Page 1, Red Hornet-- Page 2, X5 -- Page 3, PSI

Back To Projects





LiFePo4 ( Oh how I should have paid more attention in Chemistry class )

Li = Lithium

Fe = Iron

Po4 = Phosphate

No more SLA, no more Nicad or Nimh..

Most people who have spent considerable time and money in pursuit of Ebike happiness will have discovered that it's all about the battery... And with that comes the next level of performance that requries the next level of battery, and so on..

The " Chaos " was a lesson leanred, after 1000kms on a 72v 20ah pack i quickly realized that I do not need 20ah worth of capacity on my daily driver.. I work 5 kms from home, I take a detour on my way home ( bike lane, and I avoid the traffic circle of death) and I never need more than 5ah when using only the motor, not pedaling at all..

For 2009, I plan to build something lighter weight, but more powerful.

The problem with this is that in order to lighten the load, I need a smaller battery.

It warms my heart to receive a nicely packed box, might be the 12 years at Purolator Courier or the number of times I received various batteries in boxes that scared me enough to consider suiting up in safety gear.. Great job on the packing PSI !!! Thank You !

3.2v - 10 Amp Hour - 362g

Many of the lithium packs I had used up to this point were all rated "2C max". Trying to keep this as simple as possible and not send you into a coma... 1 x C = 1 hour.

So a 10ah battery, drained in 1 hour, would need to put out 10 amps, hence 10 amp hours.

( still with me ? )

Another way to look at it, is knowing your power requriements.. I currently use a 30 amp limitting controller on the Chaos, so I need a battery that can provide 30 amps when I accelerate or climb hills.. This means 3C for 10ah cells.

These PSI cells are said to be capable of 10C ie: 100+ amps !

Ok.. enough with the technical stuff for now.. To make a long story short, these cells are plenty powerfull enough to get the job done !

Triage began, I started by testing every single cell for resting voltage, also weighed them all, they varied between 361g ti 367g , I picked out 16 perfect matches .

Using my LBD and a Cycle Analyst put them to the test... gently at first..

10.1 ah at "1C" and still going strong, but this is the end as the voltage will drop suddenly soon after. LiFePo4 has a very flat discharge curve, this means that it will provide very consistent speed on the bike at the start of the ride, all the way to the end.. unlike every other chemistry, sla, nimh, nicad, LiMn, LiCo, etc..

Then for the fun part, pack assembly !

First step was to make some ballancing wires.. now most people will use teenie weenie 20awg wire, but I went for thicker stuff, 16 awg with thick aluminum crimp ends, one + and - for each cell.. Not taking any chances, soldered each one for premanent " Stay " power..There is a large amount of energy contained in each one of these cells.. let alone 16 of them.. I don't want nothing comming loose !

These wires are used to monitor the voltage of each cell, as well as " Ballance " the pack when requried, by charging any lower voltage cell directly with a single cell charger.

At each end I used 10 awg wire to connect into the bike, this is a bit overkill but I like it that way.

Snady keeping a watchfull eye on me, with her squirrel.

Next step was to fabricate something to hold these cells together ..


I salvaged a piece of thick transparent plastic from work, made a template on masking tape and drilled it out..

Next step was to find a way to secure the cell bodies to each other, preventing pressure on the threaded studs, this is very important as it helps prevent the cells from leaking electrolyte. There is a plastic seal between the brass nut and the cell end, considering what these cells cost ( ie: 40 $ per cell by the time they ship, duties, brokerage, etc.. + chargers + everything you see above to assemble the cells into packs .. ) I don't want any premature failure, so experiments are in order..

I found a product called " Great Stuff ", basically styrofoam in a can, on the left you can see how much it breaks down if you mix it up as it sets, you get less volume but more density.. Then I tried it on a cell to see if it would actually let go in the event I ever need to repair a pack.. turns out it's just right.. sticks to the shrink wrap preventing the cells form rotating but will come apart with some muscle..

Then I made a big mess, yeah, started out good, until I underestimated the amount of expansion, the foam oozed out the ends and got between the cell ends and the copper tabs, I then had to take one side off and remove the excess foam. It did however fill the space between the cells nicely.

Then I tried Hot Glue !

Much better, clean, sets quickly, removable if need be, tough stuff..

I also tried silicone but it takes way too long to dry and was too flexible, allowing the cells to move and stress the studs.

Curious person that I am, I had to open one up and have a look under the shrink. Aluminum housing (not magnetic) with laser welded ends, there is a vent at each end under the black and red stickers, in the event of something bad the pressure created by the suddend release of energy will exit thru these vents instead of exploding ( that's the plan at least.. I hope to never find out )

Something else I also discovered ( woops ) is that the threaded studs are hollow !

6.9 in/lb or .78 nm is the recommended torque value by PSI I foolishly did not use my torque wrench...I will in the future.

(More to come shortly.. check back in later ! )

Page 1, Red Hornet-- Page 2, X5 -- Page 3, PSI

Back To Projects