Playing with A123 20Ah lithium cells

[bwilson4web]

Originally Posted by Daox

The aluminum and aluminized copper tabs are simply extensions of the anode and cathode. You might find this video interesting...

http://ecomodder.com/forum/showthrea...deo-15502.html

I haven't seen it yet but now it makes sense. The higher, electromotive potential of lithium protects the aluminum. . . . makes a lot of sense.

I'm still reading up on aluminum spot welding and it is not as easy as I first thought but not impossible either. My hypothesis is a rectangular, tungsten tip joining on the edges of two sheets will give the heat concentration needed to make a nugget joining the two. A capacitive discharge, short-length, welder should then spot weld the metals together. Testing will prove the approach.

One other interesting approach is to use a pair of sheet steel strips, "|<tabs>|", around the two tabs and spot weld the strips "|<tabs>|". The steel provides the resistance and heat to hold and form a proper weld puddle between the aluminum and copper. Certainly a clever approach but I have to admit I'm looking back at your mechanical clamp with more insight.

Driving home, the answer came, aluminum press rivets:

• Equalize cell charge
• Clean tabs
• Fold tab over end conductor
• Rivet and/or clamping machine screws (solder if practical)
• Align second tabs
• Fold tabs
• Drill first rivet hole
• Press counter-sunk rivet
• Run balance wire around rivet and out side
• Drill and seat rest of rivets.
• Stack next cell and repeat riveting.

Pressing the counter-sunk, aluminum rivets provides a permanent joint. The rivet shaft diameter determines how much current each can carry, not counting the current through the mated surfaces.

A 25-pin D connector can easily handle the 16, sense wires. For cell balancing, there is no need for more than 1 amp. There will still be spare pins to operate the isolation MOSFET.

Bob Wilson

[redpoint5]

Originally Posted by Daox

You could certainly replace your lead acid battery with these cells. I don't think you'll ever come out ahead financially though, if you do you'll at least be on your next vehicle by then. Thankfully lithium batteries last longer than lead acid, so this might actually be possible...

Thanks for the great response, especially the attempt to put numbers to this project. I've been keeping my eye out for a decent clamp-meter on Ebay, but it seems DC amps aren't all that common on these devices and the ones that have the function are pricey.

I'll be converting some of my lights to LED to save some watts and improve safety and reliability. Still, I imagine you are right about a luxury car consuming more power.

I'd like the project to pay for itself eventually, but it's not the only reason I want to do it. Learning about the technology and tinkering are fun for me and worth the price.

I've also got an aluminum frame from an electric reverse-trike that I built in high school that I've considered digging up again. It was built to compete in a 1hr distance race against other schools and used lead acid batteries. These cells seem like the perfect power source for the car if I decide to build it up again. I for sure would build it if I could make it street legal for my commute. It would do 55mph with a 2hp electric motor!

[bwilson4web]

This is well worth watching:

Originally Posted by Daox

The aluminum and aluminized copper tabs are simply extensions of the anode and cathode. You might find this video interesting...

http://ecomodder.com/forum/showthrea...deo-15502.html

Further in the video he goes over how LiON cells age and the breakdown voltage of the electrolyte, 4.3V. This means I can use my gang of Zeners and a stepped charge cycle, 5A steps down to 0.5-1A, when the 15 cells approach full charge 3.3V/cell and use the 3.6V Zeners safely. This has substantially simplified my in-bike charger.

One other thing, I'm not going to fold the tabs but use them also as heat-sinks. They provide an excellent transport from the body of the cells to where I'll run outside air to help dump the heat.

Of course, we'll still need to understand the temperature effects on voltage.

Bob Wilson

[Daox]

Is there a reason you have decided to split your pack into two parts? I've seen people do this before and the results have been less then stellar vs just running 2 cells in parallel and 15 in series.

[bwilson4web]

Hi,

Originally Posted by Daox

Is there a reason you have decided to split your pack into two parts? I've seen people do this before and the results have been less then stellar vs just running 2 cells in parallel and 15 in series.

I'm building two, independent, 48V (actually 49.3V) strings:

• Primary - used on the out-bound trip.
• Reserve - return to home pack.

These will be symmetrical packs and with both charged, either can be used as the primary. But when its voltage falls too much, the isolation MOSFET will begin a 'duty cycle', reduced power, as the per cell voltage hits 3.0V. The driver will know 'time to switch.'

Right now, I'm thinking stop and move a plug but I may rig up a rider accessible switch, similar to a motorcycle reserve valve, that can be flipped on the fly. The rider then knows to 'head home ... while the reserve pack is still good.'

Bob Wilson

[Daox]

So the main reason for doing things this way is to basically know when half your range is gone?

I just mention it because even though the Peukert effect is minimized with lithium it is still there. The benefits of running the cells in parallel is that you will probably gain a few % more range, have a bit more power (less voltage sag), and the life of the batteries will also be increased as well from only drawing 1/2 the current over their life.

[bwilson4web]

Another, useful, Google indexed link:

• Lithium ion battery with lithium iron phosphates cathode - includes a 48V pack characteristics as well as temperature effects, -25C < 50% capacity.

Bob Wilson

[bwilson4web]

Exactly:

Originally Posted by Daox

So the main reason for doing things this way is to basically know when half your range is gone?

It also lets me build one; lessons learned, and; build two. For example, I'll probably build the pack tabs with:

1. tabs aluminum rivet nearest the cell - shortest distance for current flow
2. insulation separator - somehow connected to one side of the riveted tabs to insulate from the next cell. My first thought is offset rivets so they won't short to the adjacent cell but I'd really like to find a rivet-like, insulating solution.

I'll always have the option to parallel the packs.

Originally Posted by Daox

I just mention it because even though the Peukert effect is minimized with lithium it is still there. The benefits of running the cells in parallel is that you will probably gain a few % more range, have a bit more power (less voltage sag), and the life of the batteries will also be increased as well from only drawing 1/2 the current over their life.

Sometimes we have to make tradeoffs:

• range anxiety - I don't trust the 'battery fuel gauge' techniques because of the flat discharge curve. We don't really know it is empty until the discharge voltage drops below 3.3-3.4V.
• 15A max load - the eBike motor controller is rated at a maximum of 15A and these are already 17-18Ahr cells. I'm already just under 1C load. But it is very likely I won't be at 15A due to the built-in, 25 mph speed limit. So I'm not going to exceed the 1C discharge rate in typical operation.
• mitigate thermal effects - we know the packs will get warmer in operation but I don't know if it will be a range limiting effect. Having two packs means I can switch to the 'cold one' if on a hot Alabama day the pack temperature begins to approach 60C.
• use all of the first pack energy - having a second pack means I can run the first to 'empty' or 2.8V/cell level knowing I have a way home (or to a recharge plug.)

BTW, I'm finding some sources including the video that suggest over 2C discharge may be a bad thing. But as a starter battery, the high discharge load will be brief. Still, I am finding some interesting, alternative LiON motorcycle batteries.

A wild thought for your starter battery pack . . . perhaps a string of ultra-caps might provide initial CCA current in the first "n" msec to reduce the peak discharge load on the cells?

Bob Wilson

[Daox]

Ah, yeah a 15A load is pretty small. You won't see much benefit from paralleling them at that kind of current draw. In fact I'm incredibly confident you won't have any thermal issues at all at that kind of load.

From my earlier testing:

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 agree that using the voltage is not a possiblity for a SOC meter. That is why I am using a ammeter and counting the amphours in/out of the pack for my BMS' SOC meter.

[weazel]

u should drop using any battery there all rubish id go for supper caps there about a 100ish credits each and most of them are only 1~2v but id invest in them u may find replaced stock from a place that uses them look on ebay etc these are wonderful they dont waste energy because of there low internal resistance and the relly great thing about them is they can fast charge relly fast a copple of minets and can be fully charged and depending on ur capacitance they will outlast any baterry of the same size and they dont deteroate like batterys do if ur going to send money id do it on these oh and there small and very light weight