2x BMS's on 1 Pack?

[Lt.Dan]

Hey y'all, I've done some searching but didn't come up with much. Is there any downside to connecting 2x BMS's to one pack?

For more info, it will be 2x 200a Seplos BMS's, connecting to a 2P16S 542Ah pack.

The only reason I don't want to split the pack is because I already have it in a nice compression fixture, with high quality laser cut 4x bus bars, etc. The pack is also matched with highest/lowest cell capacity to average out to the nominal capacity (ie a 283ah cell is paired with a 263ah cell to average 273ah).

Main reason I'm lazy and would like to just wire in the BMS's lol.

I'm curious how communication between the 2X BMS's and between the master and the inverter are going to be handled, but I dont see how it would be a problem?

 

[MisterSandals]

Is there any downside to connecting 2x BMS's to one pack?

Just out of curiosity, what problem are you trying to solve?

The pack is also matched with highest/lowest cell capacity to average out to the nominal capacity (ie a 283ah cell is paired with a 263ah cell to average 273ah).

Nice! It takes a bit of time and effort but the results speak for themselves!

 

[Lt.Dan]

Just out of curiosity, what problem are you trying to solve?

Nice! It takes a bit of time and effort but the results speak for themselves!

I have an older model Seplos 10D BMS that won't communicate with my Sol-Ark and has some communication issues with PC as well. So I purchased 2x 10E models with bluetooth to replace it.

I also don't like running a 200a BMS almost at max current, and would prefer splitting the current between 2x BMS's.

 

[MisterSandals]

I also don't like running a 200a BMS almost at max current, and would prefer splitting the current between 2x BMS's.

Ok, got it.

So would you charge thru both BMSs?

If not, they would not be able to count coulombs (if thats what they do to estimate SoC) so maybe their estimated SoC and communication to Sol-Ark would be "suboptimal"? I dunno.

Would both BMSs be connected to the Sol-Ark and report "duplicate" info on the same battery bank?

 

[Lt.Dan]

Ok, got it.

So would you charge thru both BMSs?

If not, they would not be able to count coulombs (if thats what they do to estimate SoC) so maybe their estimated SoC and communication to Sol-Ark would be "suboptimal"? I dunno.

Would both BMSs be connected to the Sol-Ark and report "duplicate" info on the same battery bank?

I would connect both BMS's to a Bus Bar, then 1 cable to the inverter itself. Essentially, both BMS's will think they are connected to their own pack, and charging through both, assuming there is no unreasonable amount of difference in resistance of the cables, should charge at similar rates, and thus still show similar battery SOC %. One will act as the slave, which will report SOC to the master BMS, which should report overall SOC to the inverter.

Unless I'm seeing this wrong? Thats why I'm trying to make sure lol.

 

[MisterSandals]

Unless I'm seeing this wrong? Thats why I'm trying to make sure lol.

Seems safe enough to give it a go. Unless someone has actually done this, its hard to tell if there are any gotcha's. I don't see any but thats no saying much...

 

[MisterSandals]

One will act as the slave, which will report SOC to the master BMS,

This is the one thing that concerns me (master/slave BMS concept new to me) in terms of correct reporting of data to the Sol-Ark. It would likely be harmless in terms of safety if data were wrong but wrong data would make your system unworkable.

 

[Lt.Dan]

This is the one thing that concerns me (master/slave BMS concept new to me) in terms of correct reporting of data to the Sol-Ark. It would likely be harmless in terms of safety if data were wrong but wrong data would make your system unworkable.

Only one way to find out ?‍

 

[MisterSandals]

Only one way to find out ?‍

@Ampster does a LOT of out-of-the-box thinking and very experienced with Sol-Arks. Wondering what he thinks?
Hopefully this will conjure him.

 

[Lt.Dan]

@robby has experience with the Sol-Arks too.

@houseofancients has experience with the Seplos BMS's.

Any hints or problems you foresee guys?

 

[houseofancients]

@robby has experience with the Sol-Arks too.

@houseofancients has experience with the Seplos BMS's.

Any hints or problems you foresee guys?

well, let me make sure i get this right ..

you have 32 cells in 1 pack that you essentially want to split up in 2 16s packs, using 2 seperated bms's ?

kind of makes sense, inwould make sure both bms's are in comms together.
as these bms's have in rush guarding, you could even consider using busbars for both positive and negatives, make sure those can handle 400a
i would fuse every 16s pack with an individual t-class fuse , then connect to busbar .
only downside i can see is not being able to sepetately shutdown each 16s pack individually

comms to sol-arl would be 1 utp cable from master bms to sol-ark's bms port, should be pretty straightforward

 

[Lt.Dan]

you have 32 cells in 1 pack that you essentially want to split up in 2 16s packs, using 2 seperated bms's ?

No i would like to keep them in 1x 2P16S pack and use 2x seperated BMS's.

 

[houseofancients]

No i would like to keep them in 1x 2P16S pack and use 2x seperated BMS's.

that could cause al sorts of unexpected behavior, even if you disable most services in both bms's like balancing.

have not ever tested, have not ever done something like it, and dont think i ever would take the chance to do so and possibly ruin my cells
guess i am not adventures in that regard

just my 2 cents

 

[Ampster]

@Ampster does a LOT of out-of-the-box thinking and very experienced with Sol-Arks. Wondering what he thinks?
Hopefully this will conjure him.

Thanks for the compliment. I am very new to the SolArk so what I have learned is all very new but may be helpful

If this is the question at hand:

No i would like to keep them in 1x 2P16S pack and use 2x seperated BMS's.

I understand the value of a 2P16S pack but don't see a value in using two BMSs. I actually have a 3P16S pack and use a BMS and an active 2Amp balancer. I disabled the balancing on the BMS but it does the SOC calculations, and communicates to my inverter. My BMS (Orion JR2) has an EV heritage so I know it is robust but it has a very small balancing current. Its older brothers are used to 400 volt packs of many cells and do not need larger balancing packs because most EV guys buy automotive grade cells that are fairly evenly matched. That is why I used the JK active balancer.
I guess my quesiton to @Lt.Dan is what is the goal other than what I read that he does not want to push a BMS to its limit? If that is the case consider a contactor based BMS like some of the ones Docan sells. That is what my Orion uses and theoretically there is no current limit because the contactor is the gateway. I charge as high as 150 Amps and do not trust MOSFETs for that kind of control, especially under load. MOSFETs and SSDs often fail closed which is not what you want in a BMS. Did that help? If not let me know how I can help I did not read the entire thread yet for context.

 

[Delphinus]

Hey y'all, I've done some searching but didn't come up with much. Is there any downside to connecting 2x BMS's to one pack?

For more info, it will be 2x 200a Seplos BMS's, connecting to a 2P16S 542Ah pack.

The only reason I don't want to split the pack is because I already have it in a nice compression fixture, with high quality laser cut 4x bus bars, etc. The pack is also matched with highest/lowest cell capacity to average out to the nominal capacity (ie a 283ah cell is paired with a 263ah cell to average 273ah).

Main reason I'm lazy and would like to just wire in the BMS's lol.

I'm curious how communication between the 2X BMS's and between the master and the inverter are going to be handled, but I dont see how it would be a problem?

Because one of my BMS order was lost I start my build with 2P16S, but has soon I got a second I build 2 batteries of 16S, it's little work and you will get redundancy and is great for maintenance https://diysolarforum.com/threads/1s2p-48v-battery-in-a-server-rack-case.51725/post-664593

You already have the hardware you should keep it simple, keep you compression fixture only rearrange the bus bars.

 

[time2roll]

I also don't like running a 200a BMS almost at max current, and would prefer splitting the current between 2x BMS's.

Keep the second BMS as a spare and the original will never fail.

 

[Ampster]

You already have the hardware you should keep it simple, keep you compression fixture only rearrange the bus bars.

If it is a matter of rearranging the busbars I would consider that much easier that trying to run two BMSs on a single pack. Redundancy does have its advantages. My preference for a single large pack is driven by the cost of my $500 BMS and not wanting to by another. Although, now that I think about it if I ever wanted to add more capacity I could do two 2P16S packs and then another BMS would give me the value of redundancy.

 

[Lt.Dan]

that could cause al sorts of unexpected behavior, even if you disable most services in both bms's like balancing.

have not ever tested, have not ever done something like it, and dont think i ever would take the chance to do so and possibly ruin my cells
guess i am not adventures in that regard

just my 2 cents

I'm trying to understand what unexpected behavior could be presented? I don't see the problem with even leaving the balancing circuit on for both BMS's? Both BMS's will essentially be seeing the exact same cell voltages and temperatures, but they will think they are 2 separate packs all together. I don't foresee one BMS saying "HEY YOU HAVE THE SAME CELL VOLTAGES, I'M GONIG TO TURN OFF!".

If that is the case consider a contactor based BMS like some of the ones Docan sells. That is what my Orion uses and theoretically there is no current limit because the contactor is the gateway. I charge as high as 150 Amps and do not trust MOSFETs for that kind of control, especially under load. MOSFETs and SSDs often fail closed which is not what you want in a BMS.

Yes I do not like running even a high quality Mosfet above 75% load. Runing 2x BMS's would take my 150a limit (for a 200a BMS) to a 300a limit, which my inverter will never be able to get over 185a of charging anyways.

You already have the hardware you should keep it simple, keep you compression fixture only rearrange the bus bars.

Forgive my messiness in this photo, but it shows how the battery is setup with compression and bus bars etc. Its not something I can easily rearrange and split the pack into 2x separate packs. It would involve all new bus bars and more hardware that frankly I'm not interested in doing. In that case, I will stick to 1x BMS and just keep the other as a backup.




This is what it looks like currently:

Keep the second BMS as a spare and the original will never fail.

LOL, this is entirely too true!

 

[Ampster]

This is what it looks like currently:

Wow that is nice. If I had put that much effort into those nice looking busbars I would keep that configuration. I also like that your attachements for the busbars are separate from the battery studs. That makes adding the second BMS easy while the pack is in service.
If I understand the SolArk battery communications it will see it as two identical packs and the metrics will be the same. The only issue may depend on how the BMS behaves. Mine is pretty sophisticated and does some things to be able to report internal resistance, open circuit voltage and regular voltage. If yours does the same that may confuse the other which may have a different cycle of disconnecting to take those measurments.

 

[Delphinus]

Looking at your photos you also will need to arrange the cells but your compression is good and ease to do it. I will do it but at the end this is your decision do what you think is best for you, by your photos you don't seem to be a lazy person.