im here – sorry typo
regarding P section this is not white paper it was just some discussion about CLF
so this question:: “is more cells in P vs less cells in P higher risk for TR” remains open and any feedback welcome
thank you
and thank you Battery_Mooch who provided feedback in details
It makes me want to read more about it
wow i really need to go sleep but this is so addictive 
Is it draw or dump
If drawing than failed battery will draw up to own capacity and will stop – so the more cell in P the easier to handle the demand of current
Dumping is another story
as others cells dump as much as they can to the failed one than they dont carry about it what failed cell is able to handle or not
therefore the more cells in P you have dumping the current the faster failed goes into TR causing chain reaction – and the rest we all know
am i correct or wrong
You have to tell me, were you talking about the failing cell being forced into thermal runaway by the others? 
If yes, then having a larger p-group can provide more current due to the higher voltage under load for the larger p-group. Whether that causes a failing cell to go into runaway is another issue.
What is causing the failing cell to short-circuit the others?
i was considering situation when a single cell is failing/short regardless the cause
and trying to figure out would it be safer to have 3 times 12s2p vs 12s6p pack
without implementing other measure such as CLF for example
12s2p vs 12s6p is just example this could apply to 5 times 12s2p vs 12s10p
i may exaggerate but you should get my point
We need to define the situation in a lot more detail. Cells won’t fail on their own with a hard enough short circuit to draw enough current to force other cells into runaway IMO.
Forcing a cell into runaway by penetrating the cell with a nail creates an “artificial” short circuit that provides a continuing short circuit to help drain the other cells quickly. This would indeed help other cells go into runaway. Not always do so, just make it easier.
Having more cells in parallel allows for better heat transfer to other cells over the interconnect material but also means that any short-circuit current and cell heating gets divided too. I don’t think we can generalize and say that a bigger p-group aways = increased risk of additional cells going into runaway. Especially when we won’t have this forced short-circuit situation (nail penetration).
Any conclusions we might reach would have to be for the particular situation that we used for testing this. Having said this, here’s a paper that supports the more cell = bad idea…for artificially short-circuited cells, that is: https://www.osti.gov/servlets/purl/1235277
IMO, you’re going to have to do some testing to get the answers you want for the particular setup you’re interested in…CAREFUL testing.
2 Likes
Safer? That depends on many, many more things than the size of the p-groups.
Are there different cell interconnect failure points and types for those two configurations?
What is the impact on safety for the single BMS vs triple BMS setup? All that extra wiring and three BMS that all have to work properly, not overheat, not be affected by vibration/shock, etc.
Does shock/vibration affect the two configurations differently?
What are the different consequences of a single cell failure for the two configurations if nothing goes into runaway. That is, just one cell having excessive self-discharge. How will the multiple BMS’ interact?
How will the extra space required for the three BMS’ and their balancing leads affect reliability and safety?
IMO, do what 99.9999999% of what every commercial/industrial/aerospace/military pack does…parallel the cells and run off a single BMS. There are too many variables with unknown effects to be able to predict which setup might be best for you. Especially since you are only looking at forced external short-circuits of a type unlikely to be seen by an esk8 pack.
With good info not being possible to get for the setup and application you’re considering, and testing not very practical, I would recommend going with the tried&true configuration. The one used in millions and millions of packs.
Otherwise, start testing! This is the only way you will get the answers for the setup you want to use.
3 Likes
pleasure chatting with you
unfortunately we cannot think about specific situation while building the pack
we build them for esk8 and they got hammer and abuse – well not always
but we cannot bring specific scenario into equation
the cells get short – due to neglect of build — human error is mostly the cause rather than material
there is a need to find the way to minimize the error – even the most careful build may fail
BMS will not be in the pack as used for charging only
Ahhh…I strongly disagree. A good pack design considers a large number of specific situations and then the builder decides whether to take those into account or not when designing the pack.
High/low temperatures, degree of water ingress, charge voltage and current, discharge voltage and current, vibration/shock magnitudes, frequency, and directions, ease of repair, etc., can all be considered.
What causes the changes to any of those things are irrelevant though. I don’t care what causes a temperature change or a particular physical shock. I don’t care why the voltage at the charge port is too high. I only care what is happening to the pack itself and whether I want to do what’s necessary for the pack to survive that particular abuse.
I agree though that human error, i.e., pack design/assembly that was done poorly or that did not anticipate the conditions the pack would actually operate in, is by far the more common situation. Crappy BMS’ and crappy harnesses are sold though, along with other things that can cause cell failure.
Minimizing the possible human error just requires a builder who cares and who has learned enough. Not always easy to do though.
4 Likes
100% agree
1 Like
Heat is indeed the enemy and we should always consider pack temperature. But for many packs, and many builders, the temperatures the cells will reach without any type of cooling just isn’t a concern. Either because the cells never get too hot, the builder just doesn’t care about cycle life, or the builder isn’t even aware that the pack might get too hot.
Heat-sinking a pack is very, very difficult and expensive. Most esk8 packs just don’t need it though IMO due to them being pulsed most of the time rather than run continuously (for most riders).
You can easily use a heat spreader in a pack, some sort of material to spread the heat more evenly, but getting the heat out of the pack (having a heat sink) is tough to do. The pack needs to be thermally coupled to the outside of the enclosure and that is not easy to do properly. As you mentioned, water ingress is an issue along with the amount of effort needed to get the heat from all the cells fairly equally to the outside of the enclosure.
That costs a lot of money and takes up a lot of room. It is often easier to just add more cells in parallel and/or limit the battery current in the ESC.
2 Likes
Just gonna jump in here and say that I build esk8 and other PEV batteries for a living, and every build I make is tailored for a client’s specific use case.
In my opinion if anyone is building batteries for someone without adequately understanding their use case, then they are not doing their due diligence and are potentially making an unsafe pack.
The range of applications for lithium batteries are so diverse that it’s impossible to say if one pack vs another will be safe without knowing the use case.
Take, for example, powerwall packs vs esk8 packs. A powerwall can often get away with no fish paper, no vibration isolating, sharp corners on nickel strips, etc., simply because they dont move during normal use. Where as our esk8 packs have to be built to a very high safety standard, taking into account high-vibrations and micro/macro flexing.
A battery which might be totally safe sitting in your garage storing the excess power from your solar panels would be a fireball waiting to happen if you were to put it in an esk8.
/rant
Anyway, thats all I have to contribute. Everything else, Mooch said better haha. Listen to him, he is smarter than 99% of folks on this forum when it comes to batteries.
5 Likes
Thanks for the reply. I didn’t realize that the Trampa is a single controller. On the site it sounds like Bejnamin Vedder helped build the thing:
“TRAMPA BOARDS LTD has teamed up with the internationally renowned ‘YODA of electronics’ (Benjamin Vedder) to bring to the world the most advance electronic speed controller ever to be produced!”
Now that I’m looking at it closer it looks like the Makerx-tech GO-FOC HI300/HV200 75V/100V 300A/200A ESC is also a single motor ESC, is that correct?
I need a dual motor ESC as my scooter has two motors. Is the Stormcore 100D the only VESC ESC available that will handle dual 1,600 watt continuous motors on a 72 volt system?
yes, its just a single
100D can handle upto 84v (18S) safely, so I would expect it to run fine with just 72v. unless u HAVE to run it at 72v, or else spintend ubox is also a great option, but it can only do 75v (<80v for peak)
This isn’t a true statement; you can always use two single motor ESCs.
3 Likes
Make sure you know what battery you have, because often folks use “xxV system” to not mean the actual max system voltage.
The best way is to say “20S” or “17S” — both of those can be called “a 72V system” whatever that even means. Easiest way is often to look at the charger, it will frequently have the real max voltage listed on it, and not just an average estimated number.
5 Likes
yes this also, i almost forgot
3 Likes
I have two sensored motors but I really don’t need them, I’ve programmed them to run sensorless…
Should I just leave the sensors disconnected from the vesc or should I desolder them off the motor’s mainboard?