Looking for help sourcing a motor for my ebike setup

[StillDre]

I'm looking to make my first ebike. I've gotten my hands on a 2019 Commencal Meta, I'd like to make it a commuter in a really hilly area, I also want to take it on some trails so ground clearance maters to me. I want it to be a mid drive because being able to shift gears lets any motor I use in it run efficiently and it's also fun. I could maybe get a kit but I am not happy with the ground clearance with Bafang kits and the other kits I've been seeing don't seem to be as powerful as I want, the really nice CYC kits are out of my budget.

I am also looking for a challenge, I want to diy almost all of the powertrain if I can. My current idea will have the motor running legnth wise along the bike and use a 10:1 wormdrive gearbox to attach it to a shaft that then has a one way freewheel sprocket on it to deliver power to the chain. I have access to a capable 3d printer and a capable CNC mill so I'll use that to adapt a front freewheel to my current chainring. I want to be able to use the motor and pedal independantly of eachother. One of my goals is to add the ebike setup to my current chainline. The green indicates the chainline I would make Is to use a left side drive outrunner that I'll do a two stage reduction on. The first reduction will be inside the bike frame and the second will be between the wheel and the first reduction. I would like to do a reduction with belts but the second stage might need to be a chain instead.

I was looking at using a QS90 80H and a Votol EM-50-4 which I can get for a little less than $300 but if I used that motor to it's full potential, it could probably tear the spokes out of my back wheel. The QS motor there could put out as much as 350nm at the sprocket if I geared it down to 100rpm. I need recommendations for physically smaller motors that can do over 100nm and 100ish RPM once geared down. That means that a motor that could put out 14000rpm and 1nm of torque would be a match. Outrunners and inrunners are fine, as long as it can sustain the torque. I'm also looking for cheap battery recommendations reasonably priced batteries that balance high current and capacity. I was planning on stacking SPIM08HP cells on the down tube and top tube for a 20s 1p setup but if I used 1p I would end up with a smaller capacity than I want and if I went with a 2p setup (the way they come stock from batteryhookup), I'd end up with a larger setup than I want, I could maybe do higher voltages with 1s but I'm worried about the voltages I can use on other motor/controller setups. I am also worried becasue those pouch cells are quite wide and may be hard to ride with. I am open to building the battery myself, it's a skill I've been wanting to build. The problem is that I can't find cheap cells that can do the current that I want without using a lot in parallel.
My budget is about 1000usd for the conversion but I'd rather spend less if I can.

Thank you anyone who had ideas or recommendations.

 

[offGridDownUnder]

I can't find cheap cells

The cheap cells are the ones that will burst into flames.

That's not a joke.

 

[StillDre]

Well I get the issue with some really cheap ones, I also understand that lithium batteries are very dangerous if mistreated. I'm looking for something that I can build for less than $200 which feels fair given that the cells from the SPIM08HP battery pack would only cost $120 and are quite safe (as far as I know). I'm just looking for something different with less width, or recommendations to make the width less of a problem.

 

[ferret]

Worm gears have bad efficiency.

 

[StillDre]

Worm gears have bad efficiency.

My plan is to use a >9:1 gear ratio so that I can extract as much torque as I can from this system and run the whole thing on very low amperage. I believe that a nylon double worm with a stainless steel worm gear will be efficient enough not to melt and create the reduction I am looking for. I evaluated several ways of doing this gearbox and my current understanding is that a worm gearbox is somewhat simple, so it is well within my ability to see through and get the project done. If I were to find a motor small enough to use a two stage reduction or large singe stage reduction, then the design criteria would change and I would forgo the wormdrive.

 

[amberwolf]

Worm gears do work, but they're not very efficient. They use them in most of the powerchair motors, such as the ones I used on CrazyBike2 for a pedal-cadence-matching drive: 2WD Semi-Recumbent Recycled-Parts Cargo eBike: "CrazyBike2"

but ya gotta watch the chain alignment because the torque is....

Most of these motors are brushed, so not efficient themselves, but they're very simple to operate. I've been using brushless for everything since then, but am considering a brushed motor (perhaps even one of these) on a new cargo trike build Help design a DIY middrive for this heavily customized trike

 

[amberwolf]

I am also worried becasue those pouch cells are quite wide and may be hard to ride with. I am open to building the battery myself, it's a skill I've been wanting to build. The problem is that I can't find cheap cells that can do the current that I want without using a lot in parallel.
My budget is about 1200usd for the conversion but I'd rather spend less if I can.

With that high a budget, you can buy some good ex-EV cells in modules from batteryhookup, greentecauto, etc., and not have to worry about "cheap cells" that are prone to burning things down.

The modules will be built with matched cells that will probably *still* be matched in all characteristics when you get them even though they're used. (My automotive EIG cells are still doing very well after over years of hard use by me and they were all used when I got them).

The cheap cells (especially the cylindricals) won't be matched, so they'll all charge and discharge differently than each other and require constant monitoring and balancing to ensure none go too high or too low (to help minimize the chances of cell damage that can lead to a fire).

You can rebuild pretty much any of the ex-EV modules as needed; the ones made from large-format pouch or block (prismatic) cells are easiest to do this with, and the cells themselves are generally hardier and better suited for EV use than the cylindrical types....but any packs made from non-cylindrical cells need to be built to compress these cells across their large flat surfaces. There are various posts around the forum about pack compression methods with pictures, etc.

 

[StillDre]

With that high a budget, you can buy some good ex-EV cells in modules from batteryhookup, greentecauto, etc., and not have to worry about "cheap cells" that are prone to burning things down.

The modules will be built with matched cells that will probably *still* be matched in all characteristics when you get them even though they're used. (My automotive EIG cells are still doing very well after over years of hard use by me and they were all used when I got them).

The cheap cells (especially the cylindricals) won't be matched, so they'll all charge and discharge differently than each other and require constant monitoring and balancing to ensure none go too high or too low (to help minimize the chances of cell damage that can lead to a fire).

You can rebuild pretty much any of the ex-EV modules as needed; the ones made from large-format pouch or block (prismatic) cells are easiest to do this with, and the cells themselves are generally hardier and better suited for EV use than the cylindrical types....but any packs made from non-cylindrical cells need to be built to compress these cells across their large flat surfaces. There are various posts around the forum about pack compression methods with pictures, etc.

That GreenTechAuto marketplace is exactly what I've been looking for. They have a much better selection than the other ones I've been looking at.
Thanks for the suggestion.

Something else is that I guess I don't mean "cheap cells". I'm not looking for anything really nice but cells that won't cost me a grand for 1 battery pack.

 

[amberwolf]

That GreenTechAuto marketplace is exactly what I've been looking for. They have a much better selection than the other ones I've been looking at.

Keep in mind that the battery place greentecauto doesn't have an H in the name. If there is one that does, it's not the one I'm referring to.

Something else is that I guess I don't mean "cheap cells". I'm not looking for anything really nice but cells that won't cost me a grand for 1 battery pack.

That I can understand. I don't have any budget for anything, so I use whatever I can get...except on batteries I have a minimum requirement of quality / source.

It's just a point we like to make clear, since there have been so many people screwed over because of the cheapness of the cells or batteries they bought, some of them with actual fires destroying things (includng their homes), some of them just having to spend more money to buy new stuff to replace the junk after it fails or can't do the job they need done.

 

[scianiac]

You should do some research on the efficiencies of worm gearboxes before you just dismiss our comments, yes nylon versions can be more efficient but the efficiency is also RPM, torque, and ratio dependent. Like it's highly likely you'll end up in the 60-80% efficient range which will totally negate any possible gain in efficiency from your derailleur.

A derailleur system that you are going to try and put how much torque into? Based on the size of the motor sprocket and your 100nm figure it's going to be pretty worthless, always at risk of chain snapping and totally pointless in the smaller gears where it will just slip. I do think high power rear drive issues are often misunderstood though. It's not at all that bike chain isn't strong enough it's in the details. I've snapped (or more specifically popped the plates off the pins) numerous times on my 800w TSDZ2 with a derailleur (that slips in it's smallest cog even with aggressive chain wrap) and that's after I modified the BB to get chainline as optimal as possible. Meanwhile I put 5Kw and around 250 rear wheel Nm into the cheapest KMC chain on amazon and never had an issue but the difference is that has no derailleur, a single speed with good chainline and a massive rear cog. The effect is higher chainspeed, lots and lots of well engaged teeth on the rear so the load is lower and not trying to shift it around causes angled forces.

 

[StillDre]

before you just dismiss our comments

Just to be clear, your comments have merit. I'd prefer to not use a worm drive gearbox but that's the idea that I think would be within my ability to manufacture. I don't think I would be able to manufacture bevel gearbox that'll hold up. That's why I'm looking for alternative motors and powertrain ideas.

a single speed with good chainline and a massive rear cog

So if I want to keep the ability to pedal effectively, I'd have to give up my rear brake to put a sprocket on the left side? What are the alternatives there? The single speed bikes that I've seen with a left side freewheel use the disk brake mounts.

massive rear cog

That's a bummer, I was hoping to avoid this with some engineering. I find big rear sprockets really unattractive. I think I'll keep looking into alternatives but running two chains might be the only real one. I'd rather decrease complexity as much as I can. Since I have an 12 speed rear cassette, I might be able to get away with not using my small gears much and instead using a much faster chain speed. The fastest I'll need to go on the regular is 35mph. I'm also considering switching to belt drive but I'm worried the cost to replace the hub and relace the wheel will be more than I want to spend.

 

[StillDre]

I've found the Turnigy-Aerodrive-SK3-6364 and the Flipsky equivalent. They both put out about a good amount of torque and rpm. I was thinking I could buy two and use a gearbox to join two of their output shafts. I could then use a single stage gearbox and have plenty of turque or flexability in where I want it in the bike. I could connect both motors to a single controller. I just need to figure out how to deliver that power to the back wheel. The whole system would have to run off 42v. Can I overvolt motors like that to run at 56v or 72v?

 

[Chalo]

Physically small motor + low airspeed + high load + small fraction of free RPM = a great chance to burn your motor.

When nobody serious does it that way, stop to consider why not. Gearing down makes good sense for your application. Using a motor designed to run fast at low torque with lots and lots of airflow does not.

 

[ferret]

Just to be clear, your comments have merit. I'd prefer to not use a worm drive gearbox but that's the idea that I think would be within my ability to manufacture. I don't think I would be able to manufacture bevel gearbox that'll hold up. That's why I'm looking for alternative motors and powertrain ideas.

Why is worm reduction within your ability to manufacture and bevel reduction is not?
In most cases bevel reduction is easier.

So if I want to keep the ability to pedal effectively, I'd have to give up my rear brake to put a sprocket on the left side? What are the alternatives there? The single speed bikes that I've seen with a left side freewheel use the disk brake mounts.

It is possible to run a sprocket and a disk brake on the left, but it is not easy .

Astral Rune build thread

Hi everyone. This thread will document my new project. I will post photos and description of the process, feel free to ask if you want more details. This build is based on a Banshee Rune V2 frame. This is the only frame I could find that ticks the following boxes: * Modern high quality mountain...

endless-sphere.com

Avner.

 

[scianiac]

Just to be clear, your comments have merit. I'd prefer to not use a worm drive gearbox but that's the idea that I think would be within my ability to manufacture. I don't think I would be able to manufacture bevel gearbox that'll hold up. That's why I'm looking for alternative motors and powertrain ideas.

Wait so you're going to try and build the worm gear from scratch? That is about 10 times harder than all of the more common solutions. Like a two stage belt reduction to BB freewheeling cranks if you want to pedal or a single belt reduction to BB freewheel cranks and then high rear reduction is way way easier, more reliable, more efficient and probably lighter. Could also probably get away with a single stage 219 chain reduction to the BB and still pedal it in some ranges. Your design of using a separate front cog mounted above the chainring can work but you have to be careful, you can't make it too small, you need a good amount of tooth engagement and you can't put it too high or you'll cause weird suspension issues.

Just out of curiosity though what is your plan to make this worm drive gearbox? Have you looked at some commercial units to compare their size, torque ratings, efficiency, etc?

So if I want to keep the ability to pedal effectively, I'd have to give up my rear brake to put a sprocket on the left side? What are the alternatives there? The single speed bikes that I've seen with a left side freewheel use the disk brake mounts.

If you want that much power and want it to be vaguely reliable I think yes. Honestly I just don't understand why so many want a bike with 3Kw+ of power and still want to pedal it. The only result is a compromise where the bike is just bad at doing both things, heavier, less reliable, more complex, more expensive, etc. Like if you want to pedal 500-1500w is very reasonable and at above 750w with a derailleur you don't need to pedal if you don't want to. Then everybody makes the point well what if it breaks and I can't pedal home, well you fix it or you walk but chances are it will be much less likely to break and there is no guarantee that your high power but also pedalable bike won't break in a way you can't pedal it because you're 5Kw just blended your drivetrain.

That's a bummer, I was hoping to avoid this with some engineering. I find big rear sprockets really unattractive. I think I'll keep looking into alternatives but running two chains might be the only real one. I'd rather decrease complexity as much as I can. Since I have an 12 speed rear cassette, I might be able to get away with not using my small gears much and instead using a much faster chain speed. The fastest I'll need to go on the regular is 35mph. I'm also considering switching to belt drive but I'm worried the cost to replace the hub and relace the wheel will be more than I want to spend.

Wait so you want to decrease complexity but already your design is way more complicated than most and you want to keep it that way because you think a big rear sprocket is ugly?

I've found the Turnigy-Aerodrive-SK3-6364 and the Flipsky equivalent. They both put out about a good amount of torque and rpm. I was thinking I could buy two and use a gearbox to join two of their output shafts. I could then use a single stage gearbox and have plenty of turque or flexability in where I want it in the bike. I could connect both motors to a single controller. I just need to figure out how to deliver that power to the back wheel. The whole system would have to run off 42v. Can I overvolt motors like that to run at 56v or 72v?

Ah nothing says simplicity like two motors joined with a gearbox, two controllers (no you cannot connect them both to the same controller unless it's a dual controller or you very very carefully time them and this is probably a bad idea), air cooling which will be required to get any real power out of those motors. Yes you an run those motors at any voltage you want but at some point spinning them faster than the recommended voltage will start to results in high magnetic losses. Also you can easily find larger outrunner motors so you can use just one, like an 80100 for instance but you want to consider the pros and cons of RC outrunners vs a decent sized IPM motor.

 

[StillDre]

Physically small motor + low airspeed + high load + small fraction of free RPM = a great chance to burn your motor.

When nobody serious does it that way, stop to consider why not. Gearing down makes good sense for your application. Using a motor designed to run fast at low torque with lots and lots of airflow does not.

I'm well aware that all motors like spinning fast and with as little load as possible (see the thread above). I was planning on gearing it down around 100:1 which should be doable with a two or three stage (stage one being the combo gearbox) gearbox. Using two motors would mean that each one would only need to make 0.5nm which is quite a lot less than what it would need to make on an RC plane. I was also exploring using forced induction or channels to force air past my heat making components given that I can make complex geometry with my 3d printer.

 

[StillDre]

In most cases bevel reduction is easier.

Bevel gears need to have shafts that are held at both ends and two of the ends need to be almost intersecting. To support those close shafts, I'd need to machine something that is beyond my CNC mill's ability since it would need to bore holes from the top and the side. I could 3d print something but it would need to be extremely stiff, maybe doable. I also just think worm drives are cool and wanted to hear how feasible they would be in an Ebike.

Honestly I just don't understand why so many want a bike with 3Kw+ of power and still want to pedal it.

Because once you remove the ability to pedal it, you've just built an overstressed electric motorcycle. It goes from cool modding to a terrifying deathtrap and that's not something I want to confront. The police will see that too, they'll notice that the bike has no ability to pedal it and that they're almost no way it's street legal.

Wait so you want to decrease complexity but already your design is way more complicated than most and you want to keep it that way because you think a big rear sprocket is ugly

Yeah, pretty much :|

nothing says simplicity like two motors joined with a gearbox, two controllers (no you cannot connect them both to the same controller unless it's a dual controller or you very very carefully time them and this is probably a bad idea)

That's a good point, thanks for bringing that idea back to earth. The biggest thing that would make it hard would be the two controllers I think. The gearbox shouldn't be too bad because it would just need one more motor mount and one more shaft.

 

[scianiac]

Ok I think you need to better define your power and torque goals here, because they seem all over the place. On one hand you don't want to build an electric motorcycle but you are also talking about big motors and torque numbers.

Like you said you want 100Nm output of a 10:1 worm gearbox, look at this gearbox, which is only rated for a max of 70Nm and is not super compact, probably pretty inefficient and uses alloy steel gears.

Because once you remove the ability to pedal it, you've just built an overstressed electric motorcycle. It goes from cool modding to a terrifying deathtrap and that's not something I want to confront. The police will see that too, they'll notice that the bike has no ability to pedal it and that they're almost no way it's street legal.

It's ability to pedal has nothing to do with if it's a motorcycle or not. A deathtrap or not is entirely down to design. So you are going to limit it to 750w or whatever the limit is where you live when riding on the street so it's street legal? The police don't care much if you can pedal it or not, like those people who get pulled over on sur-rons with a pedal kit and go "well technically I can pedal it so it's an E-bike", they care if you were doing 45 in a place you have no business being if they care at all. Riding responsibly in speed, fashion, and location is the only real way to ride a powerful bike like this around on the streets. I mean I bet at this point the majority of ebikes riding around on the streets are not street legal.

 

[StillDre]

Ok I think you need to better define your power and torque goals here, because they seem all over the place. On one hand you don't want to build an electric motorcycle but you are also talking about big motors and torque numbers.

Like you said you want 100Nm output of a 10:1 worm gearbox, look at this gearbox, which is only rated for a max of 70Nm and is not super compact, probably pretty inefficient and uses alloy steel gears.

It's ability to pedal has nothing to do with if it's a motorcycle or not. A deathtrap or not is entirely down to design. So you are going to limit it to 750w or whatever the limit is where you live when riding on the street so it's street legal? The police don't care much if you can pedal it or not, like those people who get pulled over on sur-rons with a pedal kit and go "well technically I can pedal it so it's an E-bike", they care if you were doing 45 in a place you have no business being if they care at all. Riding responsibly in speed, fashion, and location is the only real way to ride a powerful bike like this around on the streets. I mean I bet at this point the majority of ebikes riding around on the streets are not street legal.

Sorry, I wasn't being very clear. I am speaking specifically about my mentality when I'm building it and riding it. I don't want to be a Surron guy because I feel like those people lack discipline and I don't want to fall into that hole.
If I'm being completely honest, I want a dual sport that I can build in my garage but can't go on the highway. In a way, I am trying to limit how dangerous I can make this thing by keeping some resemblance to it's mountain bike origins.

The torque figure was determined because I assumed it would be enough given that it was around what some of the bafang kits claim. My thought process was that I'd rather overbuild it in terms of power and then never have to think about the thermal limits since I don't really know how to build around that. I am open to some reevaluation but I'm not sure how I should do that.

 

[amberwolf]

The torque figure was determined because I assumed it would be enough given that it was around what some of the bafang kits claim. My thought process was that I'd rather overbuild it in terms of power and then never have to think about the thermal limits since I don't really know how to build around that. I am open to some reevaluation but I'm not sure how I should do that.

You may be able to determine actual torque requirements (without doing the math yourself ) by using the simulators over at ebikes.ca , with your specific riding conditions and goals, and vary the parts / etc until you get the results you want.

It's setup by default for hubmotors, but you can use the middrive option to use gear ratios.

 

[amberwolf]

So if I want to keep the ability to pedal effectively, I'd have to give up my rear brake to put a sprocket on the left side? What are the alternatives there? The single speed bikes that I've seen with a left side freewheel use the disk brake mounts.

You can build it using a "sprotor", or you can stack the rotor and sprocket if there's enough space between hub and frame and you can set up the chain (or belt) to clear your caliper.

I'm also considering switching to belt drive but I'm worried the cost to replace the hub and relace the wheel will be more than I want to spend.

Belt drive usually means your frame has to be able to split one of the stays to get the belt on; depends on where it has to pass thru. Or else unbolt the swingarm from the frame and pass it thru the gap there if it has one.

 

[amberwolf]

I've found the Turnigy-Aerodrive-SK3-6364 and the Flipsky equivalent. They both put out about a good amount of torque and rpm. I was thinking I could buy two and use a gearbox to join two of their output shafts. I could then use a single stage gearbox and have plenty of turque or flexability in where I want it in the bike. I could connect both motors to a single controller.

If the shafts are physically locked, and the motors are identical (actually identical, not just "the same model" ) then as long as you line up the motors' stators and rotors perfectly with each other, you can then lock the rotors together physically so they cannot move relative to each other (and the stators locked physically together so *they* can't move relative to each other; implied by their mounting to the bike but specified just in case), you can use a single controller to drive the paralleled phases of both motors.

BUT: there will be half the resistance, twice the inductance, for the controller to handle, which is harder on the controller, so be sure it's designed to take that.

If the motors are facing each other rather than inline with each other or parallel with each other, there may be differences between them from spinning in opposing directions (mirrored) that can cause problems with the controller driving them. If they're made correctly that shouldn't be an issue, but this is not always the case.

You can line up the motors by first locking the stators together (mounted to your bike) then paralleling the same phase on each, and running a small current from a power supply thru these phases, to cause them to both turn to the same spot, then lock them down while so engaged. Then you can remove the current.

How well this works depends on how well matched the motors are.

Since they'd be sensorless (presumably) then you don't have to deal with differences in sensor positions and operation; just differences in how they were wound, magnet installation, etc.

Can I overvolt motors like that to run at 56v or 72v?

What is their RPM limit (before they come apart, bearings fail, etc)? What is their kV? You can use the kV to figure out the RPM at the voltage you want to use, and see if it exceeds the RPM limit.

If they don't specify an RPM limit, then they may be implying it by stating a max voltage for them (the actual windings will probably take at least a couple hundred volts, depending on the enamel used).


Also, have you heard how loud high RPM motors are under loads like this? And gearboxes? Plus your cooling system? Is that noise level acceptable to you?

 

[scianiac]

Sorry, I wasn't being very clear. I am speaking specifically about my mentality when I'm building it and riding it. I don't want to be a Surron guy because I feel like those people lack discipline and I don't want to fall into that hole.
If I'm being completely honest, I want a dual sport that I can build in my garage but can't go on the highway. In a way, I am trying to limit how dangerous I can make this thing by keeping some resemblance to it's mountain bike origins.

The torque figure was determined because I assumed it would be enough given that it was around what some of the bafang kits claim. My thought process was that I'd rather overbuild it in terms of power and then never have to think about the thermal limits since I don't really know how to build around that. I am open to some reevaluation but I'm not sure how I should do that.

Ok first how dangerous it is has nothing to do with what it resembles or what it's built in, the only real factors that effect danger is the rider and how good it is at doing what you are doing with it. Which was part of my point, if you try and make it do everything it's going to be worse at all those things which among other things will make it more dangerous. If you bring the scope in a little to what you are actually going to be doing with it things will improve. Even still every dual sport is a compromise.

Everything is a compromise so identifying what you want from the start allows those to be optimized for. If you want high speed and high torque you have to compromise on gearing, if you want to use a derailleur to try and get around that the compromise is how much power you can put through it.

So like based on bikes you've seen are we in the 2-3kw range or the 5-6kw or up around 8kw? Then you've got some general idea. Like is it primarily a commuter with some light trails or are you slamming down double black diamonds on the weekends. Ground clearance is only an issue if you ride certain types of terrain, in many areas that terrain barely exists. I do live and ride in an area where that is an issue and it's still not a huge deal, you are trying to ride around the rocks not directly over every one you see.

 

[StillDre]

If the shafts are physically locked, and the motors are identical (actually identical, not just "the same model" ) then as long as you line up the motors' stators and rotors perfectly with each other, you can then lock the rotors together physically so they cannot move relative to each other (and the stators locked physically together so *they* can't move relative to each other; implied by their mounting to the bike but specified just in case), you can use a single controller to drive the paralleled phases of both motors.

That's very intriguing, I think that'd be a fun experiment but that is definitely beyond what I'll be doing with my project.
Thanks for explaining that concept.

Also, have you heard how loud high RPM motors are under loads like this? And gearboxes? Plus your cooling system? Is that noise level acceptable to you?

I hadn't considered the noise of the motor, maybe I'll go look that up and see if any of the motors im looking at are overly loud.
I work with robotics equipment quite a lot and the loudest things in those robots are definitely the gearboxes and pulleys. I've actually grown to like the sound of spur gears meshing so I doubt they'll be much of a bother for me. I could also look into buying bevel gears for lower noise.
Electric motor noise though I find irritating. Thanks for bringing that into my considerations.
My goal was to have a fully passive cooling system with ducts taking air from moving the bike, and challeling it over my motor and ESC. My alternative would be a Delta or equivalent high output fan and some of them are quite quiet.

 

[amberwolf]

That's very intriguing, I think that'd be a fun experiment but that is definitely beyond what I'll be doing with my project.
Thanks for explaining that concept.

Well, it's not beyond what you'll be doing if you run two motors with one controller--you'll have to do that in order to do so.

(assuming they're not brushed motors).

(and assuming it's not two motor controllers in one box that calls itself a single controller but really isn't)

I hadn't considered the noise of the motor, maybe I'll go look that up and see if any of the motors im looking at are overly loud.

When they are high RPM motors, especially non-precision-made ones like most RC motors, under load they tend to be noisy (at minimum a high pitched whine, normally).

Depends on the motor, how it's driven, mounted, enclosed, loaded, etc., for the actual amount of noise.

I work with robotics equipment quite a lot and the loudest things in those robots are definitely the gearboxes and pulleys.

I'd guess most of these aren't designed to optimize for sound reduction.

Which kind of equipment? (i'm slowly working on a robotic wolfy project...so always curious about that stuff)

My goal was to have a fully passive cooling system with ducts taking air from moving the bike, and challeling it over my motor and ESC. My alternative would be a Delta or equivalent high output fan and some of them are quite quiet.

I doubt you'll get much airflow from any scoop-type inlet intended to force air thru, with a scoop of a size reasonable to not impact your aero too much and not make the whole bike look strange. Something like a NACA duct at the rear that uses airflow over the surface to pull air thru will probably work better at fast enough speeds but still unlikely to help much.

The problems usually encountered with small-size-high-rpm motors in drives like these are that they generate most of their heat very rapidly while under load at low vehicle speeds, often at startup from a stop, and cooling that depends either on motor speed or vehicle speed is usually insufficient to get the heat out. In these cases, there's not enough thermal mass in the motor to spread that out away from the windings, so without airflow to convect it out it doesnt' have anywhere else to go and builds up until motor damage can occur.

As long as there's enough airlow, or there's enough thermal mass and heatisinking to deal with the heat until it can get away, it'll work fine.

There's plently of reasonably small high-volume fans, but the really quiet ones are large-diameter and low RPM; some have fancier blade and enclosure designs to minimize turbulence (which is what msot of the noise is from). To help with this, keeping the inlet size larger than the fan size, not having any form of screening or grillework on inlet or outlet, and having the outlet size gradually narrow down to the motor size if it's feeding into the inside of the motor, or stay as large as possible to feed aroudn the outside of the motor, helps reduce turbulence which can reduce noise.

It may also help if you can set up the duct full of drinking straws, with their ends starting far enough from the fan itself to keep from creating turbulence in the airflow right at the fan, to segregate airstream into smaller sections that may flow better. This works better for longer passages with denser materials (liquids, generally) but it may work with airflow as well.