(RESOLVED) More than 16 EG4-LLv2s not communicating to Sol-Ark

[linuxdrake]

Hi All,

Got another interesting problem, I have 12x EG4-LL with 4 DIP's, and 12 more with 6 DIP.
According to Signature Solar, have been communicating with them, and even they are stumped at the moment. The 1st Master Batt, 6DIP with ID1, is first, followed by the 12 (4dip) batterys with ID 2 - 13, then 14-24 is the 6DIP batteries.

All 24 batts using the latest firmware from 02/24/2023. All Batts are about the same % charge and voltage, within 1-2 tenths of a volt, and charging all about 4-5A right now.

But the Sol-Ark Inverter is not seeing past 1600AH, or 16 batts.

I powered all the batts off, then starting with 24, then 23 and so on, powered them back up with master last, Sol-Ark only sees 1 batt, or 100AH on the Batt Conf, with BMS lithium on 00.

Next I tried, powering all the batts off, then Batt1, Sol-Ark sees that 100AH, then powered up batt2, Sol-Ark sees 200AH or both batts, and repeated the process to batt 16, no issues, see all 16 or 1600AH over CAN from the Sol-Ark inverter.
But when I power up batt 17, issues start, Sol-Ark reads 800AH or sometimes 100AH. Very odd.

Looking at the batts Batt-Com ports, Batt1, has both Rj45 ports lights blinking very rapidly, as expected for the master batt.
All other batts up to 16, all slow flash, like they are being poll'ed which I assume expected for CAN bus.

But whats intresting is batt 17, has the same very fast port light blinking exactly like Batt1, could be an issue with firmware? not sure.
Batt 18-24, all have the slow flash, like batts 2-15, which I expected. Very odd.....

Plugging in the RS485 USB to RJ45 cable, I confirmed batt ID's were correct from 2-16, as Batt1(master) gives junk info, as its in the host state, expected.
But also batt 17 is like that too. Batt 18 shows in the BMS_Tools app as ID2, 19 as 3 and so on. I assume as the tool is built for up to 16 and is just rolling over.


Any thoughts folks? I have contacted SignatureSolar and they are not sure, I am sending them over pictures, but wanted to ask the community here if anyone has done over 16 batts and ran into the same.




NOTE: All batts are connected to a common set of bus bars, all with equal length Pos and Neg 4/0 cables, then equal length 4/0 to both Sol-Arks, double runs to use both batt terminals for maximum discharge and charge current.
Both Sol-Ark 15's are paralleled, common AC bus, common Load bus and common GEN inputs in the wireways.
System has been working without issue with 12 batts for over 2.5MWh's or 3 months, just added the 2 extra cabinets.

 

[solardad]

I thought their was a BMS limitation of 16 batts… you may have to run them dumb, no can bus connection to inverters

 

[linuxdrake]

I thought so, but after talking to Signature Solar folks they said up to 64 now, also they put it on their website

 

[Quattrohead]

I think the Solarks are limited to 1600A, at least I think that is what I have seen on mine.

 

[FilterGuy]

Got another interesting problem, I have 12x EG4-LL with 4 DIP's, and 12 more with 6 DIP.

I am assuming that means you have 12 LL-V1 batteries and 12 LL-V2 batteries.

My speculation is that the V1s can only address 16 batteries (2^4) and the V2s can address 64 batteries (2^6). I am also speculating that if you try to mix the two, the limit will be 16. The V1s are probably trying to respond to addresses above 12.

As an example, a V1 with address 0001 may see address 010001 as 0001 and try to respond on top at the same time the actual 0001 responds.

Depending on how the V1 decodes things, there may be a fix in firmware.

 

[linuxdrake]

I am assuming that means you have 12 LL-V1 batteries and 12 LL-V2 batteries.

My speculation is that the V1s can only address 16 batteries (2^4) and the V2s can address 64 batteries (2^6). I am also speculating that if you try to mix the two, the limit will be 16. The V1s are probably trying to respond to addresses above 12.

As an example, a V1 with address 0001 may see address 010001 as 0001 and try to respond on top at the same time the actual 0001 responds.

Depending on how the V1 decodes things, there may be a fix in firmware.

Thats what I thought as well, funny the "v2" has 4dip and 6dip but both called "v2"

I tried with the "master" as a 4dip model, as well as a 6dip model as the master, no change.
What's odd though, is for about 3min it did see 2000AH, then it dropped back to 800AH then 1600AH, so could some something in the firmware.

 

[FilterGuy]

BTW:
I see that there is a Firmware upgrade to V1. I have no idea what is in it.

 

[FilterGuy]

There is a V2 FW update as well.

 

[linuxdrake]

There is a V2 FW update as well.

Flashed them all with "all_in_one_16s_id4(app)_03141647.bin" from the 2-24-2023 Zip from EG4 website.

 

[linuxdrake]

Picture show the 2 "gens" of the V2

 

[DIYrich]

You can have up to 9900AH connected, but you may be limited to 16 batteries that it can read. Maybe put 8 batteries behind a smartshunt and use Solar Assistant

 

[linuxdrake]

You can have up to 9900AH connected, but you may be limited to 16 batteries that it can read. Maybe put 8 batteries behind a smartshunt and use Solar Assistant

I would agree but Engineer775 (PracticalPreppers) has 30 in closed loop!

I did send him an email and YT DM, but I doubt unless he watches this forum for a reply, he is a busy guy like us all making a living in the Electrical world.

 

[linuxdrake]

Interesting new twist, trying to further diagnose the issue with only able to see 16 batts at once, resetting the com on the master was a thought I had.

But.....Under no manual procedure or one I can search the forum or from google, can I successfully change the Inverter Protocol on the EG4LLv2's with 6DIPs with the current latest firmware.

Using the below table, I tried ID1, ID16, ID63 and ID64
Powered Off BMS, Breaker Off then on, no communication cables connected to Batt-Com, RS485 and CAN ports.
Powered On Breaker, BMS On, held button 3 for 5, 8, and 10 sec, tried Button 4, then 3+4, then 1, 2, all combinations.
Then shutoff BMS, changed DIP to a new setting and tried again.

Consumed way to much coffee to try and make this work, but in no combination I can change the protocol with the 2-24-2023 firmware bundle from EG4's website.

Exact firmware file: all_in_one_16s_id4(app)_03141647.bin

Any thoughts from the forums brain trust?



Looks similar to this youtuber who sounds like same exact issue.

 

[linuxdrake]

Bumping for help No update yet from Signature Solar

 

[Markus_EG4]

Bumping for help No update yet from Signature Solar

What’s your problem

 

[Markus_EG4]

Bumping for help No update yet from Signature Solar

Never mind i see send me your email so I can get you taken care of

 

[linuxdrake]

Update!

Great news all, after talking with the SignatureSolar Support folks, @Markus_SignatureSolar for the massive help!, even after hours, was able to reflash with the correct firmware for each "DIP" switch model.

After reflashing, was able to set Comm's following all DIP to the LEFT on the 6DIP, and all towards you on 4DIP, just to verify.

Turns out, do not flash the 6DIP batts with the id4 firmware, that is only for the 4DIP models.

The fast flashing was due to the wrong firmware on the 6DIP model, causing it (I think) to start its own bus with ID17, which looks like ID1 and compete with the master, on real ID1.

Screenshot from it all working flawless with the Sol-Ark.

Tip: Connect all the batts together, set DIP switches and reboot all the BMC's, except the master, do that one last and your good to go.
Also use the DTECH RS485 utility, not the BMS USB device, its not the correct chipset for reflashing.
Link: https://www.amazon.com/dp/B076WVFXN8

 

[JoeHam]

WOW, nicely done.

I’m so jealous, and you even had to throw the e-mower in the picture ?.

It is so nice to see a “Resolved” thread, with concise details, to help others down the road.

 

[linuxdrake]

WOW, nicely done.

I’m so jealous, and you even had to throw the e-mower in the picture ?.

It is so nice to see a “Resolved” thread, with concise details, to help others down the road.

For sure!