Q100 and Q128 with 0.35mm laminations

Jasonv8z

1 W
Joined
Aug 8, 2010
Messages
59
Location
San Francisco, CA
I've been working with the factory that makes Q100 and 128 motors. I had them make (at my cost) Q100H prototypes with 0.35mm laminations, as well as 0.5mm Q128 cassette motors with 135mm spacing, all painted black inside and out for heat dissipation.

The 0.35mm lamination Q100H resulted in about 20% improvement in efficiency (and likely power handling) over it's 0.5mm counterpart to 84% peak. I am currently running the 330RPM 36V version with 12S using a stock S06S. Top speed no pedaling is about 28mph @ ~580W input because of the S06S current limit. WIth pedaling, top speed is about 34mph. I can also cruise over 30mph all day long without any issues of overheating. I only weigh 150lbs and am a strong biker so take that into consideration.

I haven't tried the 0.5mm Q128H yet, but I'm only expecting modest gains on flat ground over the 0.35mm Q100H based on the dyno information. Looks like significantly more torque however.

I plan on having them make 0.35mm Q128H motors in the future but I could not meet their MOQ this round. I would expect roughly 40% more output over the 0.35mm Q100H based on mass (3kg vs 2.2kg).

I'm in the process of building a dyno, but for now I have attached the test data they gave me.

20150707_160047.jpg


Pricing for these motors would likely be around $175 for the 100H and $225 for the 128 cassette, shipping from USA. Do you guys have any interest in these motors? What RPM @ 36V and hole configuration (eg 32 or 36H) would you like to see?
 

Attachments

  • ZCK100F38V.pdf
    63.5 KB · Views: 323
Holy crap! it looks like the '100' motor realistically becomes a 450 watt continuous motor from that dyno!!
What a difference a lamination change makes!
6.6lbs for a 450w cont. geared motor is actually quite good anyway, you'll notice that there are no geared motors in that power range that weigh less than like, 10lbs.

You may have a winner on your hands with that one.
But.. what about making a modified version of the Q100H instead? that motor produces more power per weight due to it's much higher gear reduction..

What would be really interesting is a 135mm dropout fitting Q128H with 0.35mm laminations that's wound to produce ~350rpm on 36v. BMSBattery rates that as a 800W continuous motor, but they only sell a 201rpm @ 36v wind.. WTF?

If you could get such a motor produced and provide a test sheet, i would be interested in buying one as a tester at the very least.. telling others about it at best.
 
ps, i would change the title to 'custom 0.35mm lam. Q100 motors & 135mm Q128' or something, because the casual passerby might think that this thread is about building bikes with such motors and disregard the 0.35mm part..
 
I'd be interested in sharing vols and distributing in Australia...
 
neptronix said:
But.. what about making a modified version of the Q100H instead? that motor produces more power per weight due to it's much higher gear reduction..

What would be really interesting is a 135mm dropout fitting Q128H with 0.35mm laminations that's wound to produce ~350rpm on 36v. BMSBattery rates that as a 800W continuous motor, but they only sell a 201rpm @ 36v wind.. WTF?

If you could get such a motor produced and provide a test sheet, i would be interested in buying one as a tester at the very least.. telling others about it at best.

I'll ask about this. They sent me their "newest design" Q100 which seems pretty similar to the Q100H. motor seems a bit smaller than the Q100H (left), so we could probably gain a bit of power at the expense of weight. I requested a 12.5:1 reduction. Don't know what I actually got.

20150710_161819.jpg



I think the real winner will be the 0.35mm Q128 cassette with 135mm spacing. It's 3kg, but you can save ~150 grams from using a cassette instead of a freewheel. Most of the rest of the added mass is due to copper and iron in the motor, and a stronger clutch. I'd actually like to go to 0.2mm laminations, but they want me to pay $15k in tooling costs first. If a few more people are interested in the 0.35mm Q128, I'll go ahead and have them make a batch.
 
Very cool! Are you working straight with Ananada or going through BMSbattery?
 
I can't tell what is what in that photo.. but it does look like you've got a single reduction with the one on the right.

It's really impressive that with a single reduction, you see the 0.35mm motor become more capable than one with dual reductions.. :)

As for my self, i'd take the 4% efficiency bump at a higher weight in order to make that same power any day vs. the BMSbattery 0.5mm lam version any day of the week. But the Q100's just can't meet my power demands.

The Q128H does have big problems with being driven by most controllers to a speed where it's extra power capability can be used. It's not practical to even bother with 0.2mm laminations until you can find a controller that can push at least 500 wheel rpm in a 26" wheel ( where the geared motor will be spinning at 6,000-8,000rpm internally x what, 8 poles? that's RC Motor territory. )

I dunno, you may have a hit on your hands with the 5:1/6:1 reduced Q128 with 0.35mm lams. Perhaps it becomes a 85% efficient 600-750W rated motor?
 
I haven't been a big fan of the smaller hubs, but I have been watching them. The Q100H seems to be popular in Europe, and I imagine its because it is small enough that it is easily "believable" as a 250W "legal" electric motor. The dual-reduction means the motor spins much faster than the standard 5:1 of the larger common single-reduction hubs.

I was initially suspicious of the efficiency improvement claim, but...after some thought I realized that the standard 0.35mm lam stock is also a higher quality of metal than the standard 0.50mm thick "most affordable" lam material. That, plus the higher the RPMs, the greater the effect of moving up to a thinner lamination.

Does the Q128 have a single reduction?
 
neptronix said:
I can't tell what is what in that photo.. but it does look like you've got a single reduction with the one on the right.

It's really impressive that with a single reduction, you see the 0.35mm motor become more capable than one with dual reductions.. :)

Actually, both motors have compound gears.
It's a bit of an optical illusion, but the clutch housing is hiding the other tier of nylon gears.

What is interesting is, since the .35" laminations are thinner, the entire stator is thinner, and the extra space must be taken up somewhere.
I assume the cusomized motors must have different axle shafts with different placement of the shoulders.
Seems to me that if that is the case, the customized CST could now have a perfect 135 m/m drop-out width.
 
We don't know if the stator is thinner or not without taking it apart. Usually during an upgrade of laminations, they'll keep the motor the same width by adding more laminations to get the same stator size.
 
d8veh said:
mlt34 said:
Very cool! Are you working straight with Ananada or going through BMSbattery?
You mean Aikema, not Ananda?
Maybe we are still playing the "Don't Name the Factory"game.
The photo showing the group of motors, shows they carry the lazer etched script of Ananda and not the sticker used by Aikema.
 
Sorry. These motors are from Aikema. Perhaps Ananda makes them as well. The black prototype motor (right) in the picture is a double reduction, but smaller than the Q100H (left). There should be some power to be gained by using the Q100H motor size over whatever they sent me. I believe the Q128 is also a double, although I'm not sure of the ratio.

20150710_161819.jpg




My calculations (feel free to correct me) put the Q100 at 30mph 26" wheel = 373 RPM
373RPM x 12.5 reduction = motor 4662 RPM

4662 x 8 pole pairs/60 = 37296 ERPM = 621.6 hz


I got M19 steel core loss data for 0.45 and 0.35mm laminations at 600hz. From the data I would guess 0.5mm laminations to have well over 40 watts of iron loss/lb near saturation. Going to 0.35mm laminations reduces loss to around 27 watts. Figure 40-27 = 13 watts x 2 lbs of iron (roughly 4.5lbs total for a Q100)= 26 watts saved by using thinner laminations.
26 watts / 580 watts input = 4.5% improvement in efficiency.

I realize these calculations are very rough, but 4% is believable to me. I'll feel a bit better when I dyno the motors myself. Also note that the dyno results are at 200 RPM not 373 RPM where the iron losses would be much greater.

I agree to a point about the Q128 motors not having a suitable controller as of yet (maybe the S12S). At the very least, the extra iron and larger diameter of the Q128 should make significantly more torque. The S06S doesn't seem to lose sync up to it's 37mph no load speed for me, which seems plenty fast for a lightweight build. If you're planning on running 40mph plus, I imagine the batteries would put your bike into the heavyweight category so you might as well run a larger motor anyhow.

600hz%252520core%252520loss.jpg
 
I'm sure that 40w of iron loss at 201rpm turns into >100W of loss at high speeds.

Amazing to see how exponential these losses are. That does correlate with my bench tests of many motors.

Laminations make an incredible difference in power, don't they? One good example is the 1500W leaf motor ( 0.35mm ) versus the Crystalyte HS3548 ( 0.5mm ).

Crystalyte 35mm's Peak efficiency is about 84%..
Leafmotor 35mm's Peak efficiency is about 90.5%..

At 2000W continuous, the Leaf is about 86% efficient..
Crystalyte's motor would be 80% efficient at 2000W continuous..

Getting these geared motors into the 85-90% efficiency ranges is the next fronteir and i like what you are doing :)
 
Made a mistake here. 200 RPM is more like 300hz, so closer to 10 watts of power saved with thinner laminations at 200 rpm. Not sure where all the extra efficiency comes from in the dyno sheet. Still, I think people will be running these motors closer to 30mph/600 hz where the gains from thinner laminations will be substantial.

300hz%252520and%252520600hz%252520core%252520loss.jpg
 
580 watts for 30mph is maybe what you see when you're going downhill a bit.. that's more of the type of wattage figure you'd see for 25mph btw.

10w doesn't seem like much. I bet that the losses amplify with wattage applied? I'm no motor design expert though..
 
motomech said:
d8veh said:
mlt34 said:
Very cool! Are you working straight with Ananada or going through BMSbattery?
You mean Aikema, not Ananda?
Maybe we are still playing the "Don't Name the Factory"game.
The photo showing the group of motors, shows they carry the lazer etched script of Ananda and not the sticker used by Aikema.
When Aikema sell their motors direct, they mark them with a code that begins AKM. The BMSBattery ones are unmarked.
 
Sorry to be a complete noob here, but could someone link me to a reasonably simple paper on what effect changing the laminations have? I tried googling the term, but a lot of noise came back - such as the enamel lamination on the copper wire...

I've always been interested in squeezing more out of these mini motors.
 
What about going without a clutch? The more a motor is used continuously, the less it benefits from the clutch. No-clutch will put a cap on downhill speeds with regen, but that can be a good thing, too.

Its now been confirmed that the Currie iZip EzGo folder has no clutch on its 16" geared rear, and I must say I don't miss it.
 
neptronix said:
580 watts for 30mph is maybe what you see when you're going downhill a bit.. that's more of the type of wattage figure you'd see for 25mph btw.

10w doesn't seem like much. I bet that the losses amplify with wattage applied? I'm no motor design expert though..

My Q128H & Q100C experience kinda confirms this power observation. Although the 201 RPM Q128H in 700c runs out of speed just under 25 MPH. Which is good for hauling up hills with little loss in speed.

I run either on Lyen sensored controllers and never any timing issues. But, I know folks wanna avoid Hall sensors (heat cook-off can be a bitch) and desire cheaper controller sources…

Anyway, my coworker and I have experienced many clutch problems with the 201 RPM Q128H in 700C wheels, 67V off charger. Thicker, stronger springs seem to have solved it for the last 1-2000 miles or so and I must say the Q128 housing is much, much easier (6 screws) to R&R than the Q100 (fabricate a big wrench) should internal work be required.

I’ve ridden the Q128H with a stuck clutch on numerous occasions and it could easily be locked-on permanently if desired. I’m conflicted about that but it will regen, if you do. Not sure about extra wear on the gears, etc?

Will keep an eye on this potential USA source, thanks!
 
Yeah, that's about right for a geared hub that can handle the power.
on my 90.5% peak efficient DD, i see just around 500W for 25mph.

A clutchless geared motor is gonna have some pretty crazy drag with the power off ( just spin your geared motor backwards and you'll see what i mean, way worse than a DD ) .. I think that such a motor only makes sense if it's capable of 1000W continuous or more.
 
Back
Top