OK so all the LifePo4 stuff is very interesting, but I recently fell over some NMC cells (Nickel Manganese Cobalt Oxide) for silly money. Given that I thought I would take a flier and see what we could do. Now NMC is a slightly different chemistry to LifePo4 and has slightly different parameters. I am building a test battery with Kokan 11A/hr pouch cells which offer considerably high energy density per Kg than LifePo4. They also feature higher cell voltages of around 3.7v/cell. This leads to a number of significant issues
- 4x (4×3.7v) cells have nominal voltage of 14.8v – the starter will spin faster, and the electronics may work better at lower current.
- The charge cycle at full power requires 4.2v per cell (16.8v) which ios above any standard alternator output. This means a custom alternator / charge controller
- 16.8v on the DC line ‘may’ be too much for some systems – let’s hope not
Anyway I have ordered a total of 16 pouch cells with which to build my test battery, this should give me a 44 A/hr battery able to support 90+% depth of discharge. Equivalent to around 80 A/hr of lead acid. More than enough for testing purposes.
Given that NMC is the new and upcoming ‘kid on the block’ perhaps the target would be to move to NMC. True the voltages are not such a good a match at 12v, (the match is better on 24v systems), but they do offer more capacity at less weight. More to the point, for me at least, at the moment they offer a route into Lithium power at a price I can afford.
Currently my 44A/hr test battery will set me back the princely sum of £135 plus whatever electronics and housing I care to add. less than the cosy of a decent deep cycle lead acid 🙂
OK, so after receiving these cells and doing a bit of digging / investigation, I discover they are not NMC cells at all, but Lithium Titanate cells. These have a lower operating voltage (around 2.4v nominal) but a much MUCH longer service life – something in the order of 100000 charge / discharge cycles at fractional C rates, and 50000+ cycles at multiple C rates with over 90% retained capacity. These things can handle 15C discharge and 6c continuous, toghether with 4c charge rates. They don’t suffer from thermal runaway, and are really cool pieces of kit. The down side is the need for 5 serial cells for operation in the 12v region.
So, having splashed some cash for 16 cells, i discover I can only use 15 in a 3P5S setup to give me 36A/hrs at a nominal 12v. Testing has been ‘interesting’ to say the least – these things are awesome! a 15 cell set (3P5S) has shown excellent linearity and a near constant output voltage across sustained discharge. in all an excellent idea all round it would seem.
More on LTO cells and what I intend to do in a separate post.