How can we make custom tuned ES edition motors a reality?

[macribs]

We all done it. Thought of it, or at least dreamt about it. A motor that ticks all the boxes right from the manufacturer. That is made to meet our dreams and needs. Tailored for e-bikers first and foremost. There have even been several threads about modify motors in an effort to enhance the stock motor to better suit e-bikes. Some even tried to make it it all, better wires, axles, cooling etc. And we have even tried to see if a tailored made ES edition was possible before.

What are the criteria that people feel is important to address for tailored made e-bike hub motor? (if we can make a hub happen maybe next step is mid motor?)

Lighter motor
More powerful or at least able to go full power for longer before saturation or overheating so cooling options
Higher quality parts - magnets, adhesive, more copper fill, thicker stronger and lighter axle, heck even higher rated halls or more attention to placement.

Let the ideas flow, first we got to gather what seems to be consensus of what we wish to change. Then we need to find a manufacturer that is willing to look into our requests and are willing to go an extra mile to make an awesome product. In the end, if product turns out to be as good as we hope the manufacturer should be able to sell that motor also to other demanding customers. So we will need to convince manufacturer that this custom, hi end motor will be a win for them as a company and their bottom line as well.

If we can drum up enough interest with a group buy or at least do one big order together this could actually happen.

How do we make a hub motor lighter? Madin88 had some very insightful thoughts on that. And I quote:

sidecovers with spoke flange (instead of heavy iron ring) safes about 1-1,5kg on common 3kW motors
- magnets with gap between (will not make performance worse)
- higher pole count (means that less iron is needed for the stator core and the iron ring for the magnet return)
- thinner lams (for more efficiency at high rpm, and to counter against the higher drag the higher pole count brings along)
- titanium axle

To that I would like to add, back iron ring shaved down as much as possible to reduce weight as well as look into carbon fiber side covers with spoke flanges. Maybe carbon fiber is not possible, or maybe we can source special made carbon covers from third party? (east gem got lot of lips here on ES lately?)

Room for liquid cooling
Ready rust threated and internally painted flat mat black to aid heat removal
Pre drilled hole for FF in one of the disc brake holes.
Ventilation hole with breather/membrane to equalize pressure
Swap magnets to higher rated ones
Use adhesive with higher temperature tolerance

What would you like to see in a custom tuned up ES edition motor? When we have a rough idea, an outcast we can contact several manufacturers to see if one of them will play ball. Maybe they can not meet us on all of our demands but as long as we are met on the most important once we should be golden. And if the motor turns out light enough maybe we can also have a version for 20 mm thru axle for double power 2 wd

 

[teslanv]

Samd, Shenta and I will get right on it.

 

[macribs]

Cool - - I really hope we can make this a reality.

Reading what I just posted it seems that most or even all of this is possible without changing molds or in anyway do costly mock ups. Except the thinner lams and those thinner lams should be a focus point for the manufacturer anyway. It should be possible to take a pile of donor motors and go to town on those, machine whats needed, shave off whats need and in general just make sure a little more effort and quality control goes into each and every motor. The factor that might work to our advantages are the price of chinese man hours.

 

[teslanv]

Low Labor costs are one thing, but the challenge is that most manufacturers are hesitant to pour money into new molds, processes and parts catalogs unless there is a large enough demand. The main problem I've always had is that the "MOQ" (Minimum Order Quantity) for anything even semi-custom is higher than I would want to front the cash for.

Custom side covers (even just machining stock ones) will require an MOQ of at least a couple hundred per side for most manufacturers to entertain making them.

Contrary to the popular saying "Build it and they will come", I think in this case, we need to begin by defining what the market is for something like this.

How many motors of such a design could be sold to not only us Motorheads, but to anyone else?

The Demand must exist before a manufacturer will even consider investing the resources to make it. If we can prove that, then we have a viable path to production.

 

[Punx0r]

^This

The first question you'll get back from any manufacturer will be "how many do you want". They can't begin to indicate price without that.

I suspect the expectation here will be to get a small batch of what are really prototypes/samples made at high cost to the manufacturer and expect them to recoup those costs by subsequently selling them to an imagined mass market.

Here's a good existing thread hashing out a theoretical design for a lightweight hubmotor: https://endless-sphere.com/forums/viewtopic.php?f=30&t=57371

And another, related thread on optimising motor geometry for maximum torque output Vs. motor mass: https://endless-sphere.com/forums/viewtopic.php?f=30&t=78923

 

[Rix]

scrib'd right on mac.

 

[macribs]

@teslanv thinner lams are the way of the future, right? Thinner lams has already been done for a while, just not thin enough yet. Maybe we can speed up that process?

@punx0r thx for great links. Must read up, both unread for me I think.

I get your concerns here, but at this point we can't just stop and say there will be too many obstacles so we can't push forward. For now lets keep this on a theoretical level and create a virtual best case dd hub. Where there are no limitations. Once we have I rough idea of what most super users crave then we can take it further and start pitching to manufacturer. Lets not forget a whole lot of improvements can be done completely without altering molds, with just more time spend on each motor.

Later on when we have a more clear view of what we are actually targeting we can look into getting people to sign up for the first batch, then pitch idea to China and together with the manufacturer we can see what can be done without braking the bank. I mean if we end up with a 3.000 $ hub chances are no one will get it. So we might end up ditching some of the most costly modifications. Again, which ideas to scrap that is ahead in time.

No lets the creative juices flow, any and all ideas for what will make for a better, lighter, stronger, less heat troubled motor are welcome at this stage.

 

[cal3thousand]

Well... let's talk about the things that cause use problems with our motors. We've already identified the issues, just need to list them. The solutions are sometimes readily evident and sometimes not, but what always remains are the gripes.

So ask yourselves, what have you had to fix?

Obviously heat shedding is a big topic and anything they can do to allow heat to transfer from copper to shell more quickly is desirable.
Heat creation is also another big topic. Eddy current reduction due to stator design
Axle strength and torque handling at the dropout
Wire thickness and routing - we want fatter wires for less resistance and I²R loss reductions.


And of course some things are a compromise and we can't have both. For example, it's difficult to have a wide stator yet have room for many gears. Some want a cassette option, but that would interfere with axle thickness (as recently brought up on ES FB regarding the new MAC vs the eZee hubs with cassette). Another design issue is that we want room between the disc brake and the hub for calipers to clear, but wider and taller stators are being demanded as well.

 

[amberwolf]

If you're going to make a custom motor anyway, skip the entire axle/wire thing, use a good axle design that has no wire-exit-caused weaknesses, and use a larger bearing with an internal spacer to fill up it's ID, except for the space where the wires themselves go.

Then there is no limit to the wire size you can use, and you can add any kind of liquid cooling you like via that spacer as well if you like.

And the axle can be made in any way you like--even modular, as a thru-axle, if preferred, *and* it can be large enough diameter to enable an integrated torque arm as part of it, not having to depend on tiny little axle flats at all--you could use a normal round axle for stronger threads for the axle nut to secure it (less risk of stripping it during mounting/tightening).



If you use a large enough bearing you could even add an airflow duct inside the ID of it.

 

[Jestronix]

first thing is would u pay double for your dream motor ? Second can an after market process be done to achieve it at that cost ? Apart from thinner lams etc maybe this can happen without min order quantity.

My list,

Bigger axle - to feed thicker phases and or water cooling.
Integrated cooling fins, can hub motors actually manage the head shredding we are after with Fero and fins ?
Wider stator, I like 55 but reduce radius ?
Double temp and double halls (becoming norm)

Great thread macribs

 

[liveforphysics]

Miles already made a design I would happily pay a few thousand bucks for. Absurd low mass and efficiency and high specific torque.

 

[macribs]

If you're going to make a custom motor anyway, skip the entire axle/wire thing, use a good axle design that has no wire-exit-caused weaknesses, and use a larger bearing with an internal spacer to fill up it's ID, except for the space where the wires themselves go.

Then there is no limit to the wire size you can use, and you can add any kind of liquid cooling you like via that spacer as well if you like.

And the axle can be made in any way you like--even modular, as a thru-axle, if preferred, *and* it can be large enough diameter to enable an integrated torque arm as part of it, not having to depend on tiny little axle flats at all--you could use a normal round axle for stronger threads for the axle nut to secure it (less risk of stripping it during mounting/tightening).



If you use a large enough bearing you could even add an airflow duct inside the ID of it.

Ideally we should come up with the new wine, realistically we probably won't due to cost of such motor. Most likely the low hanging fruits and easy changes will be the starting point, and we might get producer talked into doing custom machining.

I am trying to picture this axle and a larger bearing with internal spacer but the spacer thing kind of make the visualization conflicted. The modular part is very clever, plus that will make for a good choice for both those who like to run front hub/2 wd and rear wheel dd drive.
Could you do a quick "back of a napkin" kind of visual aid drawing?

 

[amberwolf]

It's been posted before, by me, John in CR, and others; sorry I dont' have a link ATM. maybe when I get back from work.

a comment I posted in the other thread this split from in response to something about disadvantages of this method:

You want higher power to the hub with greater efficiency, so you want thicker phase wires, and you want a stronger axle; these two are mutually exclusive if you're carving up the axle to make the wires fit.

(although having an axle made of better material than recycled beer cans would help even the hollow axle designs ).

The larger bearing isn't going to make any significant extra resistance in rotation. If you're worried about that, use one with larger balls and fewer of them. There's a good number of standard bearing types and sizes that would fit this application, in various configurations; look around mcmaster-carr's site for some examples (am heading to work so no time to dig up a link).

The only potential issues are with a freewheel or disc brake--you might have to sacrifice having one or the other to get a *really* large bearing in there. But you can be creative, and have disc bolts for a larger ID common motorcycle-type disc added to the side cover outside the bearing area. You want more power to get going, you might want more power to stop anyway, right?

 

[brumbrum]

Ditch the rear brake disc mount and use only regen and front brake. Controversial, yes, but would free up space and further able modification of motor design.

 

[Jestronix]

Ditch the rear brake disc mount and use only regen and front brake. Controversial, yes, but would free up space and further able modification of motor design.

Progressive ebrakes rock, not sure the standard ebrake is good enough, either too hard or too soft. On the street its fine.

 

[brumbrum]

If these motors are going to be expensive,, which they will for what will be a tiny production run for a Chinese manufacturer, then expect the end user to have matching expensive controller with variable Regen. At the end, these bespoke motors being designed for the forum members will be mainly for running very high power with very high power equipment. Having variable regen is not too much to ask those who are likley to apply this dream product. Even the best mtb disc brakes are no good for stopping a 50+mph bike, and even if they are you would be changing rear brake pads every few weeks. Just my thoughts.

 

[Jestronix]

Has this died ?

 

[macribs]

I hope not. I am trying to read up on some of the tips n threads posted in the mxus 5k thread.
I think maybe the talk about price spooked people?

Sure if you put extra money into a custom motor, it makes sense to use a hi powered controller. That does not mean controller has to be an expensive Adaptto or a Gen4 size X. There are powerful sine waves controllers out there with regen and field weakening from a couple of hundreds up to around 400$ that would work well with 10-15 kw hub. And 500-600 bucks controllers with twice that peak.

Without making promises I think for price at the most we can risk motors to land about 30-60% on top of what they already got, ie mxus 3K, QS 205 etc. This could be be even lower if we can convince manufacturer that our changes will also benefit other customers, and also the amount of motors ordered via group buy could affect the final price. I've been chatting up this guy that do regular orders from Chinese manufacturers, and according to him most manufacturers are keen on doing custom work, as they get to charge extra plus they get a new product with even higher quality to offers their regular clients.

But before we can pitch this to China we need to get down and dirty and decide what mods we need, which mods are most important and possible what mods can we live without in case some mods will need to be dropped. If we can get a motor that runs cooler, can stay on peak power for longer that also reduces unsprung mass while having thicker wires, pre painted internals etc it will be well worth it. Someone is gonna need to make a push for even higher quality motors, the manufactures will not go to length to increase quality unless we push them forward.

In the end if it can't be done and no manufacturer will meet our requests, and we have put our best effort into it we might have to accept status quo for now and look for mid motors to fill our power urges. But we don't know if a custom ES hub motor is possible unless we try.

 

[Rix]

But before we can pitch this to China we need to get down and dirty and decide what mods we need, which mods are most important and possible what mods can we live without in case some mods will need to be dropped. If we can get a motor that runs cooler, can stay on peak power for longer that also reduces unsprung mass while having thicker wires, pre painted internals etc it will be well worth it. Someone is gonna need to make a push for even higher quality motors, the manufactures will not go to length to increase quality unless we push them forward.

So true Mac, hell just settling on axle widths, say 150mm face to face would need to be uniformly agreed upon. I am open to debate the small stuff, but mandatory mods as I see it would be high quality magnets, hi temph/strength glue for the magnets sensors, phase and sensor wires, bearings along with any other mods that reduces the mass. Should we start a list and see who is on board with what mods???

 

[marcexec]

If you're going to make a custom motor anyway, skip the entire axle/wire thing, use a good axle design that has no wire-exit-caused weaknesses, and use a larger bearing with an internal spacer to fill up it's ID, except for the space where the wires themselves go.

Then there is no limit to the wire size you can use, and you can add any kind of liquid cooling you like via that spacer as well if you like.

And the axle can be made in any way you like--even modular, as a thru-axle, if preferred, *and* it can be large enough diameter to enable an integrated torque arm as part of it, not having to depend on tiny little axle flats at all--you could use a normal round axle for stronger threads for the axle nut to secure it (less risk of stripping it during mounting/tightening).



If you use a large enough bearing you could even add an airflow duct inside the ID of it.

i.e. the way Enertrac does it for motorbikes, also with liquid cooling as an option.
Happy to supply detailed pictures - I have one out atm.

 

[macribs]

@marcexec yes pls share close ups pics and details. The one good pic at enertrac's site is so tiny it is hard to see it all.

 

[marcexec]

There you go:


(Edit: yes, I know I need to hoover the filings away)

 

[macribs]

Miles already made a design I would happily pay a few thousand bucks for. Absurd low mass and efficiency and high specific torque.

Struggle to find the thread, anyone got a link/bookmark?

 

[PRW]

Miles already made a design I would happily pay a few thousand bucks for. Absurd low mass and efficiency and high specific torque.

Struggle to find the thread, anyone got a link/bookmark?

maybe this one? Just a guess... https://endless-sphere.com/forums/viewtopic.php?f=30&t=57371&start=375&hilit=design

 

[markz]

I have casually read a few of these thread topics and its a noble idea, but price is always a factor.

From what I gathered, the laminations could be done at a reasonable cost, but to get them thinner I have no idea.
The axle could be a custom piece done by a home hobbyist with a lathe, same goes for the cover plates. Wouldnt it then be possible to outsource that end of it to China, some smaller shop could definately do it, but then how do you go about finding that local smaller shop to lathe up some axles and cover plates. I really dont see that as an issue to get done, but how much would that add to the price. Then while we are at it, why not have a huge solid piece of alum inside for a massive heat sink, there is no air movement in there, unless you drill holes, but a solid junk of alum would be good right.

Axle
Cover Plates
Bearings
Laminations
Stator
Winding