Playing with A123 20Ah lithium cells

[Daox]

Yesterday, I picked up a few A123 20Ah lithium cells while I was at Drive Green Expo in Madison, WI. I have been wanting to tinker around with these batteries for a while now.

http://www.youtube.com/watch?v=87Xak65iT84

For those who don't know, these cells are very impressive. These are the batteries that are being used in the Fisker Karma. They can handle 10C (200A) continuous discharge and up to 22C (450A) burst discharge (depends on SOC). Most other prysmatic lithium cells can only handle 3-4C continuous discharge and around 10C burst discharge. The mottcell lithium cells I have in my Prius' plugin kit which you'll see below are 40Ah batteries and only capable of I believe 2C continuous discharge. Through my testing with them I don't even really know if I would say they can do that. Even at 1/2C continuous charge/discharge they get pretty warm.

These cells weigh in at just under 1.1lbs. My Mottcell cells are roughly 3.5lbs. However, I will need to add provisions for assembling and mounting the A123 cells so it will increase their weight, but I shall revisit that when I get there.

A123 amp20 data sheet


So, here are a few images I took so you can get an idea of the size.


Here is the packaging they come in. I got four cells (enough to make a 12V battery).




I didn't have a soda can handy so I used a soup can for scale.




They're very thin!




Length.




Width.




Thickness.




This is the 20Ah A123 cell compared to my 39Ah Mottcell prysmatic cell.








I also measured the tabs. They are 1.775" wide x 1" tall. The positive terminal is aluminum and is .010" thick. The negative is plated copper and is .007" thick.

So, I have a 12V 20Ah A123 battery essentially. I'll be messing around with it a bit, developing a connecting method. Eventually I think I'll probably put it into the Paseo as the starting battery. It should weigh about 5lbs vs the current group 24 lead acid battery that is in there. While I won't be able to use it as an alternator delete battery (not enough capacity, but the wife hasn't been plugging it in anyways, and she has been driving it for the past year), it will lighten things up a bit. Funny too, the A123 cells really didn't cost a ton more than the lead acid deep cycle. If I wanted the same usable capacity I'd need 8 more lithium cells which would make it cost quite a bit more.

[Ryland]

I think this holds true for A123 system cells the same as it does for all other lithium cells, that you need to keep pressure on them so they layers don't fluff up and separate, bolting an aluminum plate to either side seems to be the popular way to do it.

I'm blown away by how cheap lithium batteries are now!

[Cobb]

Wow, looks a lot like those being sold for replacement packs for the first gen insight. I think a guy was selling 50 or 55 for a 1500 bucks.

Any reason you arent going to use these in place of your accessory batter in your prius? If your prius charging system works like the Honda IMA it should cycle off and on less frequently and hold a better charge than any lead acid battery. Most Hondas use the ELD to cycle the alternator or charging system off and on as needed vs a constant like on a typical altenrator use vehicle.

I use to charge my accessory battery before driving my Honda as a kind of mini test of the benefit of a phev and for the first few miles I had great fuel economy till that charge wore off.

How about replacing the pack in your prius with a few of those?

[Daox]

Yeah, that price sounds about right. I paid $27.50 per cell for these four cells.

While those are all nice and fancy options for using these cells, that won't be the case. The Prius' 'starting battery' is simply a small AGM to power up the electronics before the high voltage pack gets connected. To replace that small battery with these lithium cells would be a waste when I can replace a large group 24 battery in my Paseo and save quite a bit more weight. And, while replacing the 6.5Ah cells in the Prius' hybrid pack would be fun, it would be a huge pain too to convert to a new battery chemistry (thinking about the Prius' BMS). Couple that with the cells alone would cost almost $2k and I'm just not seeing that being worth it.

At some point in the future I would like to make an EV or DIY hybrid. I'm messing around with these cells because they are what I would like to use should I ever get around to it.

[ecomodded]

The battery packs would be good for portable use as well, for bicycling lights or camping or extending the laptops power.

Love the Li-ion battery technology, I just bought six 3.7 volt 18650 batteries for my flashlight that pump out 2 amps a battery, the flashlight uses 3 18650 to power 3 Ultra bright 12 watt Cree xm-l LED's.


Those A123 batteries remind me of the lipo battery packs, but bigger.

[Daox]

I've started designing how I want to connect the individual cells up. Keep in mind that I am planning ahead to use these on a hybrid or EV project so whatever I come up with must be scalable.

The cells are basically sandwiched together and wired in series. There are alternating aluminum and plastic spacers between the tabs. The whole string of cells is bolted together on each side with non-conductive (plastic) threaded rod. The aluminum spacers can be tapped for small screws, or drilled for a bananna type connector for BMS connections. In the case of my 12V, I won't need a BMS while using it as a starting battery. This setup keeps the weight down and utilizes all the surface are of the cell's tabs to keep resistance down and allow for high amperage draw.

I am still working on how I want to sandwich all the cells together, but I'll probably have some plates on each end to hold everything together. Some people tend to go overboard on this and I don't think its the most important thing unless you are really pushing the batteries.

I'll also have to add some sort of automotive type lug to the aluminum spacers on the end so I can connect it up to the car.

[user removed]

At that price I might consider using them in my vehicle. I already have a 9 HP Honda Diesel engine acquired for the project. How much would they cost per KWH?

Since they would only be on-off to generate hydraulic pressure, could I avoid having to use a BMS?

regards
Mech

[Daox]

Well one cell is 64 Wh, so 16 cells to get 1kWh. $27.50 * 16 = $440 / kWh.

You only really need a BMS if you want to come anywhere near close to either end of the charged/discharged state. Using it as a starting battery I know this won't be the case. The alternator will only charge it to roughly 50% (I'll be verifying this with testing that I will post later), and starting sure won't take that much power and then it will be topped off when the engine starts. If you can ensure that you aren't going to cycle it deeply at all as is the case for me, you can get away without a BMS. For anything else I'd suggest using one. At the very least you need a way to monitor the cells and watch it so they don't charge up too far or discharge too low. This is a pain and it is how I used to do things on my PHEV kit on the Prius. It gets old fast and you'll want something automated IMO.

[Daox]

Speaking of testing, these are my plans. I will test these cells with my powerlab6 charger. This can charge/discharge at up to 40A (2C for these cells). I will do charge/discharge cycles on these cells at .5C, 1C and 2C. I'll post up the graphs so you can see the voltage sag etc. I'll also post some graphs from my mottcell cells to compare.

[user removed]

In my application the cells would only run a motor to pump up hydraulic pressure. There is no need to monitor or control the power level since there is no necessity to control vehicle speed. The motor would be switched on and off by pressure level switches in the accumulator.
Charging would be strictly when the vehicle was sitting still from the 220 outlet in my garage.
How far the pack was discharged would determine when the Diesel engine took over. I would not use it to recharge the pack.
Similar to the Volt.
Right now I plan on using only the 9 HP Diesel, but if a battery pack was within a reasonable price range that plan could change, even to pure electric propulsion, but in either case there would be no electrical regeneration at this point. It would be purely hydraulic.
The useable capacity of a 10 KWH pack would be useful information. At $4.4k it looks interesting, but I may build the pure Diesel version first, then an electric option later.

regards
Mech