is a lot higher than what my solar can produce so I don't need any kind of charge controlers to lower the amperage being sent to the battery.

Now that being said, near the top end 4v+ side of things throwing that much power will not get you a nicely balanced pack, this is why I have my controller set to disable the GT inverters when a cell reaches 4.1v so the pack can balance itself without reaching its limits.

The second part, you don't always have to match your battery size to your dialy avg output. For around spring I have days when I produce +100kWh but my pack is only 60kWh. when that occurs my controller knows if the grid is available or not if it is, instead of triping the frequency and stopping solar production it tells the sigineer inverter to switch to grid mode (essentially stand by mode in a traditional ATS) this directly ties my GT inverters to my power company grid and sells any excess back after the pack is full.

I need to measure how much power I can get away with over night to see what my lower limit is. The PG&E power shut offs this year were 3 days each and we rarely have cloudy days in August thru November. That would set my lower battery size that will "get me by". Since I can only produce about 6-7KW peak in November, that would suggest a battery size in the 30-40KWHrs size since a larger battery would have little value in November or during winter storms when the power outages are most likely. One can buy 5KW Tesla modules for about $1000 on EBAY which would require about 6 or so for $6000. Two Tesla PW2's would be 28KWHrs with 10KW continuous power output.

Yes, having a good idea of power usage is a must here. Oversizing is always fine, under is a bad idea.

The Tesla PW2 design moved my 2 Sunny Boy's that are line side fed to a 200A panel on the house side of the Gateway which acts as an ATS. My thinking was to move it inside my ATS also. That would eliminate your concern. 200A panels are cheap.

That's correct, otherwise AC coupling would be impossible. You could even use the same Siemens panel I used that has an ATS built into it.

The quote I have from Tesla pairs my 2 Sunny Boy systems with each of one a 200A panels and 1 Powerwall 2 each. They said this was done to avoid the 120% panel current limit rule. They left my Enphase M215 system outside the Gateway so that it isn't used for charging the PW2's. It only acts as a Grid Tie system and backfeeds Grid power while PG&E power is present only.

You could always de-rate your main to get around 705.12(D)(2) (120% rule), just be careful that you don't de-rate below what you would normally use when no solar is available.

I have separate diagrams for my 2 systems. I have a hand done diagram of my entire system I sent to Tesla, I'll post next. I used to use a SMA RS485 monitor for my system which no longer works after 11 years. I only had one Sunny Boy failure in 11 years. So I stopped watching my system production years ago. Before my solar I had a $900+ per month PG&E bill. Now I get about $500/year check. So my system saved nearly $10k per year over 11 years it paid off during the third year. I occasionally walk out and look at the power production on the Sunny Boy panel. I checked it a few times last summer around 1PM or peak sun and say power in the 5.3-5.5KW range. Its possible it exceeds 11KW on the perfect day after I clean the panels but not by much.

I'll take a look at it in a few, my GF is telling me I need to get off the computer and spend time with her, so ...

If I choose to not use my Enphase based system and use less batteries such as 30KWHrs, I could use the curtail method if Sunny Boy 7000's support it. That would seem less "gross" than the shut it off for 5 minutes or more approach. I'm getting lost in details and need to get my diagram and go back and study yours.

Just FYI, when nearing full charge even the PWs use the same shut off and resume method.
Take a look at this video, He logged the status of the PW using the API

I left and looked at your diagram:

I see the inverter charger is connected to grid only power on your service entrance panel. How would the charger get solar power for a power outage longer than one day. It appears not possible for your inverter charger to recharge off solar production each day while in backup mode. Its AC input is grid only and not connected to your house loads panel where the solar power would appear each day while the grid power was down. It would appear the Sigineer Inverter acts as both an inverter with frequency control and as your charger supply. I will look at its capability. I see Sigineer appears to have a number of products that look like possible fits to this problem. How does the"BMS $2,495.00 EVTV 5/17/29" coordinate between the Sigineer Inverter/charger and the Tesla custom BMS's on each module?

I believe I answered some of the charging questions in a couple questions above, as for how to charge in a power our scenario, if you look at the diagram you'll see that the ATS upstream of the house and the sigineer defines what source is powering the home (grid or inverter) the backfeed breaker is on the same side as the house loads so solar will also move back and forth with the house.

Gotta go, I was just told a second time to get off the PC. let me know if that clarifies things.

Thanks for helping me work up to a full understanding. I am an electrical engineer and computer science graduate who spent my career in Telecom circuit, software, system and finally did 2 Telecom fiber/local Telco access startup company engineering and strategy. It’s been 25 years since I actually designed anything of significance except putting together my 2 solar systems and similar. I’m not an electrician but understand the theory but only had DIY experience and only studied specific codes to get by.

I think I’m getting closer to understanding your design. I’m trying here to restate my current understanding with some questions where I’m still not sure. I hope this makes sense. Your descriptions and answers are great. I’m trying to piece it all together and decide my application desires. I believe it’s pretty darn close and except Tesla’s APP control features, its pretty good and well thought out.

Is your original wiring diagram still accurate?

I understand you installed the manual transfer switch for maintenance and testing disconnect purposes. Does it normally operate in the shown position with the Sigineer 12KW inverter output connected to the House Load Panel and the Grid connected to the Sigineer Inverter AC input. That means your home normally runs on the 12KW inverter and the Sigineer Inverter provides the ATS function, switching between its AC input and DC battery input for power. Is that true?

The Solar Inverters are current source inverters where 100% of their output power at any time must be sunk or they will go into a fault. That 100% solar output power must be sunk into two possible places. It must provide power to the home and any excess must be sunk into the battery charger and used to charge the batteries. Since both the home power and solar power output vary all the time during the daylight hours, the Sigineer charger must be able to use the excess of solar power minus the home power and do so in a constantly varying manner. It must draw this power from its AC output alongside its own AC output since the grid AC input is not present and it’s in backup mode. The maximum power it will ever charge the battery with is 100% solar (in my case 15.5KW) when the house is near zero.

When the grid is on and present, the Sigineer 12KW inverter acts like an AC repeater selecting its grid AC input and providing up to 12KW to the home loads. How do you request the Sigineer inverter switch back to grid so that excess solar power can be sold back to the grid in the event there is grid and the battery is full and there is excess solar production that can be sold back to the grid.

You mention “your controller”. What do you use as a controller? I don’t see it in your wiring diagram. Do you have list of its functions not provided by the Sigineer and the BMS $2,495.00 EVTV 5/17/29 from EVTV. Do you have a final wiring diagram updated with any additions after your initial diagram?

Does your system provide any time shifting or arbitrage by charging the batteries from the grid during lowest tariff rate times or from solar by time of day period and discharging into the grid at night or when the tariffs payback is higher? If not, have you thought about it might be done?

Does each Tesla of the 14 Tesla 5KW modules provide a module level BMS board? Then the BMS $2,495.00 EVTV 5/17/29 from EVTV acts as the “car” level BMS manager and charge discharge monitoring that understands the module interface and coordinates BMS functions for the entire 14 module pack? It provides the high voltage (e.g. 4.15), low voltage (e.g. 3.1), charge cutoff (e.g. 3.1), charge resume ( e.g. 4.1) and high low temperature limits. How and when is cell balancing done? Do you implement the Tesla liquid cooling/heating or is that really needed for this application? Does the graphical picture you show for setting parameters come from its touch screen display monitor? Does it provide a disconnect relay for both charging and load disconnect when the various parameters are exceeded and charging needs to be stopped or load disconnected to protect the batteries.

I’m impressed? Did you get any help from Sigineer or EVTV?

I hope your GF isn’t too mad at me or your computer. You did a nice job!
 
I watched a few of Jack's videos and studied the Sigineer 12/15 KW and now have a much better understanding of most everything. I've answered most of my questions from the earlier post. I've determined both my SMA and Enphase inverters are 120V/120V split phase and yes they both support L1, L2, N and ground. I see now how Jack and EVTV have made a business out of this basic architecture and have built it for their shop. I see they implemented the BMS to operate Tesla batteries and provide everything including the batteries and resell the 15KW Sigineer inverter. I also see they are pressing Sigineer to implement the 60HZ frequency controlled GTI cutback feature now in UL1741. Mine is circa 2007 and doesn't support it. New SMA SB 7000US GTI do support it along with the new Enphase series due the UL1741 requirement.

I also see how the Sigineer inverter as a voltage source will produce whatever power is needed or sink whatever power is required to the charger from batteries to compensate for the fact that the GTIs and home power continuously vary. This cannot work once the charger has filled the batteries and excess GTI power has nowhere to go. That's when the frequency shift is signaled by the BMS until it has room.

One last question. Why the architecture which prevents any arbitrage or GTI power shifting? I suspect you will say your batteries are so large and solar is adequate to not need to do this. Your home appears limited to the 12KW from the Sigineer. Isn't this an issue with an EV?

thanks again
 
My SMA SB7000US has a neutral. I had developed this concern regarding the Enphase M215 but the manual suggests split phase with neutral. If it was true that the M215 or other GTIs did not support 240 split phase 120V/N/120V, it would not be possible for Powerwall applications to allow the GTI inverters to operate the home. PWs could use them to charge but the GTI inverters could not operate a home 120V appliance. So what is the real answer? It almost has to be that they support split phase or I would have heard about it in my PW Google investigations. I will look again.

upload_2019-12-31_13-50-10.png


upload_2019-12-31_13-52-34.png
 
My SMA SB7000US has a neutral. I had developed this concern regarding the Enphase M215 but the manual suggests split phase with neutral. If it was true that the M215 or other GTIs did not support 240 split phase 120V/N/120V, it would not be possible for Powerwall applications to allow the GTI inverters to operate the home. PWs could use them to charge but the GTI inverters could not operate a home 120V appliance. So what is the real answer? It almost has to be that they support split phase or I would have heard about it in my PW Google investigations. I will look again.
The Powerwalls will balance the 120V on split phase when running isolated from the grid, even if the generation from grid tied inverters has no neutral or carries no current on the neutral. The Enphase documentation clearly says the output is 208/240V. The Neutral might only be used for PLC communication back to the Envoy and grid parameter reference.
 
The Powerwalls will balance the 120V on split phase when running isolated from the grid, even if the generation from grid tied inverters has no neutral or carries no current on the neutral. The Enphase documentation clearly says the output is 208/240V. The Neutral might only be used for PLC communication back to the Envoy and grid parameter reference.

This makes sense. A few years ago I was thinking of using an extra M215 spare I had in a battery backup test configuration until I ran into this issue. This brings up something I've been trying to sort out with the Sigineer 12/15K inverters that L-P-G and Jack Richards uses. Its difficult to know exactly how it functions in a couple of key ways without actually testing it. Maybe someone here knows the answers.

The arrangements both J.P.G. and Jack Richards uses are not the same as Tesla's PW2. The PW2 gateway has a relay that acts as a disconnect from the grid when it fails. But with the grid operating, the PW2 inverters act like GT inverters as current sources and can feed current into the grid from the battery and even be charging from solar GT inverters at the same time. They can even convert the 240V single phase from the Enphase inverters into split phase probably by charging from the Enphase GT inverters and simultaneously producing split phase power for the home or grid from the charging battery. That makes it a very flexible arrangement allowing mixed or even simultaneous:

+ Home using battery, solar GT and PW2 inverter power
+ Grid receiving power from the PW2 inverters from the battery and or Solar GT inverters
+ PW2 charging from grid power, solar GT inverter power and doing 240V single phase to split phase all at once

When the gateway disconnects the house, The PW2 inverters become pseudo masters rather than grid slaves providing their own 60Hz yet cross synchronizing with one another. When in backup during the day, they are able to perform the balancing act between the ever varying solar GT inverters that insist on outputting their maximum solar power as current sources while at the same time making it possible to power the ever varying home loads. If the solar GT inverters exceed the home load, the PW2s use the difference to charge the batteries. If the solar GT inverters fall below the home load, the PW2 inverters make up the difference from the batteries. Should the batteries fill and the solar GT inverters exceed the home load, the PW2's use the 60Hz upshift to shut down the GT inverters until there is room in the batteries. I'm told the GT inverters are allowed to restart somewhere in the 95-99% battery full range depending on the drain down rate.

That is a very flexible arrangement.

The Sigineer 12/16K inverters are nice but its not clear they can mimic these capabilities. Both Jack and J-P-G's architecture uses the Sigineer internal transfer switch to isolate the Sigineer AC output, home loads and solar GT inverters from the grid. I believe this means neither the solar GT nor battery power can be sent to the grid. That means it really only acts in the PW2 backup mode. It also means the home power including any EV charging must not exceed the Sigineer 12KW or 16KW. This might be just fine but the architecture only allows the grid to charge the batteries or run the home in the event the solar GTs + Batteries have no power. No arbitrage or time shifting is allowed by the architecture. It would be possible to add a time of day relay that prioritizes the system to grid charge during off-peak. The system does not need NET metering since their is no way to put power back into the grid. It also isolates the home system from any attempt by the utility to use its 60Hz UL1741 GT cutback control feature recently added.

Whether its possible to use the Sigineer 12/16K inverter in a more PW2 "like" architecture, depends on its capabilities. Can it switch from a UL1741 "like" grid synchronizing and voltage following inverter when the grid is connected via a more traditional ATS - and be able to manage flexibly its charging and output by time to do arbitrage, solar time shifting - all while being on the grid?

It already appears by Jack and J-P-G's success that it can manage GT inverters including Enphase without split phase and balance the varying GT inverter power and varying home load through siphoning charging power or producing battery power. It can shut down the GT inverters when the battery is full via 60Hz upshift. The data sheet doesn't explain much about charger behavior besides its 4 phase Lithium charge profile. But what about inrush, premature stoppage, and does it accept charge from its AC out when running on batteries. By producing AC power and simultaneously allowing charging would allow the use of GT inverter power to charge plus allow the 240V to split phase conversion via charging and discharging. I assume this works or Jack and J-P-G would have said something.

Does any of this make sense?
 
But with the grid operating, the PW2 inverters act like GT inverters as current sources and can feed current into the grid from the battery and even be charging from solar GT inverters at the same time.

No, the PW inverter cannot both sink and source current at the same time. It may appear that they do because they can switch between the two modes quickly, but only one or the other can occur at any instant in time.
 
No, the PW inverter cannot both sink and source current at the same time. It may appear that they do because they can switch between the two modes quickly, but only one or the other can occur at any instant in time.

So that implies in backup mode with no grid and the solar GT inverters are not producing enough power to run the house, the PW2 inverter must supplement them AND cannot ever charge its batteries while this case occurs. The only case where in backup that the PW2 might be able to charge its batteries is with the GT inverter power > home power. But this also implies the home must run directly off the GT inverters since the PW2 inverter must be off to charge. That implies the Enphase M215 inverters will not work if they don't support split phase AC 120V/N/120V because the PW2 inverter is shut off to allow the excess to charge the batteries? If the PW2 inverters switched on and off, the Enphase GT inverters could not power the home during the off/charging time. Something must be wrong with this picture as M series inverters were used for many years. This implies the Enphase GT M215 inverters MUST support split phase. The Sigineer based system will suffer from similar issues during grid loss.
 
Now take the Sigineer architecture where the grid is off and the S inverter switches to battery input. It would be running the home off its 12/16KW output. But in the case where there is excess solar GT power, how does the S inverter know it could let the home run off the GT inverters so it can charge the batteries. I don't see any mechanism that allows the S Inverter architecture to switch to charge when the batteries need power and the solar GT inverter power > home power AND how to switch back when the home power > solar GT inverters. What am I missing? This would imply the Sigineer architecture can only power the home till the batteries are discharged and never be able to recharge on a day by basis when AC grid is off and solar GT power > home power use? Somebody explain what I'm missing!
 
I've installed the Outback Power Skybox.
It has a 5kW inverter which can run on battery, generator or PV (DC). It will also utilize AC PV and use frequency shift to control the AC PV inverters. You can stack these for more power.
http://outbackpower.com/products/true-hybrid-energy-system/skybox
It can use many different types of batteries and will charge them from grid or PV (AC or DC).
Lots of flexibility.
I've installed with SLA batteries for now since I'm only using this for power outage backup since I don't expect to demand many cycles from the batteries and they they are much cheaper.
I've set it up to isolate critical loads totaling less than 5 kW from hot tub, etc.
 
SBut this also implies the home must run directly off the GT inverters since the PW2 inverter must be off to charge. That implies the Enphase M215 inverters will not work if they don't support split phase AC 120V/N/120V because the PW2 inverter is shut off to allow the excess to charge the batteries?
No. I'm not sure what architecture the PW2 uses to handle unbalanced split phase demand using a two-wire 240V source, but it can be easily done with a transformer. I believe that is the way the SolarEdge battery solution from a few years ago did/does it.

So your microgrid could have a two-wire 240V solar GTI producing, say, 10 amps (2.4 kW), and a household demand of 15 amps on just one 120V leg of the 120/240V split phase wiring (1.8 kW), and the PW2 can be operating in island mode and charging at 600W minus losses.

Cheers, Wayne
 
Alright lots of replies here, putting them all in a single post.

Is your original wiring diagram still accurate?

It's missing a contactor on the grid input side of the Sigineer, I use it to be able to force the Sigineer into offgrid mode. Since the sigineer doesn't have an API I can use I can force its state by either giving a connection to the grid or cutting it off. If i tell the contactor to open it'll take the house off the grid (regardless of the grid being available) this allows me to charge the pack solely off the excess solar. When the pack is full, my controller tells the contactor to close so any excess solar can be sold back

I understand you installed the manual transfer switch for maintenance and testing disconnect purposes. Does it normally operate in the shown position with the Sigineer 12KW inverter output connected to the House Load Panel and the Grid connected to the Sigineer Inverter AC input. That means your home normally runs on the 12KW inverter and the Sigineer Inverter provides the ATS function, switching between its AC input and DC battery input for power. Is that true?

Yes, that puts the Sigineer into a whole house UPS mode. The internal ATS on the Sigineer can pass 19kW, then the grid goes down and power is being generated from the batteries then it goes down to 12kW

The Solar Inverters are current source inverters where 100% of their output power at any time must be sunk or they will go into a fault. That 100% solar output power must be sunk into two possible places. It must provide power to the home and any excess must be sunk into the battery charger and used to charge the batteries. Since both the home power and solar power output vary all the time during the daylight hours, the Sigineer charger must be able to use the excess of solar power minus the home power and do so in a constantly varying manner. It must draw this power from its AC output alongside its own AC output since the grid AC input is not present and it’s in backup mode. The maximum power it will ever charge the battery with is 100% solar (in my case 15.5KW) when the house is near zero.

Then you'll probably want the 15kW version of the Sigineer (if you plan on using Sigineer). My 12kW can handle about 5 min of 14kW before it starts to complain, it'll play an audible alarm after the 5 min. 60 seconds after that it'll shut down due to overload which shuts off the GT inverters. I've only had that happen once in a perfect sunny day with zero clouds and it was cold so the house AC wasn't running.

When the grid is on and present, the Sigineer 12KW inverter acts like an AC repeater selecting its grid AC input and providing up to 12KW to the home loads. How do you request the Sigineer inverter switch back to grid so that excess solar power can be sold back to the grid in the event there is grid and the battery is full and there is excess solar production that can be sold back to the grid.

No, When the grid is present the Sigineer has an internal ATS and it just bridges the output wiring to the input grid. The inverter circuitry is off in this mode. When the grid fails then it switches the ATS (this islands the house from the grid per UL) and starts to use its internal circuitry to generate power from the batteries. As to how do I request the mode on the Sigineer, I have a contactor on the grid input of the Sigineer see details on the answer before this one.

You mention “your controller”. What do you use as a controller? I don’t see it in your wiring diagram. Do you have list of its functions not provided by the Sigineer and the BMS $2,495.00 EVTV 5/17/29 from EVTV. Do you have a final wiring diagram updated with any additions after your initial diagram?

The EVTV controller has a lot of functions built in (like freq trip, contactor enable/disable, grid enable disable) but it wasn't flexible enough for my liking so I use a RPi that listens to the UDP packets sent out by the EVTV controller over wifi. These packets contain all the stats on the battery, I then wrote some code that decides things like when to charge using the built in charger in the Sigineer inverter (for example when the battery is low but it's night time and I get an alert of a storm coming I charge off the grid), When to trip the freq on the Sigineer (when in offgrid mode I can more accurately define if I want to turn on more loads like the AC using the thermostat API instead of shutting off the solar) I also use it when something on the EVTV controller doesn't work exactly as I want it (For example the EVTV controller lets you specify a cell voltage at which to trip the freq, and it'll keep tripping the freq, I wanted to change that window to a longer span and not cycle the system as often my controller listens to the SOC from the EVTV UDP packages and when it drops to set % act accordingly)

Does your system provide any time shifting or arbitrage by charging the batteries from the grid during lowest tariff rate times or from solar by time of day period and discharging into the grid at night or when the tariffs payback is higher? If not, have you thought about it might be done?

My system can because I modified the Sigineer so I can remotely enable the built in charger to use the grid to charge. My RPi controller can enable or disable the charger at any time I want. (Not that it matters there is no TOU in FL)

Does each Tesla of the 14 Tesla 5KW modules provide a module level BMS board? Yes, they're called BMB Then the BMS $2,495.00 EVTV 5/17/29 from EVTV acts as the “car” level BMS manager and charge discharge monitoring that understands the module interface and coordinates BMS functions for the entire 14 module pack? Yes It provides the high voltage (e.g. 4.15), low voltage (e.g. 3.1), charge cutoff (e.g. 3.1), charge resume ( e.g. 4.1) and high low temperature limits? Yes How and when is cell balancing done? The EVTV controller sends a balance command to the BMB boards and they do the balancing using the logic built by Telsa, When is it done I'm not sure as that is Jacks code which isn't open source. Do you implement the Tesla liquid cooling/heating or is that really needed for this application? No, the high temp ended up coming from a badly crimped cable and FL doesn't freeze so no need to heat the modules. The packs doesn't get hot enough due to usage to need cooling, environmental circumstances is another story though Does the graphical picture you show for setting parameters come from its touch screen display monitor? The one I replied to you with is a screenshot of the remote screen you can buy with the EVTV controller, the other screenshot that's more of a dashboard a few posts back is written by me and comes out of my controller. Does it provide a disconnect relay for both charging and load disconnect when the various parameters are exceeded and charging needs to be stopped or load disconnected to protect the batteries. Yes, it has contactors on both + and - plus a pre-charge resiststor.

I’m impressed? Did you get any help from Sigineer or EVTV?
No, I did a ton of research though

I hope your GF isn’t too mad at me or your computer. You did a nice job!
She's used to it, lol. Thanks.

One last question. Why the architecture which prevents any arbitrage or GTI power shifting? I suspect you will say your batteries are so large and solar is adequate to not need to do this. Your home appears limited to the 12KW from the Sigineer. Isn't this an issue with an EV?
thanks again

The Siginner doesn't have the ability to raise the freq in steps to make the GTs lower the output, it can only turn them off all together so I can't really curtail. Other inverters like the Outback Radian, SMA Sunny Island, etc do have this feature so they can lower the GT output to in turn lower the power being backfed into the battery. But yes, my pack is large enough to not need to limit it

I don't think you are correct about the solar inverters. I don't believe they have any Neutral connection, so are therefore pure 240V devices.

The SMA do, they can run on 240v or 120v. When running on 120v you have to have a N. The Enphase micros only do 240v so they don't have a N

The Powerwalls will balance the 120V on split phase when running isolated from the grid, even if the generation from grid tied inverters has no neutral or carries no current on the neutral. The Enphase documentation clearly says the output is 208/240V. The Neutral might only be used for PLC communication back to the Envoy and grid parameter reference.

This makes sense. A few years ago I was thinking of using an extra M215 spare I had in a battery backup test configuration until I ran into this issue. This brings up something I've been trying to sort out with the Sigineer 12/15K inverters that L-P-G and Jack Richards uses. Its difficult to know exactly how it functions in a couple of key ways without actually testing it. Maybe someone here knows the answers.

The arrangements both J.P.G. and Jack Richards uses are not the same as Tesla's PW2. The PW2 gateway has a relay that acts as a disconnect from the grid when it fails. But with the grid operating, the PW2 inverters act like GT inverters as current sources and can feed current into the grid from the battery and even be charging from solar GT inverters at the same time. They can even convert the 240V single phase from the Enphase inverters into split phase probably by charging from the Enphase GT inverters and simultaneously producing split phase power for the home or grid from the charging battery.

This is not correct, when Tesla installs your PWs they can either charge from the grid (installs without solar) or they can charge from the solar (installs with solar) they won't do both. They also don't allow you to push back stored power to the grid, stored power can only be used by your loads. I verified this with my rep and their engineering dept when I was debating getting PWs, if that has changed it's news to me but I highly doubt it as it would open a loop hole allowing home owners to charge at low rates and sell back at peak which I doubt many POCOs would allow.

When the gateway disconnects the house, The PW2 inverters become pseudo masters rather than grid slaves providing their own 60Hz yet cross synchronizing with one another. When in backup during the day, they are able to perform the balancing act between the ever varying solar GT inverters that insist on outputting their maximum solar power as current sources while at the same time making it possible to power the ever varying home loads. If the solar GT inverters exceed the home load, the PW2s use the difference to charge the batteries. If the solar GT inverters fall below the home load, the PW2 inverters make up the difference from the batteries. Should the batteries fill and the solar GT inverters exceed the home load, the PW2's use the 60Hz upshift to shut down the GT inverters until there is room in the batteries. I'm told the GT inverters are allowed to restart somewhere in the 95-99% battery full range depending on the drain down rate.

That is a very flexible arrangement.

The Sigineer 12/16K inverters are nice but its not clear they can mimic these capabilities. Both Jack and J-P-G's architecture uses the Sigineer internal transfer switch to isolate the Sigineer AC output, home loads and solar GT inverters from the grid. I believe this means neither the solar GT nor battery power can be sent to the grid. Battery power cannot, Solar GT can be sent back to be sold. That means it really only acts in the PW2 backup mode. It also means the home power including any EV charging must not exceed the Sigineer 12KW or 16KW. Only when the grid is not available, when the grid is available and the Sigineer ATS is in grid mode it can pass 19kW This might be just fine but the architecture only allows the grid to charge the batteries or run the home in the event the solar GTs + Batteries have no power. No arbitrage or time shifting is allowed by the architecture. It would be possible to add a time of day relay that prioritizes the system to grid charge during off-peak. You can charge during off peak using the grid power by using the built in charge into the Sigineer The system does not need NET metering since their is no way to put power back into the grid. It also isolates the home system from any attempt by the utility to use its 60Hz UL1741 GT cutback control feature recently added. This is only true if you are 100% off grid and don't connect a grid input to the Sigineer, otherwise you do need net metering and the freq of the grid will be sent to the GTs (when the Sigineer is in grid mode) if you're in off grid mode then there is no connection from the grid to the GTs

Whether its possible to use the Sigineer 12/16K inverter in a more PW2 "like" architecture, depends on its capabilities. Can it switch from a UL1741 "like" grid synchronizing and voltage following inverter when the grid is connected via a more traditional ATS - and be able to manage flexibly its charging and output by time to do arbitrage, solar time shifting - all while being on the grid? Yes

It already appears by Jack and J-P-G's success that it can manage GT inverters including Enphase without split phase and balance the varying GT inverter power and varying home load through siphoning charging power or producing battery power. It can shut down the GT inverters when the battery is full via 60Hz upshift. The data sheet doesn't explain much about charger behavior besides its 4 phase Lithium charge profile. But what about inrush, premature stoppage, (This is more of an AC coupled drawback than an inverter drawback, however an inverter like the SMA Sunny Insland with curtailing can give you a much smoother charging profile) and does it accept charge from its AC out when running on batteries. This would just be a loop, the output AC is taken from the batteries so it'd be like filling up a container with a hose connected to itself, even worse the conversion losses would mean that you'd be wasting power, there is 2 ways to charge the batteries either by using the built in charger which gets power when the grid input is connected to the grid or when off grid mode the excess solar is pushed by the GTIs past the battery transformer and into the batteries By producing AC power and simultaneously allowing charging would allow the use of GT inverter power to charge plus allow the 240V to split phase conversion via charging and discharging. I assume this works or Jack and J-P-G would have said something.

Does any of this make sense?

So that implies in backup mode with no grid and the solar GT inverters are not producing enough power to run the house, the PW2 inverter must supplement them AND cannot ever charge its batteries while this case occurs. That's correct, but if you don't have the grid available and no solar, where is the power to charge the batteries going to come from? The only case where in backup that the PW2 might be able to charge its batteries is with the GT inverter power > home power. But this also implies the home must run directly off the GT inverters since the PW2 inverter must be off to charge. Yes and no, the house does need to be on the same side of the PW2 output as the GTIs but the PW2 inverter does not need to be off to charge, it charges off the excess solar being pushed into the transformer by the GTIs That implies the Enphase M215 inverters will not work if they don't support split phase AC 120V/N/120V because the PW2 inverter is shut off to allow the excess to charge the batteries? The 215s will work fine, you only need 120v for loads that can't handle 240v, the micros output 240v over L1/L2 and the PW2 transformer is that provides the middle leg for the N so you can grab L1/N or L2/N If the PW2 inverters switched on and off, the Enphase GT inverters could not power the home during the off/charging time. Something must be wrong with this picture as M series inverters were used for many years. This implies the Enphase GT M215 inverters MUST support split phase. The Micros and the PW2 inverter never shut off, they're symbiotic relationship. The PW2 inverter provides the sinewave and the GTIs provide the power, if the PW2 inverter turns off the micros won't have a sinewave to sync to and they'll shut off. If the micros shut off the PW2 won't recharge itself. Power is simply regulated by the PW2 by increasing the sinewave frequency The Sigineer based system will suffer from similar issues during grid loss.

Now take the Sigineer architecture where the grid is off and the S inverter switches to battery input. It would be running the home off its 12/16KW output. But in the case where there is excess solar GT power, how does the S inverter know it could let the home run off the GT inverters so it can charge the batteries. I don't see any mechanism that allows the S Inverter architecture to switch to charge when the batteries need power and the solar GT inverter power > home power AND how to switch back when the home power > solar GT inverters. What am I missing? This would imply the Sigineer architecture can only power the home till the batteries are discharged and never be able to recharge on a day by basis when AC grid is off and solar GT power > home power use? Somebody explain what I'm missing!

There is no special mechanism because the Sigineer has an H bridge. in other words picture a hose, on one side you have a water pump that can pump at 15 gallons per minute (this is the GTI) on the other side you have a a pump that senses the pressure of the line and increase the pressure as needed up to 12 gallons per minute(this is the Sigineer). When the hose springs a leak of 1 gpm (your house loads) the 12 gpm (Sigineer) will increase the pressure, then someone turns on the 15 gpm pump and this now takes care of the leak so the 12 gpm pump stops pumping but now you have 14 gpm that have to go somewhere so the 12 gpm pump flows backwards and lets 14 gpm flow into the overflow tank (batteries)
 
  • Informative
  • Love
Reactions: mspohr and Utahken
Were you able to assemble a block diagram or even a schematic of the Sigineer inverter so you could verify all these capabilities before building a system. A lot of these things are subtle and not well described by the data sheet. Or were you able to speak with them to develop your understanding of all the possible issues. Or did you build it on faith and discover them as you tested? Or something else. There are a lot of possible gotchas that it appears Sigineer thought out without documenting it publicly.
 
Were you able to assemble a block diagram or even a schematic of the Sigineer inverter so you could verify all these capabilities before building a system. A lot of these things are subtle and not well described by the data sheet. Or were you able to speak with them to develop your understanding of all the possible issues. Or did you build it on faith and discover them as you tested? Or something else. There are a lot of possible gotchas that it appears Sigineer thought out without documenting it publicly.
I seem to recall that Jack Rickard AC coupled solar into the Sigineer inverters and found that it would charge the battery when there was excess power. I don't think this was an expected result because it wasn't documented. I also seem to recall that it wasn't intentionally engineered that way either.
 
  • Informative
  • Like
Reactions: mspohr and Utahken
I thought I would share some info on the Orion BMS by Ewert Energy Systems because it provides some insight into the Tesla Model S battery system and its BMS monitoring boards. I was not aware the per module boards only provide monitoring of voltage and temperature and cell taps for balancing.

Ewert Energy Systems
Orion BMS 2 | Orion Li-Ion Battery Management System
Application Notes | Orion Li-Ion Battery Management System

Tesla Model S & X Battery Module BMS Cell Tap Board


Orion 48S Cell BMS | Battery Management Systems | Stealth EV

Tesla Model S & X Battery Module BMS Cell Tap Board | Stealth EV

upload_2020-1-3_9-34-10.png
 

Attachments

  • Orion BMS Operation.pdf
    352.7 KB · Views: 40
  • orionbms2_specifications.pdf
    683.3 KB · Views: 52
  • tesla_modules.pdf
    557.5 KB · Views: 79
Last edited:
I seem to recall that Jack Rickard AC coupled solar into the Sigineer inverters and found that it would charge the battery when there was excess power. I don't think this was an expected result because it wasn't documented. I also seem to recall that it wasn't intentionally engineered that way either.

Sigineer provides a diagram for the "Charger mode" and "Discharger mode". Charging the batteries from AC from the GT inverters would imply it actually behaves in a mixed mode? In the case of backup when AC has failed and Main Relay 01 is opened, the Sigineer inverter provides AC derived from the batteries on the AC output which turns on the GT inverters ( UL1741 apparent grid AC on) and AC from solar GT inverters would be summed on the AC output. This diagram must be dramatically over simplified. There must be a completely separate charger not shown that is between the CT winding and the batteries. It sure would be nice if the provided more details and a proper block diagram.


upload_2020-1-3_10-25-17.png
 
Sigineer provides a diagram for the "Charger mode" and "Discharger mode". Charging the batteries from AC from the GT inverters would imply it actually behaves in a mixed mode? In the case of backup when AC has failed and Main Relay 01 is opened, the Sigineer inverter provides AC derived from the batteries on the AC output which turns on the GT inverters ( UL1741 apparent grid AC on) and AC from solar GT inverters would be summed on the AC output. This diagram must be dramatically over simplified. There must be a completely separate charger not shown that is between the CT winding and the batteries. It sure would be nice if the provided more details and a proper block diagram.


View attachment 496065
I think Jack deduced from the diagram that it would work either way.
He tested it and it worked.
You can see that the H circuit can run power into or out of the batteries.