Small space, big power?

[caza]

I assume that by "cylindrical options" you mean 18650 cells that output 3.7V, so not LiFePO4 chemistry.
In which case, there's something to be aware of regarding capacity: it declines differently with time between both techs. The decline in capacity is much faster with most of the Li-ion techs than the LiFePO4, meaning that while it's true that you 'll have more capacity today with this pack, there's no guarantee that it will still be the case a few months later.
In any case, capacity ratings are very tricky to interpret, some cells with lower marketed capacity will actually have more usable capacity than others under higher loads since they'll heat less, it's not as straightforward as reading whatever is in the spec sheet from the manufacturer. In my humble opinion, it is better to focus on low internal resistance cells.


22A per cell is quite a lot, one of the best cells I know is the Sony VTC5D and it's rated at 25A continuous. It can do 35A peak so you're still in the safe zone, but not a lot of cells can match this kind of specs so if you want the pack to last a while then choose wisely (or overbuild it) :wink:
Do you already have some preference regarding a particular cell ?

I replied to this but it seems the post got lost.

I have read that these cells can degrade quicker, but it's hard to find testing data on what kind of degradation to expect. The good news is I don't expect to ride this every day. Maybe a couple cycles a week excluding when it's particularly cold out. At that rate it would take 5 years to hit 500 cycles. If they can retain even 80% of their initial capacity at that point, I'll be happy enough with them.

22a is a lot, but I'm expecting that to be the absolute max peak amps I pull, not continuous. This isn't a highway cruiser! On my local roads I don't expect to be pulling more than 7000watts very often, I just want to have some margin so I can get up and go if I need to.

As for cells I was mostly looking at the Samsung 40T and Molicel P42a. I'm definitely open to other suggestions though.

I was initially really tempted by the salvaged but "new" model 3 cells BigBattery is selling, but the low amp rating and high operating temp make them not a good fit aside from the price per KWH being darn tempting.

 

[john61ct]

I'm definitely confused as to how you properly run a BMS on cells that are already in series before you connect them.

Why? The best design is a BMS that is easily removed / replaced. It's not like it is "in between" sub-packs in any scenario.

All you need is packs that gives access to the balance wires.

> incessantly monitor them

If you believe in using a BMS, why would you not want to monitor while in use?

Most riders do, even if just with a wattmeter and/or a specialised display, maybe that connects via the motor controller.

If you know your cells are reasonably balanced, all the monitoring needed for normal cycling (charge & discharge) can be done at the pack-level bulk voltage.

The 1S cell/group level testing and balancing are best performed as periodic scheduled routine maintenance.

Remember, BMSs are often the primary murderer of otherwise perfectly healthy packs, add a lot of complexity thus points of failure, and build quality often horrible.

While riding, Ah per mile or km is also a very useful metric, as is an estimate or remaining range.

> just hope they don't catch on fire

Now that's just silly.

With any li-ion chemistry,

which for most excludes the safer, lower voltage and lower energy density ones LFP and LTO

thermal runaway is an ever-present risk, and appropriate precautions are always needed especially if the packs are ever brought within a living space.

Proper care means never getting any surprises, always knowing the packs' state of health, being aware of changes as they wear, knowing when EoL is approaching, in order to proactively recycle and replace long before gross symptoms start to become noticeable to lay owners.

The **per cell** (not just group level) access LiPo sub-packs offer in fact makes such close monitoring and testing **much much** easier and reliable.

Once you have cells welded into parallel groups, you lose all visibility of per-cell condition without ripping the pack apart.

Compared to just disconnecting bullet connectors and swapping out a suspect pack, and connecting it to a standard cycling capacity / load tester to compare to your as-new benchmarks, measuring changes / variations in cell resistances etc

 

[caza]

Here's the bike with the motor in place.

The little mount the motor came with bolted perfectly in place to one of the stock mounting points and it's nearly perfectly aligned with where the old sprocket was. I'll be adding a much more robust bracket but for now it's nice to know the motor aligns and fits well.

I was hoping to put the controller in the space under the seat, but it's not going to work out, so it'll go where the tank was. DC/DC can go in that space. Maybe a small charger? Haven't decided if I want a separate 12v battery but I'm inclined to skip it.

 

[caza]

These are the top options I'm considering right now.

Molicel P42a's. Two packs side by side, 20S16P. 4.7KWH nominal. $1500~ for the cells alone.

These seem like the best choice for "vape pen cells" for my application.

At 16P, even at 200A peak load each cell would only see 12.5A, well within what they're capable of. At 100A continuous only 6A.

Bestgo 17AH pouch cells. 20S3P. 3.7KWH nominal. $1500~ for cells alone.

3P, 200A peaks would pull 67A per cell. Their rated peak is 70A so not much room for error there. Same for 100A continuous, which would be 33.3A per cell, they're only rated for 35A continuous.

These are the pouches with the best use of space I could find. But I couldn't find much info of many people using them. Are they a reputable brand?

Wish I could find a source for EIG C2020 cells, I think they would work quite well, but I can't find anyone who's selling them in the US.

I also played around with several different prismatic cells but couldn't find a configuration that would work well for the space I have.

Unless someone has a great alternative idea I think I'll end up going with the molicels. There are certainly drawbacks, but I haven't been able to find something else that really compares for price, performance, shape, density, etc.