Small space, big power?

[caza]

Hey all. Trying to plan out my battery for a kymco spade conversion. I have a QS 138 90 and EM200 on the way. I also put a pre-order in for a Nuclear 24, I really like the screen it has for tweaking settings.

I keep going back and forth on what type of batteries to use and I would love some input on my options. The space I have for the battery is pretty limited. Essentially it's a 6" x 12" space, 8" wide, with another 6" x 6" x8" space underneath it.

Ideally I'm aiming for about 3KWH if I can make it, but critically I don't want the battery to be the bottleneck of the system. I want to be able to pull those 200 amps peak if needed on the EM200.

Cylindrical cells seem like the typical go-to. It looks like I could fit 160-200 21700 cells. Samsung 40T or Molicel P42A have high current ratings that should be enough for my needs as long as the bus bars and nickel strips are sized and build appropriately.
I'm of course also concerned about safety. Has anyone done cell-level fusing on these cells? It doesn't seem to be very common with high-amp builds.

Then there's pouch cells. In theory they are ideal, but when I find a cell that's dimensions fit my needs it's often hard to find a good supplier. The EIG C020 are perfect, and have great kits for assembly, but I can't find any supplier in the states.

I've looked at other pouch cells but I'm also having trouble finding reliable information for what the best way to assemble packs from these are. I've seen people spot welding them, using PCB bus bars with through-bolts, crimping them together with rivets, etc. But there isn't a lot of info on which method is best and why, it often seems like people are just winging it. I like the idea of using pouch cells a lot, there are good reasons why Zero, Energica and others use them but it seems like the resources for DIY packs are scarce.

I think if I had a bigger bike prismatic cells would be the obvious and easiest choice, but their sizes seem to be incompatible with my size and voltage needs.

Any suggestions or insights would be greatly appreciated.

 

[john61ct]

LiPo as used in RC planes has highest power density

And energy density too but focus is on highest possible C-rates.

See link at today's post by me

https://endless-sphere.com/forums/viewtopic.php?t=104676

Note that longevity will likely be shorter, depends on your care factors, especially the avg DoD% and actual C-rates you pull for more than a minute at a time.

And LiPo has much higher fire risk, you need to know what you're doing, put in good protective infrastructure, not charge near a dwelling unattended etc

 

[markz]

Also, be careful in the exagerrated claims of C-rate for Lipo. Not all exaggerate but Hobbyking for example, a rule of thumb is half or 1/3 of whatever they claim in terms of c-rate discharge. Its all over ES, past user icecube57 did a lot of testing and the dud rate for new hobbyking lipo was like 15% or something.

Also note that any battery chemistry has the equal ability to be dangerous, more so if you have not done any research and don't know what your doing like using Lipo without a LVC cutoff of some sort.

 

[john61ct]

Also, be careful in the exagerrated claims of C-rate for Lipo. Not all exaggerate

Yes they all do, would immediately go out of business if they didn't.

The thread I linked to goes into the technicalities of objectively measuring the actual C-rate a pack can sustain at high amps, taking into account the impacts of ESIR, temperature rise, voltage sag etc

Really industry changing stuff, professional grade science going on there!

Very powerful aircraft, flown for several minutes at a time only, is their use case.

SMC is the top performer but pricey. New entrant Liperior is the value leader, but runs out of stock very quickly, and I think only ships to USA for now.

 

[caza]

I did some very basic research on hobby packs and quickly decided I don't think they are for me.

The seemingly high dud rate, fire rate, totally untrustworthy specs, And generally lack of good info on building bigger packs safely with them all make them seem far too risky to be worth it.

I do understand that any battery can be dangerous if used improperly, but it seems like starting with unreliable hobby packs to build a much bigger pack could easily be considered improper use.

So far I am most compelled to go with cylindrical cells since high quality cells are readily available and there is a huge wealth of easily accessible info on building with them.

 

[Chalo]

With vape pen cells, your options for getting lots of juice out are limited. The cells can kick ass, but the nickel strips do not. Plan on 5A per parallel string. 10A for short bursts.

Prismatic cells and big pouch cells stomp the crap out of penlight batteries in that regard.

 

[markz]

A123 26650 cans give a lot of punch, LifePO4 chemistry but again like Chalo said the strips for connecting.

Then theres old Headway cans, again LifePO4 chemistry with bolt on bus bars.

Then theres the cons with LifePO4, bulky, heavy.

https://endless-sphere.com/forums/viewtopic.php?f=3&t=85539&p=1251239&hilit=heavy+bulky+lifepo4#p1251239

 

[john61ct]

Yes LFP is double the weight and space of the usual 3.6-3.7V chemistries

maybe triple that of hobby LiPo.

Lasts a lot longer though, but here OP has made it very clear that power density is the #1 consideration.

If you use say 6S or 4S LiPo packs as your "building blocks" you're just crimping terminating connectors together with heavy gauge wires to get the xPyS layout you need

lots easier than learning to spot weld

and **much** easier to monitor cell health, diagnose any problems, replace any duds down the road.

None of it rocket science, plenty of forums hundreds of helpful members to get you up to speed.

 

[john61ct]

The seemingly high dud rate, fire rate, totally untrustworthy specs, And generally lack of good info on building bigger packs safely with them all make them seem far too risky to be worth it.

All issues easy to overcome, truly.

 

[Dui ni shuo de dui]

I think you should first come back to the origin of your problem and see if you can manage to get more space out of your frame before considering crazy options.
Though it seems to me that you should have enough space to fit a really decent battery in there, I'm pretty sure you could fit at the very least a 20Ah 24S pack of A123 cells, just like the ones I'm using on my builds. That would be the size of just one of the two blocks I'm using on my roadster (see the first link in my signature if you don't know what I'm talking about).
Such pack at 20Ah could deliver 300A peak without breaking a sweat, so that would be plenty enough for your build.
Dimension of such pack is around 165x250x230mm or so, for around 12kilos.

Also, don't hesitate to share pictures of the bike if possible with dimensions, it might help suggesting some ideas.

 

[caza]

Thanks for all the replies.

I think you should first come back to the origin of your problem and see if you can manage to get more space out of your frame before considering crazy options.
Though it seems to me that you should have enough space to fit a really decent battery in there, I'm pretty sure you could fit at the very least a 20Ah 24S pack of A123 cells, just like the ones I'm using on my builds. That would be the size of just one of the two blocks I'm using on my roadster (see the first link in my signature if you don't know what I'm talking about).
Such pack at 20Ah could deliver 300A peak without breaking a sweat, so that would be plenty enough for your build.
Dimension of such pack is around 165x250x230mm or so, for around 12kilos.

Also, don't hesitate to share pictures of the bike if possible with dimensions, it might help suggesting some ideas.

I think your right that I could probably fit those A123 cells, though I think I'd have to have a split pack. Certainly a good option.
I'll definitely share pictures once I get the build going. I want to get the ICE out of there and the QS motor in place, after that I'll have a much more precise idea of how much space I have to work with and I'll be designing the battery and its enclosure around that space. Space is definitely going to be very limited, but I knew that when I chose such a tiny motorcycle to convert and then opted not to get a hub motor. My 3KWH target might be too ambitious, and I'm okay with something lower if that's what's necessary.

Yes LFP is double the weight and space of the usual 3.6-3.7V chemistries

maybe triple that of hobby LiPo.

Lasts a lot longer though, but here OP has made it very clear that power density is the #1 consideration.

If you use say 6S or 4S LiPo packs as your "building blocks" you're just crimping terminating connectors together with heavy gauge wires to get the xPyS layout you need

lots easier than learning to spot weld

and **much** easier to monitor cell health, diagnose any problems, replace any duds down the road.

None of it rocket science, plenty of forums hundreds of helpful members to get you up to speed.

Any suggested threads to start with for researching them further? I've see a few threads of builds that used them but they were not very detailed.

With vape pen cells, your options for getting lots of juice out are limited. The cells can kick ass, but the nickel strips do not. Plan on 5A per parallel string. 10A for short bursts.

Prismatic cells and big pouch cells stomp the crap out of penlight batteries in that regard.

Vape pen cells is a funny name for them considering for a while manufacturers were actively trying to stop vape users from obtaining them.
Definitely was considering more than nickel strips if I end up going the cylindrical route. I was thinking of cutting out copper busses on my CNC and just using nickel from the copper to the cells.

But yes I'm very interested in prismatic and big pouches, my trouble was finding reliable info on building packs from pouches, or finding pitmatics that will fit in my small bike frame.

 

[Dui ni shuo de dui]

Space is definitely going to be very limited, but I knew that when I chose such a tiny motorcycle to convert and then opted not to get a hub motor. My 3KWH target might be too ambitious, and I'm okay with something lower if that's what's necessary.

I haven't paid attention to that point: why is it that you won't use a hub motor? Doesn't make any sense to me, those little bikes are the perfect candidates for it.
Hub motors can be insanely powerful, on such small motorbikes you could do both power wheelies and have a crazy top speed, provided you have a powerful enough battery to power it... for which having more space in the frame would make things much easier (and less exensive)

You should seriously give it another thought, an inboard motor makes very little sense here, many drawbacks for no real, actual benefit I could think of.

 

[Overclocker]

taiwanese bike deserves taiwanese cells :lol: molicel

 

[caza]

I haven't paid attention to that point: why is it that you won't use a hub motor? Doesn't make any sense to me, those little bikes are the perfect candidates for it.

I'm not totally against it, but I wanted everything to bolt on without modifying too much of the base bike, and none of the QS hubs I looked at were the right size for the existing swing arm. Between buying a second swingarm to modify, buying and setting up a rear disc/caliper, it seemed like more trouble and more money than it was worth to save a little space. But of course this has helped create my current problem! The mid drive is going to be easier for tweaking gear ratios, swapping out for another motor down the line, or swapping the motor to a new project later. Plus you know, all the usual arguments about unsprung weight aren't without merit.

What are the "many" drawbacks for mid-drive. The only two I can think of are space and having to maintain the chain.

taiwanese bike deserves taiwanese cells molicel

Your comment comes off as sarcastic, but from what I can see the molicel P42a's seem to perform well in independant tests. What am I missing?

 

[Overclocker]

sarcastic? why would you think that what's wrong w/ you

taiwan is awesome. they got gogoros etc. if i had a kymco to convert i'd definitely use molicels. my next build's gonna use moli p26a

 

[Dui ni shuo de dui]

What are the "many" drawbacks for mid-drive. The only two I can think of are space and having to maintain the chain.

Main drawback is of course space, which leads to several subsequent drawbacks:
Less battery space means either less capacity and/or less voltage
Less voltage usually means less top speed (very simplified)
Less capacity means that you'll have less range, but also a lower acceleration since the amps delivery depend on paralleled cells.(simplified also cuz I don't want to write too much, but that's the idea).
So that leaves you with less batttery performance, which means that the battery will likely be the bottleneck on your setup. That's always better to avoid being limited by the battery, since the consequence in most cases is tat the battery will have a shorter life.

So to summarize, you end up having less battery performance and less battery durability. Generally speaking if you want to keep good performances that leaves you with the only option of going for a less stable chemistry with a higher cell voltage and higher discharge rate, LiPo, which usually don't last very long and are less user friendly/not as safe.

The other main drawback is that it's usually harder to fit an inboard motor than a hub. Swingarms are not very hard to convert to hub motors and you can most of the time even find some compatible ones designed for this purpose. An inboard motor means custom brackets, welding and a perfect alignment. Nothing super difficult in both cases but usually it's easier to fit a hub.

Also, no chain, no annoying noise. People like to say that they have gear ratio choices but in practice very few actually go through the trouble of changing the gear ratios, since custom sprockets aren't cheap.
I'm not saying that inboard motors don't have some nice qualities, I just mean that for such a small bike it would have been wiser to fit a hub, at least in my opinion.
Anyway, I re-red your first post and saw that you already bought the motor, so I guess it's done, no need for me to insist and sorry if that bothered you, I thought the motor choice was still pending :wink: . Let's wait for the pictures, hopefully with a bit of luck we'll be able to find a clever way to pack a beefy battery in there!

 

[markz]

Noise but some have reduced it, sinewave and used their own grease, maybe even diff gear material. Stock C4k coaxial with trapezoidal was noisy. Was a mountain goat as I could stop on a 20% grade and the thing would climb Mt. Everest, good luck doing that on a hub motor. People could hear me coming up behind them on the Cyclone 4kw, but on a Leaf 1500W or any other direct drive on trapezoidal controller they couldnt hear anything, but the sinewave controller always helps out with noise.

 

[caza]

Personally I enjoy the noise of a chain on an emoto, but I understand not all share my enthusiasm.

As you said, maybe a hub would have been a more efficient choice, but the motor is on the way, for better or for worse.

It seems more and more likely that I go down the cylindrical route. I would cnc copper bus plates, nickel plate them, and solder nickel strips to the plates, and spot weld those strips to the individual cells. I've done some rough layouts for how cells could fit, and it would be trivial to fit as little as 180 cells for a 20S9P pack, and the max I could reasonably fit is 300 cells at 20S15P. If I'm aiming for a 200A peak the smaller pack would pulling 22A per cell and the larger pack would be pulling 13A per cell. Both are well under continuous ratings for cells that are available so I'm pretty comfortable with that.

I've done a bit more research on pouch cells and am no longer intimidated by building with them, but sourcing cells with dimensions that suit my project still seems like a gamble. If I had a good source for EIG cells in the US that is the route I would likely take, but I cannot find any. I should be able to fit A123 20AH's, but the pack would be incredibly wide for the bike and still have less capacity than the cylindrical options.

Sorry for the misread on the sarcasm! It would be fitting to use cells from the same place the bike was made.

 

[Dui ni shuo de dui]

I should be able to fit A123 20AH's, but the pack would be incredibly wide for the bike and still have less capacity than the cylindrical options.

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.

If I'm aiming for a 200A peak the smaller pack would pulling 22A per cell and the larger pack would be pulling 13A per cell

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 ?

 

[Dui ni shuo de dui]

Was a mountain goat as I could stop on a 20% grade and the thing would climb Mt. Everest, good luck doing that on a hub motor.

We are talking about a kymco spade here, it's a street bike so that's a very different use case, requiring different specs :wink:
Good luck doing both 20% climbs and 130+ km/h with the same gear ratio and a small battery, it's always a matter of compromizes
If the OP wanted to build an electric dirt bike then I'd totally agree with this point of view though.

 

[caza]

To be fair streets still sometimes have hills and the ability to deal with them is just as relevant on a street bike. I have some pretty decent hills to get to the other side of town where my mother lives, it would be nice to not be bogged down on them.

We also have a small local mountain with a paved road to the top, it would be fun to ride up on the spade, though I doubt I would be able to reach the summit with the space I have for batteries, even if I had used a hub motor.

 

[markz]

Some prefer silence, others dont no matter what they are riding whether it be going slow in small tight townships or on the highway. Its the same as regular gas motorcycle riders who dont care about noise and disrupt the peace of everyone around them by modifying their motocycles to others who keep them stock and the noise is bareable.

Harley Davidson purposely made their electric motorcycle produce more sound intentionally, whereas the Zero Motorcycle in one article the first line in the first paragraph mentions [Zero motorcycles] "prove quiet, efficient, and fun"
Another article seems to be a promo article based on quiet and silence - https://electrek.co/2019/02/08/zero-electric-motorcycle-tesla-police/

Lots of people are putting a C4k mid drive motor on motorbikes, just read a post here on ES where a guy put one in an old diesel quad. Myself I'd drop more money and get a 16kw Cyclone motor but all depends on what your using it for. I got an old Suzuki motorbike up at the cabin thats collecting dust for a decade now. Its all flat land up there, lake cabin community with families so no need for noise, no need for speed. 4kw would do just fine. If your in a San Fransisco with its steep hills, I dont know how a QS/MXUS 50H hub motor would handle a stop on a hill without enormous amounts of amps and lots of voltage to get the overall watts high enough. You may not stop on a hill all the time, but there might come that time.

QS 138 90 and EM200 on the way. I also put a pre-order in for a Nuclear 24

http://www.qs-motor.com/product/qs-motor-138-4000w-90h-pmsm-mid-drive-motor/
Looks to be a good motor, specs say limit 200A and Nucular 24Fet should be plenty of power.

 

[Dui ni shuo de dui]

Not sure hy there's so much focus on the noise "problem", it's not like it's really important. IMHO chain noise isn't really a nice noise to hear (and I'm a big gas engine noise enthusiast, I even have playlists of engines sounds I like to hear sometimes, so no one can call me a noise hater :wink: ) but I guess it's subjective. Anyway, by focusing on this you are missing a bit the important point in my post which was the battery.

Regarding hills and climbing, we do not have a lot of these in Shanghai that's for sure, but climbing from standstill from an underground parking steep ramp is no trouble at all, even with my very heavy motorbike (and very heavy 95kilo ass :lol: ), so I don't think this is actually a problem unless maybe if the hill is kilometers long. I think you seriously underestimate how much torque those hubs deliver!

However, I do agree that I wouldn't run a hub motor in San Francisco, I don't think it would last very long indeed!

 

[john61ct]

Any suggested threads to start with for researching them further? I've see a few threads of builds that used them but they were not very detailed.

Because you're not doing much more than when "building" with Lego. Choose the connector type, and wire gauge, figure out physically arrange them within your 3D space available, deciding on your enclosure.

Maybe post a thread dedicated to the idea, both here and the rcgroups forum.

If using a BMS, then that's its own thing too

 

[caza]

Any suggested threads to start with for researching them further? I've see a few threads of builds that used them but they were not very detailed.

Because you're not doing much more than when "building" with Lego. Choose the connector type, and wire gauge, figure out physically arrange them within your 3D space available, deciding on your enclosure.

Maybe post a thread dedicated to the idea, both here and the rcgroups forum.

If using a BMS, then that's its own thing too

So the short version is just crimp on your connectors, incessantly monitor them, and just hope they don't catch on fire? It just seems like a more high risk way of building a bigger pack.

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