diy solar

diy solar

EG4 18kPV Q+A general thread

What's the minimum resistance that the RSD sense circuit in the 18Kpv will detect as a closed circuit? I'm using 75-100 feet of CAT6 cable (only one pair), and the total resistance is around 5 ohms, which makes the inverter think the RSD switch is open.

I'm currently working on using all four pairs to drop the resistance a lot, but if 1 ohm is right on the borderline then I probably want to add a small relay or SSR or something...

Thanks!
 
Just starting to commission (2) 18Kpv with (6) PowerProOutDoorWallMount batteries, and had a protocol question:

In the Inverter, the default Battery Type is set to 2: Lithium, but the default Lithium Brand is “2”.

@EG4_Jared recommended setting the Lithium Brand to “1”, but it feels like “0:EG4” would be the more logical choice.

I set the Brand to “1” as per Jared’s suggestion, and now I can see Battery 0 [ID=1] and Battery 2 [ID=4] on the monitoring website, so that’s better than the default.

What’s the difference between “1” and “0:EG4”?

[Yes, I also set the Primary battery [ID=1] RS485 protocol to P02-GRW so it won't be able to use RS485, and used a straight-through Ethernet cable, which doesn't unscramble the RS485 pins between the inverter and batteries, and I can see both connected batteries on the website, so that's all good.]

Thanks!

With the WallMount batteries on firmware version Z02T17 or newer, setting the CAN protocol to EG4/LUX and the lithium brand to ID No. 1 will allow for additional information to be displayed in the monitoring system.



1718289610105.png

1718289586343.png
 
With the WallMount batteries on firmware version Z02T17 or newer, setting the CAN protocol to EG4/LUX and the lithium brand to ID No. 1 will allow for additional information to be displayed in the monitoring system.
Yes, I have all that running, just wondering what the difference between "0:EG4" and "1" is. My logical brain wants to use "0:EG4", but maybe "1" is really "1:EG4_Multiple_Batteries" or something.
 
If I’m talking Modbus RS485 to two parallel 18Kpv inverters, do I need two RS485 interfaces (one for each inverter) or can I just connect to the master inverter with a single RS485 interface and talk to the two distinct Modbus addresses?

I’m already talking to the master, but hope to install and configure the slave today.
 
So I've got (2) 18Kpv and (6) PowerPro Outdoor Wallmount batteries. Last night when I hooked everything up, the cells were a couple of mv apart. This afternoon, after charging to 100%, they are showing:
MaxCellVolt_BMS 3624
MinCellVolt_BMS 3357
so a delta of 267 mv. Is that normal for charging/balancing, etc?
 
If I’m talking Modbus RS485 to two parallel 18Kpv inverters, do I need two RS485 interfaces (one for each inverter) or can I just connect to the master inverter with a single RS485 interface and talk to the two distinct Modbus addresses?

I’m already talking to the master, but hope to install and configure the slave today.
You need a second RS-485 interface for the second inverter. Well, _maybe_ you could connect the Meter485 ports together and set different Modbus addresses, but there's no clarity from EG4 on how that would work.
 
So I've got (2) 18Kpv and (6) PowerPro Outdoor Wallmount batteries. Last night when I hooked everything up, the cells were a couple of mv apart. This afternoon, after charging to 100%, they are showing:
MaxCellVolt_BMS 3624
MinCellVolt_BMS 3357
so a delta of 267 mv. Is that normal for charging/balancing, etc?
That's extremely out of balance.
Recommended holding voltage at 56.4 for at least a day and letting them try to balance.
Monitor the cells, if you get close to 3.65V/ cell reduce charging voltage by 0.1V until you stay under 3.65V.
 
That's extremely out of balance.
Recommended holding voltage at 56.4 for at least a day and letting them try to balance.
Monitor the cells, if you get close to 3.65V/ cell reduce charging voltage by 0.1V until you stay under 3.65V.
I can’t tell the individual cell voltages, much less what battery the high and low cells are in. Maybe @EG4_Jared woukd know?
 
What were the voltages when they were a couple mV apart?

In middle range of SoC (how batteries are stored and shipped), they will be similar voltage even if of different SoC.
Then when charged toward full they will diverge.

3.624V sounds close to target maximum; do you know what max voltage BMS protection allows before disconnecting?

As 42 Ohms says, setting inverter/charger to limit max voltage so highest cell doesn't exceed 3.65V and holding that for extended time should let balancing take place.

You've got 6 batteries. Do you have max/min cell data for each?
 
Unfortunately I don’t seem to be able to query the individual batteries with Modbus, though the monitoring portal at EG4 shows individual battery min/max and they are now within 7 mV in individual batteries. Here’s a graph of voltage from SA, looks like they spent some time in the balancing region. IMG_7033.jpeg
They now have two cycles, guess I’ll let them get some exercise. 🤓 Mostly just wondering if I should be concerned, but 50mV delta within a battery is OK?
 
Unfortunately I don’t seem to be able to query the individual batteries with Modbus, though the monitoring portal at EG4 shows individual battery min/max and they are now within 7 mV in individual batteries. Here’s a graph of voltage from SA, looks like they spent some time in the balancing region. View attachment 223055
They now have two cycles, guess I’ll let them get some exercise. 🤓 Mostly just wondering if I should be concerned, but 50mV delta within a battery is OK?

There should be no issue. After cycling the batteries a few times and taking some time to balance the cells, their voltages are expected to align more closely.
 
On the changelog page there seems to be a new update (FAAB-1B1B) but I am unable to download it. Anyone know when this will be available?
 
I tried starting a thread for this but got no replies.
I want to use a load shedding system with an 18K. To do that, I need a grid status relay output.
The 6000XP apparently has this.
The 18K has an extra set of relay outputs that are just noted as "reserved" in the manual.
I can use a power quality monitoring relay to do this on the grid side of the inverter and then set the high and low grid settings on the 18K to match the relay, but this still leaves some edge cases where it is possible for the grid to be disconnected and the load shed system to be disabled, namely during the "monitoring" period. This could potentially cause an overload and a shutdown, something I want to avoid.
Is there any way to get this status from the 18K, up to and including via RS485?
 
I have 4 powerpro batteries... 2 EG4 18kpv... Hooked up the recommended config for this setup... any idea why one battery shows 98% with same voltage as the 100% batteries.

1719084743343.png
 

Attachments

  • 1719084592895.png
    1719084592895.png
    183.1 KB · Views: 5
Don't worry about it unless they diverge further.

3.3xxV could be a settled voltage, with little current draw. Maybe what current was drawn did come a bit more from one than the others.
Or they differ a bit in accuracy/offset of current sensors.

3.6xxV, then I'd expect all to be closer to 100%.
 
Just got an 18Kpv installed on my test circuit with 2kW array and batteries.

Been playing with the settings to familiarize myself.

Multiple firmware updates and finally unit seems to work ok for now…. There are quite a few quirks I have noticed and haven’t figured out yet.

Here’s 1 that could be better.

Under battery backup mode, settings are
PV charge priority - Enabled-
PV Charge Power - 1kW

Time set to whatever to run…

Issue - if array is producing 2kW, of course only 1kW routes to battery which is correct, but the rest is lost???

For my example, running 1200W load, If setting is under normal consumption, ~800W goes to charge battery, ~1200 goes to run my test load.
As soon as I switch to PV Charge Priority, with the 1000W settings as above, the systems shows 1kW is going into battery ( battery showing taking in 9xx Watts) but then shows pulling from grid about 1200W. That means the other 1000W from array is totally lost.

Need to fix this so any overage from PV Priority settings feeds the loads.
 
Just got an 18Kpv installed on my test circuit with 2kW array and batteries.

Been playing with the settings to familiarize myself.

Multiple firmware updates and finally unit seems to work ok for now…. There are quite a few quirks I have noticed and haven’t figured out yet.

Here’s 1 that could be better.

Under battery backup mode, settings are
PV charge priority - Enabled-
PV Charge Power - 1kW

Time set to whatever to run…

Issue - if array is producing 2kW, of course only 1kW routes to battery which is correct, but the rest is lost???

For my example, running 1200W load, If setting is under normal consumption, ~800W goes to charge battery, ~1200 goes to run my test load.
As soon as I switch to PV Charge Priority, with the 1000W settings as above, the systems shows 1kW is going into battery ( battery showing taking in 9xx Watts) but then shows pulling from grid about 1200W. That means the other 1000W from array is totally lost.

Need to fix this so any overage from PV Priority settings feeds the loads.
I have a hard time believing that 1200W of power simply disappears. The only other place it could go is heat so I am willing to bet something in the way it is being displayed is incorrect.
 
I have a hard time believing that 1200W of power simply disappears. The only other place it could go is heat so I am willing to bet something in the way it is being displayed is incorrect.
The pv modules are not "pushing" the energy to the inverter. The inverter only draws what it needs or is told to.

In my case the inverter determines it only needs 1kW, so that is all it draws, even if the pv is capable of 2kW. So yes, 1kW is "lost".

Perhaps, EG4 can modify the firmware to have the 18kpv draw a %, or a min and max. On cloudy days, there's no telling how much power is available at any given time so a static setting less than pv array potential will not be optimal.
 
Last edited:
The only other place it could go is heat
If there is no display issue and it is accurate, then the power is dissipated as heat in the panel.

The MPPT should be able to estimate the power available from the data it collects during tracking, even if it is not drawing all of it. What these UIs may need to get better at is explaining what is actually being pulled and what is the estimated potential. OTOH, people could still get confused 🤷
 
I’ve got a parallel pair of 18Kpv inverters with a half-dozen PowerPro batteries, and my PoCo has just swapped out the meter from a mechanical one that did Net Metering to an electronic one.

I’d like to test pushing some power back to the grid to see how the meter reacts, it it as simple as turning on the grid breakers, waiting (how long?) for the inverter to sync up, and doing a forced discharge for a few minutes?

Do I need to have any settings preconfigured?

Thanks!
 

diy solar

diy solar
Back
Top