96V battery build. BMS Help.

[chofmeister]

Hey I am trying to build a 96V pack for a project and I haven't been able to find some concise info on building packs of this voltage using available BMS technology. If there is a good thread on this I would appreciate a link. It seems like higher voltage packs
wold require lower voltage BMS systems arranged in series and then possibly a battery controller monitoring the individual BMS?

 

[amberwolf]

96v doesn't fit any of the usual numbers of series cells**, so except for BMSs that are modular*** there may not be much out there for it.

Is 96v your full charge voltage, or your nominal voltage? How many series cells are you using, of what chemistry?


**assuming NMC, LiCo, etc then at 3.7v nominal, 96v would be about 26s. at 4.2v full charge, 96v would be about 23s.

***there are some BMSs that use a balancer/sensing board on each individual cell (group) that each talk to a "master" BMS unit; IIRC Energus is one of those. I don't remmber any other names but they shouldn't be too hard to find. Expensive, though.



If you break the battery into smaller series chunks with their own BMS, you can run into problems with total voltages across points being too high for the components in them, especially if each one has it's own charge/discharge FETs, because as soon as the BMS turns those off, then the total voltage of *all* of the packs is now across that open-circuit, and if the FETs aren't rated for that voltage...pow.

 

[j bjork]

I havent seen this as a problem, I plan to build a 30s battery with a 32s ant bms.

 

[chofmeister]

96v doesn't fit any of the usual numbers of series cells**, so except for BMSs that are modular*** there may not be much out there for it.

Is 96v your full charge voltage, or your nominal voltage? How many series cells are you using, of what chemistry?
96V nominal. Looking to use 18650's or possible a LiFePo cells.

**assuming NMC, LiCo, etc then at 3.7v nominal, 96v would be about 26s. at 4.2v full charge, 96v would be about 23s.

***there are some BMSs that use a balancer/sensing board on each individual cell (group) that each talk to a "master" BMS unit; IIRC Energus is one of those. I don't remmber any other names but they shouldn't be too hard to find. Expensive, though.



If you break the battery into smaller series chunks with their own BMS, you can run into problems with total voltages across points being too high for the components in them, especially if each one has it's own charge/discharge FETs, because as soon as the BMS turns those off, then the total voltage of *all* of the packs is now across that open-circuit, and if the FETs aren't rated for that voltage...pow.

96V nominal. Looking to use 18650's or possible a LiFePo cells. My idea was to make three packs 26S16p divided by heat transfer plates, which are linked together to make a 26s48p pack. I've seen a lot of options for 48V BMS up to 60V but nothing in the higher voltage range and I'm not sure how to approach it. I figured there must be members on here making higher voltage packs with specific management systems but I really haven't found anything difinitive.

 

[amberwolf]

Well, once you decide what you're going to build the battery as, then you can decide on BMS solutions.

If you aren't sure how you're going to build the battery (how many series cells of which chemistry) then you probably shouldn't buy a BMS yet, as you might need a different one for one setup vs another.

If the chemistry you want to use is supported by them, there are apparently BMSs like the ANT referenced above that do up to at least 32s
https://www.aliexpress.com/item/32849893644.html
though you would need to find out from the manufacturer if it can be used with less series cells/groups than that.

 

[chofmeister]

I havent seen this as a problem, I plan to build a 30s battery with a 32s ant bms.

Have you posted a build thread about this? I'd like to see different approaches to building these larger packs.