diy solar

diy solar

EG4 new AIO rated 12K output and 18kPV aka "EG4-18Kpv-12LV"

go to 1:11:38. Again, this is the Fortress Envy which appears to be spec on spec to the EG4 18K PV.
OK, thanks for sharing that Henderson. it seems more complicated than it could be. we are in talks with them regarding the application in that video. but now, you may please neglect that, and refer to what I replied above. Thx
 
You are correct, I hadn't realized this point. Thx
Not entirely true. Look at my signature. You can use 3 EG4-6500 effectively. I was going to do split phase x 2 (4 eg4's required) but after running 2 of them, not necessary to do split phase x2. Even though I have 4 Eg4 inverters, #3 has become a 100amp battery charger (low sun or faster charging when desired). Inverter#4 has become a spare if necessary. 2 EG4's in split phase runs my entire house and has a ton of spare power. Most of the time power usage is less than 20%. Running freeze dryer and 3.5T A/C today with power draw at only 40% max.
 
Last edited:
I've heard this name Adam De lay in a couple of threads recently, who is he, and why is his opinion important?
He has a methodical approach to a lot of the technical aspects of inverters. Especially the software and it’s functionality. And that can be extremely helpful to people. That’s likely why you’ve heard his name come up.
 
The new EG4-18k looks very promising, but we need more testing to see if it can deliver.
We need to see if it has similar problems as the SolArk in certain applications:

1. AC coupling is disabled for now in off-grid mode ( according to the manual)
this effectively means it is not working!

2. Will AC coupling work with stacked units in off-grid or backup mode?

3. How about AC coupling with legacy inverters, that do not support FW, VW or VV, is there a PV disconnect relay for AC coupled inverters in off-grid mode and for compliance with EPO?

4. Need load imbalance testing at full imbalance, SolArk has problems with max imbalance.
( see Dean Abney's video at: SolArk Testing)

5. Is the MID relay rated for 200 A?
Is the relay the top-of-the-line 200A Gruner relay as in a Tesla Powerwall MID panel or as in a FranklinWH aGate or it is a chinese knock-off?

 
Last edited:
The new EG4-18k looks very promising, but we need more testing to see if it can deliver.
We need to see if it has similar problems as the SolArk in certain applications:


5. Is the MID relay rated for 200 A?
Is the relay the top-of-the-line 200A Gruner relay as in a Tesla Powerwall MID panel or as in a FranklinWH aGate or it is a chinese knock-off?
I took some pictures of the internals of my LP12k.
Unfortunately there is a sticker on top of the relays so I can not make out brand/model
There is another relay that is 90 Amp rated that does have a manufacturer/model number
churod char-112A90C
*update*
Looking at additional photos I took, the relays with the sticker on it are the same : 90 amp rated.
Probably the AC passthrough relays are located closed to the hookup of the grid and are not visible on my photos
IMG_20221021_180809620_HDR.jpg
 
Last edited:
The new EG4-18k looks very promising, but we need more testing to see if it can deliver.
We need to see if it has similar problems as the SolArk in certain applications:

1. AC coupling is disabled for now in off-grid mode ( according to the manual)
this effectively means it is not working!

2. Will AC coupling work with stacked units in off-grid or backup mode?

3. How about AC coupling with legacy inverters, that do not support FW, VW or VV, is there a PV disconnect relay for AC coupled inverters in off-grid mode and for compliance with EPO?

4. Need load imbalance testing at full imbalance, SolArk has problems with max imbalance.
( see Dean Abney's video at: SolArk Testing)

5. Is the MID relay rated for 200 A?
Is the relay the top-of-the-line 200A Gruner relay as in a Tesla Powerwall MID panel or as in a FranklinWH aGate or it is a chinese knock-off?

We're currently doing a multitude of edge-case testing for some of these scenarios that people are looking for. We have some videos in the works as well to demonstrate the results of our testing, so stay tuned for that!
To answer your questions with what I know,

1. AC Coupling works just fine in off-grid mode. It'll utilize the GEN port and the hybrid inverter will work as the power source instead of having the grid-interactive inverter look at the GRID port.

2. Yes.

3. The AC Coupling software is built around UL1741SA compliance, so the frequency shifting should raise to a point that will disconnect a legacy inverter but it is not something I've seen tested. The RSD on the inverter shuts down all of the ports, and you can wire in an initiator for remote operation if the inverter is indoors or it is needed for code, etc.

4. We've done a lot of imbalance testing with the 18kPV, getting up to about 8kW of imbalance. You can output up to 70A on a single leg, even with 0A on the other leg.

5. Yes, the relays are 200A to allow for the typical household to have full bypass capacity with just a single unit.
 
Thanks fromport for the pics and the relay info.

A 200A MID relay is much larger and the fine print reveals the details. Many physically smaller relays can carry/sustain a higher amp rating, like 90A, but have significantly lower make and/or break currents like 30A, see the manufacturer link.
From the AC relay board pic I can tell that the relays are not anywhere close to break or make 200A.

Braydan, please give us the 200A MID full relay specs and manufacturer datasheet link - break and make ratings - so we can verify.
 
This is a pity. I hope that this will be a feature one day. There are some markets that use split phase and 50hz in Latin American and the Caribbean, for example Jamaica where Sol-Ark has been doing very well.

Can someone from EG4 confirm that it is 60Hz only? I see the Fortress version can do nominal 50hz/60hz...
I know Eddie answered, but yes it does support 50hz. It will autodetect based on grid frequency. The spec sheet will be modified to clear up confusion!
 
Thanks a lot for the link!

The 112A200 relay is a 1-pole 200A 12V relay, with a 50A make/break rating there is no 220A max break! Where do you get the 220A max break rating from? If you break/make 200A with this relay it won't last very long! And it looks like the grid and backup terminals are not rated for 200A either! Additionally, you need 2 relays for the split-phase MID disconnect. And there is no feedback contact to verify that the relay is still operational and not stuck!

Schneider XW inverters have a long history and expertise on stacking inverters and the problems with relay arcing on parallel connected relays! 2 XW Pro's can be stacked, but if more than 2 are stacked an external 200A transfer MID switch is required, that's why they came out with the BCS 2200.
The Tesla Powerwall-2 MID panel - aka gateway - the FranklinWH aGate, the EP Cube MID panel and the Enphase IQ controller use a fully rated 200A MID latching relay, 200 A make/break, with an auxiliary contact for feedback, see pic for the top-of-the-line Gruner 200A MID relay.
The MID relay size is about 4" x 3" x 1.2", the plated copper bus bar in a full MID assembly is about 20mm x 3mm for 200A, and the lugs are 200A rated too, about 1" x 1" x 1.125".

Gruner 200A MID Relay: 4" x 3" x 1.2" - green wires are aux contacts for feedback
Gruner MID Relay.JPG
200A MID Switch assembly: plated bus bars are 20mm x 3mm and larger for 200A, 200A lugs are 1" x 1" x 1.2"
Generator 200A Switch.JPG

There is a lot more metal, copper and technology required to have a fully rated fail-safe 200A MID switch, and a mechanical overwrite will be useful too. We'll see what problems users run into if they start stacking 2 or more 18 KPV inverters and think they can switch 200A with a 50A rated make/break relay!

Having a single inverter UL listed is one thing, UL testing of stacked inverters with 200A MID rating is a completely different undertaking!
 
Last edited:
The 112A200 relay is a 1-pole 200A 12V relay, with a 50A make/break rating there is no 220A max break! Where do you get the 220A max break rating from? If you break/make 200A with this relay it won't last very long! And it looks like the grid and backup terminals are not rated for 200A either! Additionally, you need 2 relays for the split-phase MID disconnect. And there is no feedback contact to verify that the relay is still operational and not stuck!
May I suggest you spend a little more time on the PDF document linked to. It can handle 830 volts at a maximum 200 amp and is indeed rated to 220 amps. Would I trust it to do this more than half a dozen times, absolutely not. And there looks like there's quite a few of them in the inverter so I'm assuming there's one per phase.
You are right to be skeptical as this is no contactor, just a barely adequate relay.
 
He has a methodical approach to a lot of the technical aspects of inverters. Especially the software and it’s functionality. And that can be extremely helpful to people. That’s likely why you’ve heard his name come up.

I know Eddie answered, but yes it does support 50hz. It will autodetect based on grid frequency. The spec sheet will be modified to clear up confusion!
Awesome, thanks.
 
I looked at the pdf document and I have documents for top-of-the-line 200A make/break MID relays! It has a max switching current of 220A, but no documentation how many switching cycles at 220A! There is no 220A rated make or break current for the contacts!
Look at the contact rating: 50A make, 50A break, that is the rating that is important for reliable switching. The 220A max switching current is "smoke and mirrors", probably more smoke and ultimately a fire hazard! Sure, you can make/break 200A or 220A, but the contacts as you've said won't last very long. What happens if one relay for one leg is stuck but the other leg is still operational? If you time the make/break exactly to the 0-crossing/phase transition point it may handle many switching cycles, at phase-0 transition the current is 0! But no mechanical relay with a 30ms switch time can be exactly timed for a near zero-crossing transition that lasts less than 2 ms!

Some day we'll see solid-state zero-crossing relays for 200A that won't need active cooling to handle the 200A current.
 
Last edited:
Would I trust it to do this more than half a dozen times, absolutely not.
That was my first reaction as well. However, when I look closer at the data sheet, things seem more reasonable.

The spec sheet says
1684261063308.png

1684261699276.png

The 50A rating is at 830Vac. It is not specified, but at 240V it will be able to break higher currents and still meet the 30,000 electrical operations that are in the specification. However, since the sheet does not specify a 240V break current, we can't quantify an increase and we can only evaluate operation cycles at the 830V.

What is interesting is the Electrical Life chart. It gives operation count information for switching much higher currents.
1684258726031.png

At 50A the chart indicates almost 50,000 operations, but they only rate it for 30,000 operations for resistive loads. They gave themselves plenty of margin for variations from what they tested.

At 200A the unit has ~ 20000 operational cycles. However, the chance of it ever switching at 200A is pretty low. In most households, even the peak is closer to 100A or 30000+ operational cycles.

1684261905934.png
Even if we dived by 4 to cover inductive loads, it is 5k-7.5K operations. Since the switching event will normally only occur during a power outage, this seems reasonable.
 
I just made quote for a person with 600 amps main and a 200 amp essential load panel.
I wouldn't want to make a bet that with their new EV they might be able to pull 400 amps
I’m in the same position right now, putting together a proposal for a business that requires 400 amp service at its main structure (lots and lots of walk-in freezers and refrigeration, heavy AC loads, etc). I don’t want to spec only two units instead of three, and then get a call on a hot July day telling me that the system tripped.

So, we’ve got one distributor (Fortress) telling us that two units might pass through only 300 amps; another distributor (Signature) saying they “believe” it might be 400. A definite answer from someone who has actually load-tested this issue would be nice.
 
Since the switching event will normally only occur during a power outage
No. Unless I’m missing something, switching will occur every single time the load calls for more than the inverter’s rated output of 12KV, which triggers the need for AC pass through.
 
That was my first reaction as well. However, when I look closer at the data sheet, things seem more reasonable.

The spec sheet says
View attachment 149454

View attachment 149455

The 50A rating is at 830Vac. It is not specified, but at 240V it will be able to break higher currents and still meet the 30,000 electrical operations that are in the specification. However, since the sheet does not specify a 240V break current, we can't quantify an increase and we can only evaluate operation cycles at the 830V.

What is interesting is the Electrical Life chart. It gives operation count information for switching much higher currents.
View attachment 149442

At 50A the chart indicates almost 50,000 operations, but they only rate it for 30,000 operations for resistive loads. They gave themselves plenty of margin for variations from what they tested.

At 200A the unit has ~ 20000 operational cycles. However, the chance of it ever switching at 200A is pretty low. In most households, even the peak is closer to 100A or 30000+ operational cycles.

View attachment 149457
Even if we dived by 4 to cover inductive loads, it is 5k-7.5K operations. Since the switching event will normally only occur during a power outage, this seems reasonable.
I would still feel much more comfortable with the German made Gruner Relay.
50A Make and Break without a 240V rating is kind of a bit of guess work as to how it will perform when 100A or more is being connected or disconnected. One can only hope that they tested it through a lot of cycles to make sure the Relay contacts remain intact.
 
No. Unless I’m missing something, switching will occur every single time the load calls for more than the inverter’s rated output of 12KV, which triggers the need for AC pass through.
The inverter can back-feed and it has current transformers on the input. In normal operation, the grid will remain connected and the inverter will operate however it has been programmed to behave.
  • If it is programmed for zero back feed, it will ramp down output as the demand goes down....keeping backfeed at zero.
  • If the output load is greater than the inverter can provide, current from the grid will start flowing into the system and the loads will be serviced by power from both the grid and the inverter.
  • If the inverter is configured to allow back-feed and the load is less than the inverter can supply, the extra output will back-feed onto the grid.
  • If the grid goes down, the relays must be opened in order to for the inverter to safely continue to supply power to the house.
 
Back
Top