Converging on General Solution BMS features

methods

1 GW
Joined
Aug 8, 2008
Messages
5,555
Location
Santa Cruz CA
I have lost 2 handfuls of hair at this point trying to solve the unsolvable riddle.

All the obvious - budget, complexity, timeline, compatibility, feature creep

* I need to be compatible with "advanced components" that over-reach... such as the Sevcon Gen 4... which controls its primary contactor. In cases like this... I think my best bet is to incorporate the ability to drive a primary contactor, but not require it, such that I can communicate protection events in the form of CAN messages, throttle inhibit, or indicator. I believe the BMS should ultimately have its own contactor in every case (in order to make the battery system true stand alone) - but its not my job to ENFORCE this. Really all that is lost is a couple hundred bucks and some volume due to redundancies in the system. All aerospace is covered in redundancy... it just has to be... OEM avoids this in the cost race... but we are solving a new riddle now.

* Our market has a huge range of options... looking just at controllers... some are incredibly dumb. Some incorporate basic pack level LVC/HVC. Some do all sorts of weird I/O... I need to support the best and the worst... and allow the customer to drive down cost through the systems approach.

* Our market covers a HUGE range of voltage... from 12V to 1200V... that is like... saying I want to build a carburetor... that will work on a 5HP briggs and stratton as well as a Dodge Viper V10. Not a perfect analogy... but the problems encountered (mostly budgetary) are very big hurtles.

* Everyone has already solved the specific riddle... the solution to THEIR system. There are hundreds of BMS solutions out there ... all totally proprietary. Those will be useful in the future... but what I am looking at is 5, 10, 20 years down the road. Junkyards are going to be PILED with old components. We will desperately want to leverage them. Controllers, chargers, BMS units, batteries, DC-DC's, pumps, vacuums, lights, harnesses... think of it as a subsidized flea-market paradise. Huge companies are dumping 10's of millions into high-end, race to market, solutions. In pretty big quantities... and all in a very adversarial way. No way in hell is a simple black box going to easily make a Tesla primary controller work with a Leaf motor with a GM battery with a Porsche BMS... just aint going to happen... even with CAN. There are dozens if not hundreds of small signals, dependencies, and incompatibilities. Of course we will eventually converge on a set of standards.. . but that is a LONG WAY OUT... and a VERY FAST MOVING TARGET... as tech develops. We cant bound it now - we dont even know what we are trying to bound. New batteries, new parts, new motor designs, new charging standards, new power levels... changing changing.

Doing my best to try and make something that will support DIY in a cost effective and non-capturing way.
I want to work WITH all that junk... not work against it or try to capture the market in any way.

Farfignugin.

At least I have settled on a few things:

* We will eat the cost of IP66 and just make one flavor - the one that can be zip-tied to a bike
* We will support nearly any voltage... in one flavor... to work with changing requirements and legacy gear
* We will support any Zero bike design, even pre 2012... as these batteries and components reach end of life
* We will absolutely not kill batteries... even if we do suck some juice during operational periods
* We will try our damnedest to pick interface features up front and not change them... such that as V2 and V3 roll out... that they are plug and play compliant with at most, pin swaps.
* We will try to actually make it affordable - on the scale of $10/series cell (good luck... and certainly no way to recoup engineering investment)
* We will work out in the open for others benefit. There is less honor in my personal win than in a group win. If the group wins, I win... so its a WIN WIN and not a Zero Sum Game.

Bills to pay.... bills to pay... stay on topic methods :|

On that note... a rad friend came up with a nice side project for me that will put gas in the car. Arduino's and Androids and fans and resistors and temp probes... sweet. Get paid to do super fun hobby stuff.

Wasn't that the dream I was trying to live back in 2007?
Damn... got here... grass is no greener :oops:

-methods
 
Back
Top