SHENGYI new style lighter 9c type DD motor

miuan

10 kW
Joined
Nov 26, 2009
Messages
992
Location
Slovakia
Hi,

I was looking for a substitute for my smoked em3ev DD hub and discovered this motor offered by a local vendor in Slovakia.
http://www.elektrobike.sk/elektrobike/eshop/6-1-Prestavbove-sady-nad-250W/0/5/73-prestavbova-sada-Sinus-500W-bez-baterie
Looks to be same diameter as 9c, with a cassette spline, disc mount and fully 135mm compatible. Also weighs only 5kg, which assumes narrower stator.
The vendor also said it was more silent than 9c he sold before.

Found a similar looking motor here:
http://www.cycleman.cn/Pro_shen.asp?ArticleID=252#

I will have it in my hands tomorrow, time to make some measurements!

Anyone familiar with this motor?
Viktor
 
Just got the motor in. It has a 12mm axle, probably due to the cassette splines requiring a smaller bearing. More pix to come in the evening.
 
Very interesting. Lighter is good for many, as is being able to use a 9 speed cassette.

The motor, if it has a narrower stator than a 9c style motor, should still be able to handle 48v 20 amps. If a 28mm stator can take 3000w for awhile, why couldn't a 22 mm stator take 1500w is what I mean.

Still quite sufficient for those running a typical street commute, and now it's easy to have dd reliability and 9 speeds.

Lets see a really detailed review of this, once you have 500 miles or so on it.
 
Website of the producer:
http://www.syimotor.com/products_detail/&productId=332e51fa-ec5d-4570-a45d-5832ec911e54&comp_stats=comp-FrontProducts_list01-004.html

Photos of the motor:
https://drive.google.com/folderview?id=0B90X1vErDqMLQkdGZTFjcmo1d0E&usp=sharing

Haven't opened the motor yet since the vendor warned me of an o-ring that was extremely difficult to put back on upon reassembly. Eventually I'll open it though.

Cassette spline looks good, freewheel has 24 degree engagement intervals. It weighs 4.9kg including wires, which puts it close to MAC territory.

Side covers look good. The magnet ring is not a separate part woth two spoke flanges like on 9c, instead it's a part of drive side cover/spoke flange unit, bolted to the other side unit. The seam between side covers is sealed by silver tape.

There is ample space for brake caliper. I measured the inner spoke flange spacing at 27-28mm (maybe it's actually the same 29mm as 9c), outer is almost 10mm wider. Flanges are almost equally distant from each dropout, so wheel dishing seems well sorted. Hub Flange Diameter seems to be the same as 9c = 233mm between opposing holes which are 3mm wide.

The axle is a disappointment. It has a mere 4mm section of 12x10mm axle flat on each side, beyond which the axle is machined into a 9mm round threaded rod that fits a tiny 15mm nut, compared to 18mm on most geared and 21mm on 9C DD motors. You can either utilize the provided torque washers for low power applications, or susbstitute them with 4mm steel torque arms.

Why the producer chose 9mm axle remains a mystery to me. The good thing is, wires exit from the hub rather than end of axle as was the case with 9c. The provided waterproof plug is also a welcome addition and will serve most who run a typical kit controller. I would not trust the skinny pins for more than 60 phase amps though.
 
The 9mm round axle should be strong enough, and with threads all the way around, less likely to strip like the half a thread axles.

But the problem I see, is that with this design, you will have to have dropouts strong enough to need no torque arm. This is a real problem if you plan on running 3000w. Bolt pressure alone would be sufficient for 36v and 20amps, in a steel frame.

I see this as a minor improvement. If the magnets are narrower than a 9c, you should be able to feel it when you ride it, assuming the winding speed is similar. Looking at the caliper pic on that web page, it seems like it's plenty wide enough to have a similar stator to 9c, yes, conhis, golden etc. About 28mm.
 
Finally took the hubbie apart today and updated the photo album:
https://drive.google.com/folderview?id=0B90X1vErDqMLQkdGZTFjcmo1d0E&usp=sharing
Most of the photos are self explanatory.
The stator has about one half of total motor weight. Stator steel is 25mm wide and 205mm in diameter, Magnets are short of 24mm. It's too late today, more comments and speed test later.
Bearings are 6904RS for non drive side and 6001RS for drive side + 7901VRD for freehub support, the latter turning a little hard to begin with.
There is an extra wire that acts as a speed sensor. It uses a magnet fixed to the inside of the non drive side cover to switch a solid state sensor on the hall pcb.
The white device epoxied on the windings is a 105C temp switch that shuts off +5V hall supply in case of overheating.
 
Thanks for the update. 3-4 mm narrower than regular 9c or muxus then. But that doesn't mean it's wimpy, just less likely to tolerate a huge overwatt, like 3000.

Bet it still runs like a champ at 1000-1500w.
 
This motor can't compare to the torque of 9c or Magic pie, it's rather a no-nonsense fully compatible motor (135mm dropouts, disk brakes, 9-10speed cassettes) for low to medium power solutions, barely heavier than a 500w gearmotor, yet less noisy, robust and regen ready. Would also be a killer choice for 2WD conversions.
 
Thanks for the pics and info!

24mm stator isn't too bad. More like a front 9C; not really powerful. I don't know what the point of the motor is other than it being cheaper i'm sure due to less materials. A big geared motor like a MAC/BMC already kicks the 9C's ass.. even this is a couple pounds heavier than a MAC/BMC.. but she'll regen, so hey..
 
? My front and rear 9c and muxus motors all have the same 28mm. I have front rotors running on rear stators.

But I do agree, a 24mm motor should be fine for most road use. Lighter can be very good.
 
Performance wise it should be very close to the front H series Crystalytes that also happen to have a 24mm stator of similar diameter.
 
dogman said:
? My front and rear 9c and muxus motors all have the same 28mm. I have front rotors running on rear stators.

But I do agree, a 24mm motor should be fine for most road use. Lighter can be very good.

I might be thinking of the crystalyte H24xx front motor... my bad :oops:
 
miuan said:
This motor can't compare to the torque of 9c or Magic pie, it's rather a no-nonsense fully compatible motor (135mm dropouts, disk brakes, 9-10speed cassettes) for low to medium power solutions, barely heavier than a 500w gearmotor, yet less noisy, robust and regen ready. Would also be a killer choice for 2WD conversions.

Why would that be a killer choice for 2WD when the motor has like 2 time less torque than a single magic pie? I'd rather put a nice good DD rather than 2 of them
 
Done a lot of progress yesterday, with a lot of frustration in the end. Built the wheel with a 26in Rigida Andra 40 rim and Sapim Strong spokes, put on Stan's FR rim strip and goo under the massive 2.5in Hookworm with utter ease.

Then came the difficult part - fitting a torque arm, yeah.
Realized the dishing was not so perfect (too far to the non drive side, as usual) so I decided to skip the torque arm on drive side to allow the cassette to sit as close to the frame as possible. The only way you can fit a TA on this motor's axle is from the inside of the frame on either side, and using a 135mm dropout, I realized there is no stretching needed if you fit a TA on one side only. I used a BMS Battery 4mm steel TA on the disc side. Drilled the frame dropout on 2 points to ensure the TA has nowhere to move. Due to the disc rotor, there is very little space for a nut on the inside of the TA eyelet, but with the proper tool, it can be done. Added 4 pics to the gallery today (no.3314 to 3318) that should explain how exactly I did it. Feel free to ask questions if anything's unclear, or suggest a better method of securing the axle.

Finally, the real bummer. After I got the motor spinning only very rough, I realized one of the hall sensors put out 5V all the time. So the goddamn motor was broken to start with. I smoked a phase on the previous DD motor using this 12fet controller, so chances were it was the controller's fault, but I seriously doubt it. Will definitely give a sensorless controller a try before getting into the motor again. D'oh!
 
Arrgh!

I agree with the above comment, good for a two wheel drive. Why? 9 speed cassette for one, and for two, at 1000w each a double motor would have a very robust 2000w pull up hills. Plenty, even for my gigantic cargo bike loaded for camping. So it would not require a pie or any other motor if you run two.

Of course, more is better, but two of this motor should do all you ever really need.
 
For less weight, you could get a bigger DD that is more powerful, more efficient, and not have to think about the issues that come along with mounting a hub motor on a front fork.

Two of these motors would be 22lbs and be 1000W nominal combined.
One crystalyte H40xx would be 18-20lbs and 2500W nominal and have 5% higher peak efficiency, and better efficiency under a high load.

Or compared to a Crystalyte HS35xx / magic pie; you get about ~1500w nominal power and a few % more efficiency for 16lbs, versus 1000w nominal with two of these, but 6lbs more weight.

There's really no point to running two unimpressive hubs.
 
neptronix said:
For less weight, you could get a bigger DD that is more powerful, more efficient, and not have to think about the issues that come along with mounting a hub motor on a front fork.

Good points about the fork issues, however they can be avoided altogether if using the right equipment and hardware. Front wheel skidding IS definitely a problem on some setups though.
Compared to a single H40XX, two of these would be equally powerful, better cooled and much better balanced. Also less prone to broken axle/spoke problems due to half the torque and lower weight.
I rode a MP2 on a Kona Stinky with 24S lipo and 40/100A. It had great torque, but at some point the extra torque only wanted to flip the rear heavy bike over.
 
miuan said:
Compared to a single H40XX, two of these would be equally powerful, better cooled and much better balanced. Also less prone to broken axle/spoke problems due to half the torque and lower weight.
I rode a MP2 on a Kona Stinky with 24S lipo and 40/100A. It had great torque, but at some point the extra torque only wanted to flip the rear heavy bike over.

No.. a h40xx is rated at 2500w continuous. These are going to be rated at about 500w continuous, due to their width.
The h40xx has much higher efficiency, so both the shengyi and big DD motors are going to operate within their thermal design the same, really. The magic pie isn't the most efficient motor in the world, but it has an awesome air cooling design. It will for sure provide 50% more continuous power than a pair of these motors, still.

If you spoke up a wheel properly, you will never have spoke problems even at extremely high power.

Yes, front/rear balance is a problem. The longer the bike, the better though. The more upfront you can put the rider, the better, as well. Put the battery up front also... and get a throttle tamer.. you can combat the effects of high power with a rear hub; it just takes a little planning.
 
neptronix said:
No.. a h40xx is rated at 2500w continuous. These are going to be rated at about 500w continuous, due to their width.
The h40xx has much higher efficiency, so both the shengyi and big DD motors are going to operate within their thermal design the same, really. The magic pie isn't the most efficient motor in the world, but it has an awesome air cooling design. It will for sure provide 50% more continuous power than a pair of these motors, still.

The thermal design, as you mention, is the bottleneck, as is efficiency. To me it seems though that your assumption of 5 times the sustained power of HX40 vs 24XX is exaggerated, as both get very similar cooling surface. Now get two of them and the cooling surface clearly favors the dual hubbies, which may offset even more than the difference in efficiency makes up.

But let's get back to OT. Yesterday I swapped the original 12fet for a sensorless one with 25A and regen, and today I will take it for its maiden ride.
 
Thermal design and efficiency are what generally determine what a motor can handle continuously.
Just 1% of efficiency difference under a high load has a dramatic effect on heat output.

Attached is a spec sheet for the Crystalyte H4080 under 72v/50a. It hits it's peak at 90% efficiency at 2,600w input, and only has ~500W of heat to shed. A motor that large can do that for a long period of time.

The thinner the stator, the more end turn losses you end up with typically in most hubs; that's why efficiency tends to increase the wider the motor is. The shengyi motor is likely 80% efficient at peak. It has a good amount of thermal mass in proportion to the stator, but your efficiency lower. It just can't be pushed as hard past it's rating for that reason either.

Anyway, if you bought the motor, enjoy it; discussion over. I just wanted to point out that a dual drive is not the way to go for power ( unless for some reason, you want a less power dense motor ).

miuan said:
The thermal design, as you mention, is the bottleneck, as is efficiency. To me it seems though that your assumption of 5 times the sustained power of HX40 vs 24XX is exaggerated, as both get very similar cooling surface. Now get two of them and the cooling surface clearly favors the dual hubbies, which may offset even more than the difference in efficiency makes up.

But let's get back to OT. Yesterday I swapped the original 12fet for a sensorless one with 25A and regen, and today I will take it for its maiden ride.
 
My ride is a hybrid 2WD - a DD rear with regen, geared front with freewheel that I disengage while cruising or at lower power scenario (most of my riding). It is a production 135mm dropout bike (Kona Ute) that has no means of fitting anything with wide stator without losing the rear gears. That's why I have to use narrow motors. The Magic Pie is a nice choice, but I'd have to up the voltage to 24S to get the same speed and performance. HS3540 would be nice, but according to the simulator, it seems no more efficient, and overheats sooner than 2 of these (pic attached).
A heavy rear motor is a problem also because I regularly carry the bike up/down staircases, and while I can cope with high weight of the bike, I need it to be balanced well. I ride all winter, frequently in the snow, and the 2WD bike is so much nicer in these conditions, actually mandatory to climb the sorts of hills that I encounter, unless I change to stubby tires.

The H4080 sheet you included surprised me. I have no knowledge of this motor, but it seems strange that its peak eff. is so much higher than its barely narrower 3540 peer, and so much farther down from the no load speed. I suspect Justin must have used a much different approach, or perhaps the controller on the H4080 was a sine wave type.

So much for theory, today's ride to work was quite pleasant. The single shunt chinese controller has nowhere near the power of the original 12fet, but since both the motor and controller were barely warm to touch after the ride, it still has some more punch to offer. And the regen works at 20s lipo without any mods!
 

Attachments

  • SimClyte.png
    SimClyte.png
    35.2 KB · Views: 1,658
Simple answer:
- dual hubbies are more efficiency when torque or acceleration is required. It reaches faster top efficiency. In city commuting, dual hubby is most efficient.
- for long straight road, single motor is more efficient
 
Back
Top