New Battery technology

[asulil]

well you can just use the same batteries that are in the Tesla they are mass produced and in bulk cost about $1 (see for example: http://www.alibaba.com/product-detai...702288953.html)

Also:

http://www.orbtronic.com/batteries-c...cell-ncr18650b

to replicate the 23KWh battery in the Karma you would need about 1,879 of these - that would weigh about 190 pounds.

[asulil]

well you can just use the same batteries that are in the Tesla they are mass produced and in bulk cost about $1 (see for example: http://www.alibaba.com/product-detai...702288953.html)

Also:

http://www.orbtronic.com/batteries-c...cell-ncr18650b

to replicate the 20.1KWh battery in the Karma you would need about 1,652 of these - that would weigh about 167 pounds. so theoretically 3200 of these would weigh only 340lbs and give you twice the range. And with Elon building his gigantic factory to turn out these little batteries - expect price to plummet even more. our battery technology may be better in some ways - but this battery technology will win the day - simply because it will (and currently is) cheap.

I am not suggesting that is as easy a sunday stroll - but theoretically there is nothing stopping anyone from buying the cells for $1 a piece - (plus you need controllers) stringing them together in a form factor that matches the karma battery (**** just use the karma battery take out all the modules) - I don't know about the software side and what you would need to do to get it to work.

[smoothoperator]

From a rough calculation of a spare battery enclosure I have laying around the shop I measure a volumetric capacity of 260L or so (without the wing portion). The Karma uses 315 cells I believe so this is a cavernous amount of space for such a few amount of cells (even taking into consideration the thermal systems necessary so that the pack can perform optimally). The pack that closest resembles the Karma pack is that of the Spark EV (21.3kwh, 336 prismatic cells etc). The interesting thing is the volumetric capacity of the Spark A123 pack is only 133 liters.

http://www.gmgoelectric.com/product/...batteries.html

Taking into consideration Fast DC provisions have been implemented in the Spark design it would not be unreasonable to assume that the C heat dissipation is adequate in the smaller enclosure and in the same ballpark as that of the Karma's enclosure.

IMO it would be the path of least resistance to use some variant of the Spark EV modules in the Karma enclosure to increase the all EV range. From a technical standpoint this would be the easiest retrofit barring a clean sheet pack re-design.

[Hockeydad]

Wow Smooth! You really know you're stuff. Now this is what I was hoping to get from this thread... some good sharing of thoughts to try and resolve a problem we will ALL have at some point. We can't depend on anyone making a new battery for the few thousand Fiskers on the road so we are left fending for ourselves.

Please, please, keep this conversation going, share thoughts and let's try and put our energies towards a solution rather than trying to find arguments of why it won't work.

So just to keep things on track... we're not looking to replace any software, temperature control system etc., we're simply trying to look at replacing the rechargeable power source leaving everything else intact. There could be issues of different voltages so an adapter/transformer may be required but that should be about it, as long as it can fit in the space.

Hopefully we can look at more power and range for the same weight or the same power and range but less weight which still means more range

Would someone be able to do a little experiment and hook up these batteries in a mini-pack, enough just to power a few electronical systems at first to see what issues and what needs to be done to connect the new power source to the car? Or is that not the right first step to do?

[Hockeydad]

Sorry, I mixed up the two responses...

Smooth is saying that we should use the Spark EV batteries rather than the Asulil's idea of taking Tesla batteries and creating our own pack.

So my first response should have been to Asulil's where I was wondering if there are members out there who could look at creating our own pack of Tesla batteries if we believe the Tesla batteries are superior to the Spark EV batteries and test a mini-pack to see if they can work with the Fisker Karma on a small scale basis and then go upwards from there, testing for heat etc and seeing if our current temperature control systems are adequate.

My second response should have been to Smoothoperator where I wanted to ask if we could hook up 2 of the Spark EV batteries together given we have almost the space to get two of them in there? Also, what's the range on one of these Spark EV batteries?

[smoothoperator]

First step would be to get some more in-depth technical information about the Spark EV battery. It is my assumption that it uses a variant of the AMP20 cell used in the Karma (7P3S layout).

The Karma consumes in excess of 500wh/mi but lets use this figure for a normal case scenario. A few assumptions first:

• We are able to make contact with someone with knowledge of Catia that is able to optimize the Karma enclosure for these modules
• These are the same modules that are used in the Spark EV. If this is the case, which is likely since the Spark EV uses 4 modules. With each module producing 5.38 kwh that puts the Spark at 21.52 kwh which is very close to the manufacture stated 21.3 kwh.
• The assumption that each module takes about 35L (1.2 Cu Ft) of space. If this assumption is correct then about 7.4 of these modules could fit in the Karma enclosure.
• If the above assumption is close to reality then it should be possible to get 39Kwh of power out of the pack (probably 36 or so usable). This will mean a range of roughly 72 real miles or 92 "Fisker miles". I read somewhere that it costs GM something like $400/WH so that would put the complete module price at $15k or so.

These projections and ideas are based heavily on assumptions which may or may not be valid due to the limited information I have on these packs. If someone is willing to do the legwork and collect more information I would be happy to help.

[asulil]

$15K seems like a lot. I am 100% all for discussing any idea on how to improve our cars - i think its great and am really appreciative of all the informed comments.
Using the spark system certainly seems more straightforward - but at $15K ? when you can buy the same amount of "energy" with the tesla batteries for around $5k.

[smoothoperator]

Originally Posted by asulil

$15K seems like a lot. I am 100% all for discussing any idea on how to improve our cars - i think its great and am really appreciative of all the informed comments.
Using the spark system certainly seems more straightforward - but at $15K ? when you can buy the same amount of "energy" with the tesla batteries for around $5k.

The ESS of a Tesla requires much more management and is incompatible with the Karma thermal system. The engineering cost to put such 18650 batteries in a Karma enclosure, along with the thermal systems that would need to be installed/engineered would be astronomical. The whole car would need to be re-engineered. This along with the new BMS and custom engineering and integration work needed to implement a completely different battery chemistry make it cost prohibitive to go that route. It would cost millions. From my experience it is best to use battery technology that is closest to what the car was designed for. We are talking about off the shelf modules used in other cars vs custom engineering a small form factor battery that Tesla has poured many hundreds of millions into perfecting. Heck even Rimac uses these A123 batteries and they are an extremely lean performance EV company that does most everything in house.

[valerun]

Originally Posted by smoothoperator

The ESS of a Tesla requires much more management and is incompatible with the Karma thermal system. The engineering cost to put such 18650 batteries in a Karma enclosure, along with the thermal systems that would need to be installed/engineered would be astronomical. The whole car would need to be re-engineered. This along with the new BMS and custom engineering and integration work needed to implement a completely different battery chemistry make it cost prohibitive to go that route. It would cost millions. From my experience it is best to use battery technology that is closest to what the car was designed for. We are talking about off the shelf modules used in other cars vs custom engineering a small form factor battery that Tesla has poured many hundreds of millions into perfecting. Heck even Rimac uses these A123 batteries and they are an extremely lean performance EV company that does most everything in house.

+1

you don't want to use Tesla approach. Among other reasons, check out YouTube videos of LiCo fires.

What we are doing is junking the gasoline side altogether and fitting Enerdel batteries into the original pack and the spaces that are cleared in the front. A quick photo of our project below. You can see a native pack on the right.

Replacing cells in the main pack might seem simple but it isn't. Couple of reasons:
1. Central BMS expects certain signals from the cell-level BMS boards. So you want to keep the same BMS electronics. This means you need the same cell count. This means that you are now constrained on the AH per cell. This means that you might not be able to realize any gain in pack size after all as cells come in very discreet capacity levels
2. cell-level BMS will be designed to fit the module that's in the car now. It won't fit (physically) on anything else.
3. If there is any difference between the chemistry used in the car and chemistry you are using, BMS will not be happy and may result in you not being able to utilize full capacity of your new battery
4. You would be lucky to achieve 50% space utilization of that 260 liters. Connectors, cables, cooling, BMS, cell hardware, etc. Datasheet for A123 cells used in Fisker. You will notice that at the cell level, these have ~250WH/liter density. If you could utilize 100% of 260 liters available, you would have a 65kwhr pack. That would be nice. There is a reason why it's not 65kwhrs...

My [well informed] assessment is that it will be very hard to achieve anything beyond 25kWHr usable in the native pack space without going to dangerous chemistries such as Tesla's.

V

[smoothoperator]

Originally Posted by valerun

+1

you don't want to use Tesla approach. Among other reasons, check out YouTube videos of LiCo fires.

What we are doing is junking the gasoline side altogether and fitting Enerdel batteries into the original pack and the spaces that are cleared in the front. A quick photo of our project below. You can see a native pack on the right.

Replacing cells in the main pack might seem simple but it isn't. Couple of reasons:
1. Central BMS expects certain signals from the cell-level BMS boards. So you want to keep the same BMS electronics. This means you need the same cell count. This means that you are now constrained on the AH per cell. This means that you might not be able to realize any gain in pack size after all as cells come in very discreet capacity levels
2. cell-level BMS will be designed to fit the module that's in the car now. It won't fit (physically) on anything else.
3. If there is any difference between the chemistry used in the car and chemistry you are using, BMS will not be happy and may result in you not being able to utilize full capacity of your new battery
4. You would be lucky to achieve 50% space utilization of that 260 liters. Connectors, cables, cooling, BMS, cell hardware, etc. Datasheet for A123 cells used in Fisker. You will notice that at the cell level, these have ~250WH/liter density. If you could utilize 100% of 260 liters available, you would have a 65kwhr pack. That would be nice. There is a reason why it's not 65kwhrs...

My [well informed] assessment is that it will be very hard to achieve anything beyond 25kWHr usable in the native pack space without going to dangerous chemistries such as Tesla's.

V

Interesting project!

Here is my take on your points:

1. From my brief conversations with A123 engineers I believe the pack hardware in the Karma is not proprietary to this vehicle and this architecture is used in all AMP20 Packs/Modules. The only thing proprietary is the shape of the battery enclosure and some of the external HV connectors. The Battery Control Module (BCM) is the same as on the Spark EV (with just a different flash than the Karma). It is possible to flash this component to correlate with the correct cell voltages, AH etc. A simple data dump from Canalyzer provides quite an informative look at how customizable the A123 BCM is.

2. For the actual modules you can replace (or flash) the monitor balance boards that correlate with the correct cell specifications. The Spark EV uses this same setup of MBB's inside the Cell Modules that communicate with the BCM. I think all that many are looking for is a ~35-40kwh pack. Anything else would make the car very heavy.

3. I think by replacing the Amp20 Prismatic pouches with higher capacity ext ones inside the modules itself may end up being the most cost effective solution.

I have been trying to find some information on the AMP20 Modules (i.e dimensions, weights, capacities, cooling etc). Do you have any information on these energy modules?