[mdrobnak]

I saw in another thread (that I can't find now), if the BMS slaves 'in the middle' weren't connected to a fully populated pack, it would kill cells..

Two things:

Which are the 'middle' modules?
Was there a solution ever found?

I am building a 160 cell pack, so I will be using a full set of slaves and then some...I'm worried about the 'and then some' part.

Given my cell configs are 16 cells per module, I assume some measurements are going to cross the measuring modules, similar to how Thunderstruck is wiring with their BMS being in groups of 8. (See http://www.thunderstruck-ev.com/images/companies/1/LGChemConnectorPinouts.pdf)

Any info is greatly appreciated.

Thanks.

-Matt

 

[swoozle]

I've done work with the Gen 1 Volt BMS and modules. The slaves (BICMs) require that all of the inputs for a given BICM are populated with cells. For instance BICM 1 has two 12-cell inputs and one 6 cell. If you try to run it with just two 12-cell modules it will not work correctly and drain some (4? it's been awhile) of the cells.

In addition the cells need to be in series but just within a given BICM. I tried to hook up multiple 12-cell modules that were wired in parallel (a large 48V pack for a house inverter) and it does not work.

The BICMs do draw power from the cells, so any cells that don't have a BICM attached will slowly get out of balance with the pack.

So 160 cells won't work. That's not evenly divisible by 6, the smallest Volt BICM input size. With 16-cell modules the closest you could get is 144 cells.

Another thing to note is that while you can use multiple identical BICMs ( for instance two BICM 1's), the CAN messages will be duplicated. You'll get all of the cell data, it'll just be tricky to tease out which numbers belong to which BICM.

 

[mdrobnak]

I've done work with the Gen 1 Volt BMS and modules. The slaves (BICMs) require that all of the inputs for a given BICM are populated with cells. For instance BICM 1 has two 12-cell inputs and one 6 cell. If you try to run it with just two 12-cell modules it will not work correctly and drain some (4? it's been awhile) of the cells.

In addition the cells need to be in series but just within a given BICM. I tried to hook up multiple 12-cell modules that were wired in parallel (a large 48V pack for a house inverter) and it does not work.

The BICMs do draw power from the cells, so any cells that don't have a BICM attached will slowly get out of balance with the pack.

So 160 cells won't work. That's not evenly divisible by 6, the smallest Volt BICM input size. With 16-cell modules the closest you could get is 144 cells.

Another thing to note is that while you can use multiple identical BICMs ( for instance two BICM 1's), the CAN messages will be duplicated. You'll get all of the cell data, it'll just be tricky to tease out which numbers belong to which BICM.

I'll shift the ID on the secondary set, that's not a problem. (Using a CAN man-in-the-middle type of thing)

I'm ok with 540V vs 600. However, what about wiring cells 160,161,162 to cell 160? (Let's assume we can configure it for 162 cells) Won't that work? EDIT: Things like the Orion BMS and wire up incomplete sets like this.

-Matt

 

[swoozle]

I'll shift the ID on the secondary set, that's not a problem. (Using a CAN man-in-the-middle type of thing)

I'm ok with 540V vs 600. However, what about wiring cells 160,161,162 to cell 160? (Let's assume we can configure it for 162 cells) Won't that work? EDIT: Things like the Orion BMS and wire up incomplete sets like this.

-Matt

Sorry, not following the last question. It sounds like you want to short circuit 3 cells in a serial pack into one input which I know I must be misunderstanding.
And what's that about the Orion BMS?

 

[mdrobnak]

Sorry, not following the last question. It sounds like you want to short circuit 3 cells in a serial pack into one input which I know I must be misunderstanding.
And what's that about the Orion BMS?

So, you've got the most negative terminal, then cell 1, cell 2, etc..up till cell 6 or 12. Each tap is previous voltage + cell X..

So, 0, 3.3, 6.6, 9.9, etc...

So if you have 4 cells, but 6 cell taps.. cell 4,5,6 would be wired to cell 4. Hmm but thinking about that then it would see cells 5 and 6 as 0V effectively.

The user guide for the Orion BMS says to to that:
"If fewer than 12 cells are populated in a group, unused cell taps must all be connected to the highest potential cell in that group. For example, if 6 cells are populated in a group, taps 6 – 12 all must be connected to the positive tap on cell 6.*"

Maybe it just needs to not have an open circuit is the thing. Not sure.

I'd laugh if LG / Chevy are using the same chips as Orion.

 

[swoozle]

So, you've got the most negative terminal, then cell 1, cell 2, etc..up till cell 6 or 12. Each tap is previous voltage + cell X..

So, 0, 3.3, 6.6, 9.9, etc...

So if you have 4 cells, but 6 cell taps.. cell 4,5,6 would be wired to cell 4. Hmm but thinking about that then it would see cells 5 and 6 as 0V effectively.

The user guide for the Orion BMS says to to that:
"If fewer than 12 cells are populated in a group, unused cell taps must all be connected to the highest potential cell in that group. For example, if 6 cells are populated in a group, taps 6 – 12 all must be connected to the positive tap on cell 6.*"

Maybe it just needs to not have an open circuit is the thing. Not sure.

I'd laugh if LG / Chevy are using the same chips as Orion.

Ah, I see. I never tried that. I have seen some info out there on the chipset that the Volt BMS is based. I forget how I got to it but I'm sure a little diligent googling would get you there.

 

[mdrobnak]

Turns out this is going to be academic for at least now anyway - I've pivoted on my battery pack voltage because I'm going to a different inverter that won't do 600V (but will do more current.), so back to 96 cells and not worrying about this - this time.