Signature Solar EG4 48v 100ah Batteries Alarming Out

[AleksK]

Do you have a grid connection to the growatt? And are you in the usa?

 

[mrall]

Sorry, I do not. The best thing to do would be to use a combination of an RS485 listener (so you can see the packets being emitted by the battery BMSs) in combination with the PC software provided by EG4. This should allow you to reverse engineer things. Also, if you're trying to sort out Inverter<->BMS communications, you probably want to use the address 1 protocol, which I understand is specific to allowing the inverters to communicate with the battery. @RichardfromEG4 might be able to help further. It should not require reverse engineering line protocols and writing a custom adapter. There are others with Growatt inverters that I believe have gotten comms to work correctly.

Thanks. I have have reverse engineered 95% of the the protocol using a sniffer. Only a handful of bits elude me including Protection Status and CMOS/DMOS state. Those conditions are hard to force or fault and I don't want to generate faults on purpose. I plan to use an M5Stack ESP32 to poll the batteries and upload the data in JSON to a web server. I did the same thing with my weather station so most of the code is already created.

 

[mrall]

@mrall Solar Assistant is able to read from the 48V LiFePower4 batteries over the RS485 port using the "Narada" protocol. Maybe you'll have some luck finding it?

I would also point out that there is some inconsistency in the DIP switch address settings and it's particularly important that you get address #1 correct since that is supposedly used to communicate with inverters. See this thread:

EG4 dip switch settings

I seem to remember a post a while back where it was mentioned that the user manual for the eg4 batts 48v, no lcd, had reversed the position on the dip switch settings from reality ie... if the book said for a particular ID number set up up down up, it should actually be down down up down...

diysolarforum.com

The protocols are apparently different if using address #1 vs. any of the other addresses according to Solar Assistant (https://solar-assistant.io/help/battery/eg4). On my LiFePower4 batteries (no LCD), if I use the manual's setting for address 1, I actually get address 7. The best thing to do is hook up the BMS software and plug into your #1 battery (the one that interfaces with the inverter) and scan all addresses to see what "PACK ID" is showing up. If it's not #1 then the battery is trying to talk to the inverter using a different protocol and that could explain why it's not working.

There are also reports that some DIP switches work correctly with the manual's table, while others do not. None of my batteries worked according to the manual's table but all worked correctly using the table I posted in the thread above. I checked all addresses from 0-15. Hope you're able to solve this and hope it doesn't require reverse engineering the protocol and writing an adapter (!).

I have 2 48V LifePower4 batteries (no display) purchased in May 2022. With dip switches on battery 1 set to Up-Down-Down-Down, I can access it from BMS_Test_V1.0 at device address 1 and the same for battery 2 set to Down-Up-Down-Down using device address 2. Both are speaking the Narada Protocol. I have yet to get them to communicate with my SPF3000TL LVM...

 

[goldserve]

Thanks. I have have reverse engineered 95% of the the protocol using a sniffer. Only a handful of bits elude me including Protection Status and CMOS/DMOS state. Those conditions are hard to force or fault and I don't want to generate faults on purpose. I plan to use an M5Stack ESP32 to poll the batteries and upload the data in JSON to a web server. I did the same thing with my weather station so most of the code is already created.

Get this working yet?

 

[niktak11]

Can you check the individual cell voltages? 56.5V is pretty high for a pack that isn't very closely balanced. The (non-eg4) pack I received a couple months ago would hit single cell OVP at about 54.5V.