File this under “typing myself smart”. Or “smaaaht” as we say in Boston. Anyway, I’m trying to work out the maths on the lipo cells I looked at here. Now, after reading a bit more closely, it seems like to get a good model I should go with the Turnigy single cells at 40C. This is just for the sake of crunching some numbers, and the Turnigy are a pretty good, standard cell.
So, here are the specs for those cells.
- Cell voltage: 3.7 (nominal)
- Cell capacity: 5Ah
- C rating: 20C
(note: Chris mentioned a 40C rating on the ES thread, which he said was 20C real-world, so I’m going with that.)
Total current per cell: 100A
Pack configuration:
For 200V we need to run 54 cells in series. Let’s say we want around 40Ah and see what that gives us. 8 cells paralleled? So 8 x 100A? 800A? Can this be right? If it is, it’s crayzeeeeee, so let’s dial it down to 400A, or 4 cells in parallel, because 400A at 200V = 80kW. Or 5 cells. Sorry, coffee needs to kick in.
So we have a 100kW pack. It’s made up as 54s5p, or 270 cells. Each cell weighs 130g, so 270 of them is about 35kg, or 77lbs.
Now, here’s the thing. That’s 100kW peak for the pack, but 100kW continuous for the motor (I have to check that to confirm.) This pack is crazy light, so let’s add some to get some more juice. If we go to 6 cells, we have 600A at 200V, so about 120kW. That’s enough buffer, I think, and now the pack has 324 cells and weighs 93lbs.
Also, we can re-configure this for higher voltage, which, I believe will give me better performance out of the Rinehart. Let’s go with 108S and drop it down to 3p. Now we have 400V at 300A.
This all adds up to a 15Ah pack running at 400V, which gives us a 6kWh pack and weighs 100lbs.
Funny. Wasn’t that just about the size Chris was talking about?
The price? $2000.
Guess what. You know that pack configurator calculator thing I thought should exist? Well I made one. My skills are weak in this regard, but if you want to play with it, here it is: pack-config
The Electric Chronicles: Power in Flux 
Thanks for the info Ted. Don’t forget that when fully charged, those cells could be as high as 4.2V. For the PM100DX, is it 360V maximum operating voltage? So that would mean about 84S would be about the maximum number of cells you should go?
I think the inverter has an operational voltage of 360 V, but will tolerate being exposed to a 400 V bus. I’ll have to check with Larry or Chris as to whether it needs a firmware upgrade or if it can tolerate starting off on this voltage. So you might be able to get away with an 88 s pack. 88 s is a great number – you can divide by four to get four sub-packs of 22 s, and this is divisible by two to give an odd number (11 s). The odd number means the pack terminations will end up on opposite sides, making the adjacent connection much easier.
I recently did some tests on some cells very similar to these ones, comparing the power density to some high power 18650 cells. I was specifically comparing power density: although the LiPos kill in terms of C rating, the higher energy density of the 18650s is a factor when considering power density. My test was a little complicated to get the information that I wanted, so bear with me:
My LiPo cell was a Dow Kokam UHP cell with specs very similar to your Turnigy (though much more expensive!) It was 4.5Ah, weighing 125 grams with a C rating of 30 continuous, 50 peak.
My 18650 was a Samsung INR18650-25R (NCA chemistry) that was 2.5Ah and 45 grams.
To compare these cells, I wanted my test to be outputting the same number of Amps per weight. My test bench only went up to 50 amps, so I tested the Kokam at that level. Certainly not stressing them too hard, but they did sag and heat up.
To determine the power for the 18650 test, I had to take into account the different capacity AND the different weight, so I wanted the same Amps per gram, so 50A * 45g / 125g = 18Amps. When I tested the 18650 at 18Amps, I was surprised to see that the voltage drop and the temperature rise was (very slightly) lower than the Kokam at 50A. So the power delivery is about matched, if not slightly favoring the Samsung cell. Now, what I’m not sure of is whether the extremely high C rating of the LiPo implies that it can handle a lot more heat, but I tend to be wary of published C ratings and prefer to look at voltage drop and temperature rise. At any rate, the matched power density, as well as the 1.5 times higher in energy density (and the much lower cost) led me to choose the 18650s for my project.
Just one more data point 🙂
The Rinehart should be good to ~410-415V before it wont turn on if I recall correctly. I believe most people run 96s packs (4.2*96=403V) as it is divisible by 12 which is really good for BMS integration
410V is the hard limit, from what I’ve been told.