Re: [EVDL] BMS questions

[John Lussmyer via EV]

Though, how will this setup works will depend on what kind of failure 
happens.


One possible configuration would be with Big Cable cross connects at one 
end of modules, and little cross connects between the 12 cells of the 
module:


A) Single Cell Opens - in either module - All pack current comes from 
one module  It runs down faster, though in the process of doing so, the 
other cross connected cells will end up trying to charge the ones from 
the discharging module, and all of the small connects will end up fried.


B) Single Cell Shorts - in BMS connected module, will set off alarm.  
Good.  In other module - once again cascading melted cross connect wires.


Without Big Cable cross connects - i.e Strings are only connected at the 
ends. (other than the little cross connects)


A2) Single Cell Opens - all current flows through other string. As 
string voltages diverge, little wires start frying.  If the open cell is 
on the BMS side, it will get detected once the relevant cross connect 
fries.  If on the non-BMS side, we're back to cascading melted cross 
connect wires.


B2) Single Cell Shorts - BMS side detects it quickly.  Non BMS side - 
cascading melted wires.


Thinking about the more-likely scenarios - different modules at 
different temperatures resulting in small differences.  I think that it 
still runs the chance of melting the cross connects during heavy 
acceleration.  Note that each "cell" in this case is a 200AH unit which 
is actually 20p.


On 10/9/2022 1:09 PM, (-Phil-) wrote:
Well, it depends on how closely matched the packs are.  If they are 
good quality and closely matched, then you don't really need the cell 
interconnects between the 2 strings to be all that stiff.   Consider 
that most BMS balancing loads are 100ma or less, it's unlikely to have 
much more than that flowing between the 2 strings unless there is an 
actual failure.  I've used silicone-insulated 20awg wire for this on a 
pack before.   In the event of a short, the 20awg will fuse and the 
silicone will not burn and allow the conductors to melt through.  If 
you have an outright failure then this could happen, but otherwise the 
resistance of the 20awg interconnects will limit the cross-flow 
current during high pack draw, but allow them to still cross-balance.



___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

Re: [EVDL] BMS questions

[(-Phil-) via EV]

Well, it depends on how closely matched the packs are.  If they are good
quality and closely matched, then you don't really need the cell
interconnects between the 2 strings to be all that stiff.   Consider that
most BMS balancing loads are 100ma or less, it's unlikely to have much more
than that flowing between the 2 strings unless there is an actual failure.
I've used silicone-insulated 20awg wire for this on a pack before.   In the
event of a short, the 20awg will fuse and the silicone will not burn and
allow the conductors to melt through.  If you have an outright failure then
this could happen, but otherwise the resistance of the 20awg interconnects
will limit the cross-flow current during high pack draw, but allow them to
still cross-balance.

On Sun, Oct 9, 2022 at 9:51 AM John Lussmyer  wrote:

> I can get a cable in to do cell taps, but running a bunch of cross-module
> interconnects that can carry high current is unlikely.
> I'd need to use something like 16ga interconnects for between
> module/cells.  Then would I need to cross-connect with big cables at each
> module pair, as well as cable them in series?
> On 10/9/2022 9:23 AM, (-Phil-) wrote:
>
> If it's sealed with no cell taps, how can you build a proper BMS?
>
> On Sat, Oct 8, 2022 at 8:09 PM John Lussmyer via EV 
> wrote:
>
>> Can't do in this case.  Each module is a sealed box of 12s.
>>
>> On 10/8/2022 5:55 PM, (-Phil-) wrote:
>> > The best way to do a parallel setup is cell buddy-pairs if possible.
>> > There is no problem doing it this way.   Could also "cheat" with 2
>> > separate strings by tying the cell taps together as stiffly as
>> practical.
>> >
>> ___
>> Address messages to ev@lists.evdl.org
>> No other addresses in TO and CC fields
>> HELP: http://www.evdl.org/help/
>>
>>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

Re: [EVDL] BMS questions

[John Lussmyer via EV]

I can get a cable in to do cell taps, but running a bunch of 
cross-module interconnects that can carry high current is unlikely.
I'd need to use something like 16ga interconnects for between 
module/cells.  Then would I need to cross-connect with big cables at 
each module pair, as well as cable them in series?


On 10/9/2022 9:23 AM, (-Phil-) wrote:

If it's sealed with no cell taps, how can you build a proper BMS?

On Sat, Oct 8, 2022 at 8:09 PM John Lussmyer via EV 
 wrote:


Can't do in this case.  Each module is a sealed box of 12s.

On 10/8/2022 5:55 PM, (-Phil-) wrote:
> The best way to do a parallel setup is cell buddy-pairs if
possible.
> There is no problem doing it this way.   Could also "cheat" with 2
> separate strings by tying the cell taps together as stiffly as
practical.
>
___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/


-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

Re: [EVDL] BMS questions

[(-Phil-) via EV]

If it's sealed with no cell taps, how can you build a proper BMS?

On Sat, Oct 8, 2022 at 8:09 PM John Lussmyer via EV 
wrote:

> Can't do in this case.  Each module is a sealed box of 12s.
>
> On 10/8/2022 5:55 PM, (-Phil-) wrote:
> > The best way to do a parallel setup is cell buddy-pairs if possible.
> > There is no problem doing it this way.   Could also "cheat" with 2
> > separate strings by tying the cell taps together as stiffly as practical.
> >
> ___
> Address messages to ev@lists.evdl.org
> No other addresses in TO and CC fields
> HELP: http://www.evdl.org/help/
>
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

Re: [EVDL] BMS questions

[John Lussmyer via EV]

Can't do in this case.  Each module is a sealed box of 12s.

On 10/8/2022 5:55 PM, (-Phil-) wrote:
The best way to do a parallel setup is cell buddy-pairs if possible.  
There is no problem doing it this way.   Could also "cheat" with 2 
separate strings by tying the cell taps together as stiffly as practical.



___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

Re: [EVDL] BMS questions

[(-Phil-) via EV]

The best way to do a parallel setup is cell buddy-pairs if possible.  There
is no problem doing it this way.   Could also "cheat" with 2 separate
strings by tying the cell taps together as stiffly as practical.

On Sat, Oct 8, 2022 at 4:57 PM John Lussmyer via EV 
wrote:

> I've been looking into both Thunderstruck and Orion BMS docs.
> One interesting doc is from Orion - describing the possible issues with
> parallel strings (which I will have), and some possible mitigation
> strategies.
> https://www.orionbms.com/manuals/pdf/parallel_strings.pdf
> I'm thinking about the one on page 14 - Paralleled strings with two
> separate DC busses
> Since I'll have 2 strings, this would entail a couple more contactors
> and some BIG diodes.
> a pair of 300A+ ones for the discharge side, and a pair of 50A ones on
> the charge side.  (probably about $100 of diodes from digikey)
>
> Though, I'm not sure i'd use contactors on the charger side, instead I'd
> just have it shut the charger off and notify me that it happened. (hmm,
> might allow a single string to shut off - but if the 2nd string does it
> would shut the charger down instead.)
>
> An Orion (pair) could shutdown each side independently, the
> Thunderstruck can probably only shutdown one side.
>
> Any thoughts on either system, or the parallel string setup?
>
> ___
> Address messages to ev@lists.evdl.org
> No other addresses in TO and CC fields
> HELP: http://www.evdl.org/help/
>
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

[EVDL] BMS questions

[John Lussmyer via EV]

I've been looking into both Thunderstruck and Orion BMS docs.
One interesting doc is from Orion - describing the possible issues with 
parallel strings (which I will have), and some possible mitigation 
strategies.

https://www.orionbms.com/manuals/pdf/parallel_strings.pdf
I'm thinking about the one on page 14 - Paralleled strings with two 
separate DC busses
Since I'll have 2 strings, this would entail a couple more contactors 
and some BIG diodes.
a pair of 300A+ ones for the discharge side, and a pair of 50A ones on 
the charge side.  (probably about $100 of diodes from digikey)


Though, I'm not sure i'd use contactors on the charger side, instead I'd 
just have it shut the charger off and notify me that it happened. (hmm, 
might allow a single string to shut off - but if the 2nd string does it 
would shut the charger down instead.)


An Orion (pair) could shutdown each side independently, the 
Thunderstruck can probably only shutdown one side.


Any thoughts on either system, or the parallel string setup?

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/