search

byte4geek / SEPLOS_MQTT

Read data From Seplos BMS and send them to the Home Assistant
GNU General Public License v3.0
39 stars 13 forks source link

Extending the scope to cover a second pack #1

Closed championc closed 1 year ago

championc commented 2 years ago

The Master is always "Pack 00", but if you daisychain a second Seplos, it will be given the identity Pack 01

It would be good if the script can pull the data for Pack 0x

Great work

byte4geek commented 2 years ago

i did think to this but i do not have a second battery pack for now, i can't test it.

championc commented 1 year ago

Hey again,

I can test. The packs are 00, 01, 02, ......0x

Enjoy your coffee (or beer) :-)

byte4geek commented 1 year ago

I will try to write a script for 2 packs but I can't test it, I have only one pack.

championc commented 1 year ago

Just a quick update. I can run the script with ADDR=00 and later with ADDR=01 and the results are different (as I expect). So it is correctly able to extract the results from each pack. Of course, the best solution is to pull all results at the one moment

championc commented 1 year ago

Further Feedback. If I query with ADDR=02, it returns an error, which is correct. Entering ADDR=01 should assume that you have two packs 00 & 01

So ADDR should represent the Last Pack in a setup

championc commented 1 year ago

Further Update

I have made a crude hack version to call query_seplos_ha.sh, output all data, then run a modified query01_seplos_ha.sh on Pack01 and output all data with 01 before each data variable (so \"01lowest_cell\":\"Cell $lowcellnumb - $lowcell mV\",\ etc...)

byte4geek commented 1 year ago

good, i can add 2 scripts one for each pack executed by the first script

championc commented 1 year ago

UPDATE - Seplos have confirmed that you CANNOT obtain the details from a Primary Master pack when it has an additional Secondary unit in parallel.

So it's maybe best to stick with just your original script ONLY

ztardik commented 1 year ago

@championc

UPDATE - Seplos have confirmed that you CANNOT obtain the details from a Primary Master pack when it has an additional Secondary unit in parallel.

Hi, can you explain in more details what they meant with this?

@byte4geek

good, i can add 2 scripts one for each pack executed by the first script

Could you write a short how-to on this? I'm also with dual pack configuration.

byte4geek commented 1 year ago

You cant do this in parallel, but you can do this in sequence.

I'm writing another script that call query_seplos_ha00.sh and after call query_seplos_ha01.sh for example and send the data to HA in different naming.

byte4geek commented 1 year ago

i added the scripts you simply download and add to your SEPLOS_MQTT folder.

update the user password, mqtt server ip and so on

chmod 700 run_bms_query_dual.sh 
chmod 700 query_seplos_ha01.sh

and try to run ./run_bms_query_dual.sh

you must create all entity manually in Home Assistant like the first pack. all mqtt entity in home assistant are called with 01_ prefix see below:

01_lowest_cell 01_lowest_cell_v 01_lowest_cell_n 01_highest_cell 01_highest_cell_v 01_highest_cell_n 01_difference 01_cell01 01_cell02 01_cell03 01_cell04 01_cell05 01_cell06 01_cell07 01_cell08 01_cell09 01_cell10 01_cell11 01_cell12 01_cell13 01_cell14 01_cell15 01_cell16 01_cell_temp1 01_cell_temp2 01_cell_temp3 01_cell_temp4 01_env_temp 01_power_temp 01_charge_discharge 01_total_voltage 01_residual_capacit 01_soc 01_cycles 01_soh 01_port_voltage

byte4geek commented 1 year ago

If it works change the entry in crontab with:

@reboot nohup ~/SEPLOS_MQTT/run_bms_query_dual.sh &

championc commented 1 year ago

The problem with two packs in parallel is that, when querying Pack00 and Pack01, the results are the same. The only solution is to make 2 x Master.

Personally, I am now going to change my cells configuration to 2p16s and only use 1 x Seplos BMS, and sell the other one

byte4geek commented 1 year ago

Sorry but I don't understand, if you use the vendor tool, you can switch between each pack. This assume that you have address 00 and 01 and the cable connected PC -> pack 00 -> pack 01, this is correct?

championc commented 1 year ago

Seplos say that if you have a Pack01, Pack02, Pack03 ....... then you can NEVER read the details of Pack 00 via RS485 :(

They say that the dip switch must be changed to all off, to read the Pack00 details

byte4geek commented 1 year ago

Seplos say that if you have a Pack01, Pack02, Pack03 ....... then you can NEVER read the details of Pack 00 via RS485 :(

They say that the dip switch must be changed to all off, to read the Pack00 details

Do you have any update regarding 2 packs?

MrAlgo commented 1 year ago

I also wanted to monitor 2 packs / BMS but it’s not possible via their software as confirmed by Seplos and presumably no other software due to hardware restrictions of the BMS? This video explains it further.

https://youtu.be/liu86NoFiQc

MrAlgo commented 1 year ago

I wonder if the data the screens receive is also RS485? Could you tap into the data sent to them for those of us with two packs?

ztardik commented 1 year ago

The screen is probably rs232, that is the cheapest option for short links. But there is no option to edit parameters on the screen. One other option is to try listening on the internal 485 bus. Maybe the master shares it's data with the slaves. The last option is to use CAN bus. I've got an adapter but could not test it until April.

MrAlgo commented 1 year ago

That would work for me as I only need to be able to monitor both packs remotely when away rather than change any settings.

Presumably, I would just need an RS232 to USB adapter for the hardware but could this bash script be modified to work with the screens output?

Got to say that I am a novice with all of this and my skills stop at implementing what others have created and some slight modification or addition of code to make it work with my HA instance so I'm not going to be much help in that department, but I can test.

Seplos have just come out with a Bluetooth version of this BMS. Some new hardware on the new revision of screens and some on the BMS itself.

I spoke to Seplos to see if replacing my current screens would give me BT read-only access and they confirmed it would so I ordered 2 which are en route to me now. This at least solves checking both packs while I'm within BT range via their app. Just need to sort the same but when I'm away.

It really is a bit ridiculous that Seplos have created a system where two packs + can't be monitored by their Battery Monitor software... If that worked I would just put a cheap laptop next to them and remote in to check and adjust when needed...

markvk42 commented 1 year ago

Ive just been watching the off grid garage set of vids on this, as far as im aware if you have the same type of BMS and you are not trying to use can bus then you can monitor all of the batteries within their software. If you are using Can then you cant monitor the master only the slave. To get around this they introduce the BT version, but its stupidly buggy!

byte4geek commented 1 year ago

i removed the script for 2 batteries since seplos made this questionable choice that not to allow tracking of multiple packages connected together, i am very disappointed with this choice of seplos.

markvk42 commented 1 year ago

Ok Thanks, so how would i monitor 3 banks? 3 ras pies or i pi running containers some how?

ztardik commented 1 year ago

@byte4geek Have you looked into the CAN protocol? Maybe it would be a viable option to just listen on the CAN bus in case of more than one pack. I have MKS CANable Pro that I bought some time ago for 25 Eur, I can test it in April.

bbbowden commented 1 year ago

Concerning multiple packs... We are connecting multiple batteries in parallel and utilizing the CANBUS to aggregate the data from those batteries into a pack message. In the CANBUS message, the high and low cell and which battery the cell is in is indicated in the CANBUS message. Unfortunately, the individual cell data is not relayed in the CANBUS message.

So, I am looking at not using a "Master Battery" (meaning that its address is set to 0) and rather gathering the RS485 data myself and then generating the CANBUS message to be used by the inverter. This does mean that the maximum number of batteries I can have in parallel is 15, but that shouldn't be a problem.

I'm in the process of modifying these scripts to sequentially query each subsequent battery. Using the Seplos BMS software, you can see how the query message changes i.e.: Battery 1: ~20014642E00201FD35$ Battery 2: ~20024642E00202FD33$ etc... The things that change are characters 3 and 4, characters 13 and 14, and then the checksum.

So, modifying the script shouldn't be too hard.

Now my question:

Looking at the communication log in the Seplos BMS software, there are two commands sent for each pack query: Get pack remote communication data and Get pack telemetry data. Do you have the protocol description of the response to the Get pack remote communication data message? It's not critical (I don't think) but I'm curious what is in there.

By the way, thank you for parsing the pack telemetry data. That is a huge help in my efforts.

byte4geek commented 1 year ago

I found only this document on internet about the protocol, try to find all information inside. https://www.dropbox.com/s/myvjpi74prjkcgz/SEPLOS%20BMS%20Communication%20Protocol_V2.0.pdf?dl=0

bbbowden commented 1 year ago

I have created a barebones Node Red flow to query the Seplos BMS across multiple batteries. Once you set the number of batteries to query, set the Multi Inject Time stamp to re-occur at your sample interval. The battery data comes out on debug 7 and can be used however you want (i.e. inject into MQTT).

You will need to have the Buffer Parser node installed in your Node Red instance. It's the only non-standard node used in this flow.

Hope someone finds this helpful.

SeplosBMS_Multi_Query.zip

bbbowden commented 1 year ago

Opps, missed a couple of nodes (the ones to set the number of batteries) when I exported the flow.
SeplosBMS_Multi_Query_Set_Number of batteries.zip

markvk42 commented 1 year ago

Opps, missed a couple of nodes (the ones to set the number of batteries) when I exported the flow. SeplosBMS_Multi_Query_Set_Number of batteries.zip

bbbowden and all others. Thanks for your work, when i get my batteries im certain ill be back to ask questions!

pedro37 commented 1 year ago

Opps, missed a couple of nodes (the ones to set the number of batteries) when I exported the flow. SeplosBMS_Multi_Query_Set_Number of batteries.zip

I have created a barebones Node Red flow to query the Seplos BMS across multiple batteries. Once you set the number of batteries to query, set the Multi Inject Time stamp to re-occur at your sample interval. The battery data comes out on debug 7 and can be used however you want (i.e. inject into MQTT).

You will need to have the Buffer Parser node installed in your Node Red instance. It's the only non-standard node used in this flow.

Hope someone finds this helpful.

SeplosBMS_Multi_Query.zip

Thanks, is it possible for a step by step guide?

navesile commented 1 year ago

Opps, missed a couple of nodes (the ones to set the number of batteries) when I exported the flow. SeplosBMS_Multi_Query_Set_Number of batteries.zip

awesome work. I got it up but keep getting weirdness for cid. any insight? up using 8s system

image

navesile commented 1 year ago

thanks bbbowden for the the heavy lifting. I figured it out. messed it up, cut things, added things, sent to db, just started to play with grafana.

image

here is my flow. 28V2battery.zip

byte4geek commented 1 year ago

Monitor 2 or more batteries is possible only via CAN protocol

navesile commented 1 year ago

i I don't think so much details is available in can

On Tue, Oct 24, 2023 at 2:48 AM byte4geek @.***> wrote:

Monitor 2 or more batteries is possible only via CAN protocol

— Reply to this email directly, view it on GitHub https://github.com/byte4geek/SEPLOS_MQTT/issues/1#issuecomment-1776624226, or unsubscribe https://github.com/notifications/unsubscribe-auth/AP3TPXQVNF5A7NVLIPLEVEDYA5QDVAVCNFSM6AAAAAARAKPK7CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONZWGYZDIMRSGY . You are receiving this because you commented.Message ID: @.***>

-- Regards,

Andy Miller OneAndAHalfLegs.org