Playing with A123 20Ah lithium cells

[bwilson4web]

Your efforts really helped me an in particular:

At 1C dis/charge I saw a pretty constant 4F above ambient in cell temperature. The cell heated up very uniformly. The terminals I thought would heat up as well, but they always remained at a lower temperature than the cell. This may be due to the large aluminum blocks on them.

I'm going to treat the tabs as heat radiators as much as possible. Compared to the poly bag, they should do a reasonable job of dumping the latent heat from the battery. When I get my cells, I will test this by using an IR thermometer to see if there is any detectable temperature gradient from the tabs down. The video suggested this approach for heat management.

I have a boat-load of things to test but I suspect that if I group balanced cells, I can detect the earliest dV/t drop. I may also go with the Prius trick and measure voltages in pairs which a single battery monitor chip can handle. I'll still go with the per-cell, Zener balancing array but the voltage monitoring is just to disable the string when the first cell pair begins to fall off the discharge edge. <grins>

Every time I can simplify the design, the parts count and risks go lower. <grins>

Bob Wilson

[bwilson4web]

Hi,

Originally Posted by weazel

u should drop using any battery there all rubish id go for supper caps . . .

I had looked at "supper caps" (ultracaps?) before but found there were problems:

1. voltage energy storage device - to extract energy, we have to convert the stored ultracap voltage into usable voltages and currents. In my case, a pretty massive DC-to-DC converter that generates 48VDC at 15A (720W max) in addition to the ultracaps. Furthermore, the DC-to-DC converter will have to deal with either small voltages and massive currents or large voltages that have a huge swing. Let me give an example for my 720W maximum load:
   • 2.5V (1) -> 288A
   • 5.0V (2) -> 144A
   • 12.5V (5) -> 57.6A
   • 25V (10) -> 28.8A
   • 50V (20) -> 14.4A
   • 125V (50) -> 5.76A
   • 250V (100) -> 2.88A
   • 500V (200) -> 1.44A
   • 1,250V (500) -> 0.58A
   The wide voltage swing means a DC-to-DC converter that already will be running ~90% efficiency ... Opps just lost 10% of the energy ... that has to have a very wide input and close to discharge, some massive currents. They can be built (look for inverting switching regulator) but in this energy and voltage range, pretty expensive.
2. energy density - most of the ultracaps I've seen are cylindrical. This means intra-cap space of cylindrical units is unusable. In contrast, these poly-cell batteries are flat, stackable plates so we'll get a minimum sized power pack which is important on a space limited, electric bicycle. Worse, we're looking at .7-1.8 MJ/kg versus .56-.1 for supercapacitor which means a heavier ultracap for the same amount of energy. My electric bike needs to be lighter, not heavier.
3. series ultracaps voltage balancing - the cheap solutions, accurate resistors, work by losing energy. All other solutions are terribly expensive.
4. surface versus volume energy storage - all caps require two surfaces with a dialectic between them and store the energy in the voltage field. In contrast, these lithium iron phosphate batteries store their energy inside the mass of the electrodes, not a surface layer.

To contrast, these lithium iron phosphate batteries have:

• 90-100 Wh/kg
• 220 Wh/L
• 3.3V constant discharge voltage

So for ultacapacitors, we are looking at:

• 0.5 to 30 Wh/kg
• 0.33 to 3.89 kg/L -> 117 Wh/L (best case)
• 0-2.5v per ultracap

There are applications where an ultracap makes sense. It has extremely high discharge rates but that is not needed in my eBike application. If I were making a 'rail gun,' an ultracap would be the way to go. They might also work with a battery to provide a high start up energy surge that these batteries don't do as well in providing.

Bob Wilson

[Daox]

Ack, how could I forget. Eric also sent me this link not too long ago. Mike Dabrowski (Insight guru, MIMI inventor) has also been tinkering around with these cells. Hes has some excellent testing results on his site including thermal testing. As you can see, they heat up very uniformly, and the tabs do tend to stay cooler than the rest of the cell.

http://99mpg.com/workshops/designingandbuildi

[bwilson4web]

LOL! I love Mike's work since I first saw it in 2006. The last time I checked, he was modifying a Prius transaxle to be a rear-drive, electric motor for his Insight ... replacing the 5th wheel. <grins>

Originally Posted by Daox

Ack, how could I forget. Eric also sent me this link not too long ago. Mike Dabrowski (Insight guru, MIMI inventor) has also been tinkering around with these cells. Hes has some excellent testing results on his site including thermal testing. As you can see, they heat up very uniformly, and the tabs do tend to stay cooler than the rest of the cell.

http://99mpg.com/workshops/designingandbuildi

Awe man! He has great toys!

One thing he brought up that is obvious once you think about it ... compression of the polypaks. But I didn't see a plot of compression force vs. cell capacity.

Well I had another insight this afternoon (Ok, old man finally remembered something) about my eBike load:

There are two potential loads:

• 350W/48V = 7.29A - rated power current
• 15A "Maxmum current"

I've used the maximum current for all of my earlier, draft calculations but at 350W, we'll see half the current load. But as for the A123 cells, they are going to have an easy life.

The other thing to share is I bought my first airplane, a Cherokee 140, and a Yamaha 550 motorcycle to drive to and from the airport. I was always going to be a Visual Flight Rules pilot which meant if it was nice enough to fly, it was nice enough to ride the bike. Although I had a full set of leathers, motorcycles and bicycles don't do well on snow and ice. Realistic temperature range:

• 40F (4C) - assumes gloves and layers under rain suit
• 100F (38C)

If I remember correctly, Mike lives in the NorthEast and has to deal with greater temperature swings than I do. Combined with the relatively light load of the eBike, this should be an awesome combination. <grins>

In bad weather, I'll have to take one of our Prius or stay home.

Bob Wilson

[bwilson4web]

Sad news. Not only is there a delay but I was offered not only some blemished cells but worse, short tabs with one looking like someone had punched holes to the point of looking more like a sieve . . . too much metal removed.

My eBike application means vibration, motion and temperature changes. I am not willing to take a blemished cell unless there is some reasonable replacement for early failure and the tab with all the holes has too much metal removed.

I have a boat-load of NHW11 modules and seven would give 50.4V stacks close enough to the 48V lead-acid stack. If the bike controller doesn't blow up with 56V, I can fully charge them and I'll have 5.5Ahr per stack.

Alternatively, six modules fully charged to 8V would give 48V but discharged to 6V it would be 42V, close to the 41V cutoff voltage of the eBike motor controller. Not as good as the A123 but 'less bad' than the lead acid stack.

The nice thing is I can test the Zener balancing and battery management approaches without having to risk the more expensive A123 cells.

Bob Wilson

[Daox]

That is too bad. I specifically asked when I heard of my cells if they had full tabs or not. I too wasn't willing to dink around with blemished cells.

In other (IMO) sad news, looks like A123 will be majority owned by the Chinese pretty soon. Too bad nobody from the US (or any other country who doesn't blatantly steal from the world) was willing to plunk down any money on some of the most advanced battery technology available. I suppose at least this time they're actually buying it.

Wanxiang To Invest In A123 Systems | Hybrid Cars

A123 Systems announced yesterday that it has signed a non-binding memorandum of understanding (MOU) with Wanxiang Group Corporation establishing the framework for a strategic investment through which Wanxiang would invest up to $450 million in A123.

A123 Systems saved by the Chinese | ChargedEVs

Wanxiang will provide a bridge loan of $75 million or so, and eventually invest up to $450 million in A123, which would give it an ownership stake of about 80% in the company. Wanxiang’s global empire boasts some $13 billion in revenue and over 45,000 employees.

[oil pan 4]

I ordered 8 AMP20 cells off ebay after I figured out A123 and their authorized dealers to be useless, apperantly they dont sell batteries to you and me.
No wonder they are bankrupt. They earned it.

The idea is to fit 8 of these cells into a 30cal. ammo can and make a sub 10 pound 40 amp hour, 900 cranking amp battery to replace my weakened 20 round cell 12 amp hour battery that use to put out about 300amps.

Now I will have enough power to run the welder for a few minutes or kick the air compressor on once or twice.

My question is how much compression should I try to load the cells under so they stay together?
I am sure they wont like floping around inside their housing.

I was planning on soldering the tabs together, any one try this yet?

[Daox]

I don't think you need much compression at all, just enough to keep them from swelling. They aren't going to be super taxed (high loads for extended durations). If they're properly cared for they shouldn't even swell I believe. Its only upon over-charge or over-discharge that they become a problem.

I doubt soldering will work. The one tab is aluminum and the other is copper plated with aluminized copper I believe. I'd also be very careful about heating the cells up.

[oil pan 4]

I plan on using them teamed up with my 28 round cell battey I made about 6 months. I will have 2 strings of AMP20 cells with 7 strings of round cells to:
Run glow plugs for up to 10 seconds, 100 amps
Crank the diesel, 350 amps
Occasionally use the intake heaters for 30 seconds up to 1 minute, 100 amps
Some times i use the intake heater and glow coils together.
The heaters are on manual controls so the batteries wont have to crank with heating elements on.

I plan on installing a charge limiting resistor on a time delay, mostly to protect the alternator from the full load the lithium batteries tend to put on it after start up.

I have a 12 watt soldering pen on a varrible voltage controler, I usually run it about 6 watts, I can keep the heat down and use heat sinks.

[oil pan 4]

Here is what happens when they are used with out a compression load to hold them in place:
http://ecomodder.com/forum/showthrea...tml#post338583