Hub Motors on Everything, Part V: Hub Motors on Even More Things

A month ago, I received 12 skatemotor-sized motor cans that I had commissioned through mfg.com by a semirandom Chinese machine shop. It was mostly an excercise of curiosity, but with 12 motor cans and no other motor parts, I decided to keep going with it – I went back to the semirandom Chinese machine shop and had them fabricate 12 sets of skatemotor parts: both endcaps, the center shaft, and the big ring nut thing that mounts the wheel.

That’s alot of little aluminum bits. What I have now learned is that someone else will actually hold the tolerances I indicate on my drawings, whereas I usually would just ignore them. The result is that the endcaps fit very tightly in the can (one is supposed to be looser) and the shaft-bearing fit is also a little tight, but it’s within sandpapering reach, not second lathe operations. The next time I do this, I have to specify a different tolerance band… though for now I do not mind making a single lathe pass and freeing up said endcap, or heating up the bearing and freezing the shaft for a real bearing fit.

Before I sent the dimensioned drawings off for quoting, I did a little re-engineering of the motor in the interest of easier assembly.

The shaft was remade as a straight pass-through, single-diameter part with key slot to clear wires. I actually designed it to mimic a 15mm metric keyed shaft such that it can form the starting stock, and the only operations done to it would be drilling and grooving.

Otherwise, the method of axial alignment has now switched to retaining rings. Yes, I know – snap rings are terrible, and I hated them the most when I was younger and eagerly disassembling appliances which always seemed to use them and I had no retaining ring pliers to cleanly remove them with and and… well, normal circular snap rings didn’t fit in this design anyway. I designed in two “low clearance” retaining rings (e.g. McMaster 97414A670) on each side of the stator. The inner ones keep the stator aligned, and the outer ones both space the bearings out and act as a restraint to prevent wire insulation from grinding on the bearings. I think this is a nice and simple system that is easy to assemble.

The stator is still a 50mm copier stator mounted on a custom 3d printed adapter hub –  there is no more “stator bore” shoulder on the shaft. I did try getting quotes for fully custom punched and stacked laminations designed to a slightly different geometry with thicker teeth to maximize torque production, but it would have cost way more than I wanted to spend (or had) – $1200 for tooling alone. While the stators would be cheaper per unit afterwards,  I think my best bet is still to find a supplier of completed (and epoxy coated) stators. The problem is that most places want minimum orders of 1000 or 10000 – small quantities for this kind of value added item is always difficult, and even if they cost a dollar each after that I still can’t dump that much.

The endcaps also got a little fatter. To fit the extra width of the retaining ring and wire-clearing gap and to give myself more room to put windings, I pushed out the interior width by 6 millimeters, necessitating protruding the endcaps out on each side. This makes the motors 42mm wide – while that’s just as wide as the old version 3 RazEr wheelmotor upon which this design is based, it does make them incompatible with the existing RazErBlades frame.

Gee, my stuff is always so much shinier when someone else makes it. Here, I’ve arbor pressed the static endcap into the can and pressed bearings into both. The bearing bore is actually a somewhat tight slip fit – again, another fiddle factor to adjust – so I dropped them in with a dose of 609 Loctite.

I have 2 sets of custom curved magnets remaining from the ‘blades project which didn’t see use in the final vehicle because I ran against a deadline. This set installed perfectly – the final one was a not-too-tight press fit into the bore, so this set doesn’t even need any glue to hold it in. However, it IS a little close for my comfort – there’s no guarantee this diameter will be the same across all the cans; so again, a dimension to change very slightly (on the order of 0.03-0.05mm or less).

I only have 1 50mm stator (the one shown), so either I’m gonna need to buy some more copier motors, ravage some more recycled office equipment, or find a source for them Yo Real Quick. I’m not keen on selling them to strangers over the internet just yet due to these sourcing issues, but it’s either that or build infinite hub motor powered vehicles…

Hub Motors on Everything, Part IV: Crap, Now I Really Mean Everything

I did something bad.

About a month and some ago, I made an account on mfg.com out of late night curiosity into the world of sourcing fabricated parts out to random people you have never met or talked to beforehand. For a while now, at least a year, I’ve occasionally gotten suggestions from people around the MITs to try putting some of my parts on there to be made in volume, or at least in quantities above what I cared making by myself. From those impressions, I figured mfg.com was like some kind of dating website for factory foremen, shop heads, and engineers. I now think that assessment is pretty spot on. It really is like a dating website. You make a profile and create a Request for Quotation, which filled out like one of those personal quiz things that abound on the Internet, except about your part…..as in, the part you want fabricated: what materials, processes, distance from you, and oh yeah include a picture and/or 3d model. The RFQ remains open for quote for a time period you determine, and you also set things like anticipated delivery date and payment and logistics terms (such as “T/T Ex Works”, which I took to mean “Paypal only, and Fedex it or something”).

But musing about the website is besides the point. I elected to at least try the system out, and spent a few minutes thinking about what part I found really bitchy to make that either took a long time or required precision operations, or both. The answer I arrived upon in my early morning stupor was motor cans. Specifically, I uploaded a CAD file of the skatemotor can, and to avoid the potentially really expensive one-off sample pricing, I said I wanted 10; they were to be made of whatever steel you found on the floor (though I specificied low carbon such as 1012-1018 or equivalent), I am the end user of the parts and do not really expect to need more of them (answers to such questions as “Anticipated Usage Per Year”) and I whipped up a basic 3-view drawing with labeled tolerances and finish requirements. Then I closed the window and let the magic happen.

The problem is it did.

I received in the neighborhood of 20 quotes, since I opened the floor up to all and besides the host of shady Chinese backwater shops (all privately messaging you information about how shady and backwater they are and linking you to their own websites, just like a dating website but with engineers), there were also plenty of local American houses and even a Croatian one. From my experience in mechanical engineering and machining & fab, the U.S. quotes were essentially what I had guessed they would be (mfg.com lets you set a target price, which is not visible to the bidder). In other words, I can’t afford it – sure as I am that “graduate student” is not the most highly paid of engineering positions, I’m neither a company nor have seed funding nor a rich uncle. And so American manufacturing slips yet another small delta: I elected to concentrate on Chinese quotes, which fell into three very distinct bands: they ranged from not much less than local Yankee-made parts, to reasonable considering the RMB to USD exchange rate and material costs, to… seriously, are you kidding me?

At the close of the quoting round, I was totally ready to ditch a nontrivial (for me) amount of money to see if getting my 10 cans machined at the seriously? rate was even a legitimate physical phenomenon. I was both intensely curious at how inexpensively outsourced parts could be, as well as a little scared at the realization that maybe the complacent U.S. manufacturing industry was deservedly upset about Chinese manufacturing.

Three weeks later,

It’s a real thing. I actually got updates as the parts were finished in rounds, and received an inspection report for 5 of them along with the order. A little cross checking with calipers found that at least one of them is closer to my specifications than I ever cared to hold.

They’re steel. Q235 type steel was the selection, material to be provided by supplier. Q235 is a Chinese standard roughly equivalent to type 1012 low carbon steel, so it was what I would have used anyway.

They’re shiny. Definitely a separate lathe and 4th axis/rotary indexing job, since I could see the faintest of grip marks on the inside of the very well finished cans – a 4-axis combo machine would have been able to do this in one setup. I designed this generation of motor to be made on such combo machines if it came to be the case.

They’re actually threaded very consistently and with great finish.

I actually got 12 of them – either 2 somewhere did not meet the inspection standard or they decided that just making 2 more to fill the damn box correctly was worth it, but hey, free cans. For the total price quoted, I was really not going to be disappointed if I got a chunk of cast scrap iron that was vaguely motor can shaped. My expectations have been well-exceeded.

The problem that I now must confront is, what now? I have 12 motor cans and no end plates, stators, shafts, and only one set of magnets to match.


and like, one spare tire.

next steps

While having metal shaped to my liking is nice, there are other very hard to find and/or bitchy parts involved in making one of these motors that I am actually curious about bypassing through the magic of sourcing.

1. Wheels. I’ve had to machine out wheels to fit over the rotors, and if I can get a “ring of polyurethane” kind of part molded, then that makes everything much better and more convenient.

2. The stator. Wound. Finished. Maybe not pre-terminated and with pigtail cables, but winding the motor takes forever. I know this is the kind of work that gets farmed out to marginally-paid migrant workers in Hobbyking Southern China somewhere, but if I can get wound stators for a reasonable price, that is actually the make-or-break for me regarding whether or not I will pursue this as a serious endeavour.

2a. Raw stators. Having a dependable source of stators that isn’t “this random-ass laser printer I pulled off the loading dock” is a start.

This might end poorly.