Very Impressed by fast motor winding with gear reduction vs. Slow motor Winding without gear reduction

Joined
May 19, 2012
Messages
611
H3525 vs. H3548 with 2:1 reduction:

Screenshot-27.png


MAC 12T vs. MAC 6T with 2:1 reduction:

Screenshot-26.png


Notice how the very impressive gain in efficiency at the lower and midrange speeds in both the Crystalyte direct drive motor and MAC geared hub motor after applying a 2:1 gear reduction to the fast "low torque per amp" winds.

As we know fast winds sacrifice torque per amp, but gain rpms per volt.....but looking at these results I think I would want the lower internal resistance (and thus less heat) of fast wind coupled to gear reduction than the higher internal resistance of a slow wind without gear reduction for any given speed range.

P.S. The 12T and 6T MAC Motors both have a 5:1 internal reduction.
 
Sure, but the benefits trade off against additional cost, weight, wear, maintenance, space onboard, noise, filth, pinch hazard, and points of failure when you add an extra reduction stage.

Whether it's worth it depends on application.
 
Chalo said:
Sure, but the benefits trade off against additional cost, weight, wear, maintenance, space onboard, noise, filth, pinch hazard, and points of failure when you add an extra reduction stage.
Or just go with a wheel that's half the size . . . .
 
With just about any electric motor pole switching speed plays a huge part of efficiency. It starts at 0 RPM where efficiency is zero.

BLDC motors can't pack enough poles to be efficient in a hubmotor envelope to be very efficient below 150 RPM or so, and even then aren't anywhere close to their peak efficiency. A switched reluctance similar to a stepper motor can but they suffer from low peak efficiency.

For example on the MAC, 36 poles and 5:1 gearing, the poles are switched 180 times per revolution ("Virtual Rotor Speed") . For a 26" tire that comes to 1 pole switch every .45" of travel. Compare that with a Direct Drive 48 pole (with larger OD & weight) of the same power rating, with a pole switch every 1.74" of travel. That shorter distance of pole switching increases efficiency and torque significantly.

Adding more turns per pole doesn't change the wattage required for the same level of torque, it merely takes more voltage and less amps.

As for wear & tear, I recently replaced the gear/clutch assembly in a MAC with over 5000 miles, 52V-35A limited, thats been run hard and put up wet. Gears & clutch were fine, easily had another 5000 miles left to go. Not sure the Grin version will last as long due to the constant reversing backlash, but it might.
 
JackFlorey said:
Chalo said:
Sure, but the benefits trade off against additional cost, weight, wear, maintenance, space onboard, noise, filth, pinch hazard, and points of failure when you add an extra reduction stage.
Or just go with a wheel that's half the size . . . .
Sure!
[youtube]3DCiATL-kbc[/youtube]
 
Gearing has advantages and turnoff. Each builder has to choose his own priorities.
 
JackFlorey said:
Chalo said:
29" wheel up front, 12" in the back. Much easier to get the power you need out of a small motor, no endos - and think about how many questions you'll get!

I seen one cat pull something similar, it was many years ago but I think he was on a mountain bike with 20" bmx rear and 26" front. However from chatting with him, he didnt know much about his setup as he had someone else build it all. Never seen the guy since, only other ebikes I see near my pad are a handful out for sunday rides on rare occassions. Further from home, ebikes everywhere but 15% diy built. Magic Pie cat is the obvious one, and a fwd cat.
 
Chalo said:
Sure, but the benefits trade off against additional cost, weight, wear, maintenance, space onboard, noise, filth, pinch hazard, and points of failure when you add an extra reduction stage.

Whether it's worth it depends on application.

Perhaps now with fat bikes......which have a greater O.L.D. (e.g. 135mm QR and 150mm thru axle for front and 170mm QR (177mm thru axle) + 190mm QR (197mm thru axle) for rear) we'll see some new hub offerings using that extra space for adding another stage of planetary reduction inside the hub.
 
ebike4healthandfitness said:
H3525 vs. H3548 with 2:1 reduction:

Screenshot-27.png


MAC 12T vs. MAC 6T with 2:1 reduction:

Screenshot-26.png


Notice how the very impressive gain in efficiency at the lower and midrange speeds in both the Crystalyte direct drive motor and MAC geared hub motor after applying a 2:1 gear reduction to the fast "low torque per amp" winds.

As we know fast winds sacrifice torque per amp, but gain rpms per volt.....but looking at these results I think I would want the lower internal resistance (and thus less heat) of fast wind coupled to gear reduction than the higher internal resistance of a slow wind without gear reduction for any given speed range.

P.S. The 12T and 6T MAC Motors both have a 5:1 internal reduction.

Glad somebody else noticed the MAC...I am the world's biggest MAC fan :lol: .

ANY fast wind motor does NOT give up torque per amp...the torque a motor can produce is a factor of the motor diameter, stator width, number of poles....and the number of turns has NOTHING to do with it. When I say "amp", I am talking about battery amperage. A lower turn count DOES require more phase amperage to produce the same torque as a higher turn count motor that has the exact same design...other than the number of turns on each pole.

A lower turn count means lower internal losses...that is the advantage in addition to a higher rpm. The disadvantage of a low turn count is you need a bigger/better controller to provide higher phase amperage if you want the same torque.

In general, the faster a motor turns the more power it can produce...hence the advantage of a geared hub motor BUT because the stator has an air gap between the outer edge and the housing, it WILL tend to run hotter than a Direct Drive hub motor so you have to be careful about over heating...just my general rule of thumb, if you want to go over 30 mph on level ground then a geared hub motor probably won't work for you because it will over heat. That basically correlates to a 10T being the fastest motor you can run without over heating if you want to run a 26" or larger diameter wheel. At least that is true for me and I weight about 200 lbs, my bike weighs about 70 lbs, and I ride in very moderately hilly areas.

For me, the big advantage of a MAC is the torque due to the 5:1 reduction or actually multiplication :wink: ...turns out the MAC will produce about twice the torque a large DD hub motor will produce (at low speeds) as long as they both have the same input amperage/voltage. It is about twice because a MAC is smaller in diameter and width than a 205/50H DD motor.
 
I have thought for years now, that it would be great if pedicabs and other heavy cargo bikes could mount a simple hub motor with the 5-1 planetary gearing, but the magnet size and weight of copper of a big clyte direct drive.

The way to go for now, John in CR showed us quite some time ago. Big power motor, fast wind, tiny wheel. I bucked the trend and went tiny wheel, low speed motor with only moderate power. But that was for a specific cargo bike I never intended to ride faster than 18 mph.
 
dogman dan said:
I have thought for years now, that it would be great if pedicabs and other heavy cargo bikes could mount a simple hub motor with the 5-1 planetary gearing, but the magnet size and weight of copper of a big clyte direct drive.

The way to go for now, John in CR showed us quite some time ago. Big power motor, fast wind, tiny wheel. I bucked the trend and went tiny wheel, low speed motor with only moderate power. But that was for a specific cargo bike I never intended to ride faster than 18 mph.

Dogman Dan...I agree.

Although I am always raving about the MAC, it does have some serious limitations due to the inability to reject heat. Steady state, I can't run more than about 1,000W or 19-20A @ 52v without over heating (experimentally verified :wink: ). Shorter term you can run about 2,600W for 5 minutes according to the GMAC page that has Justin's data:
https://ebikes.ca/product-info/grin-kits/gmac.html
I regularly run 60A battery/150A phase with my 12T MAC for roughly 5-10 seconds then the amperage drops because the motor is spinning a lot faster but that gets me about 200 NM of torque when starting and I have never had any problems with the clutch or the gears.

You can add liquid cooling to a MAC/GMAC and run more power but it complicates things too much in my opinion :D . You need to use about 5 ounces of cooling fluid and if you use ATF and a drop gets on your brake rotor, it can get exciting...I know :lol: .
 
The truth about hubs is that they have an optimal wheel size they should be operated at. AKA they have sweet spots in the RPM ranges where iron/cogging losses are at their lowest, and you see the highest efficiency. They also have a sweet spot where power output is at it's highest... which tends to happen in smaller wheel sizes on higher voltages.

Each of these points depends on stator design. The ebikes.ca simulator will help you find these two sweet spots per motor... :)
Pole count is a big determining factor.

But i'll give you some clues..

MACs seem to do best in 24 inch wheels and larger.
The larger of the shengyi geared motors probably like being in 20-24 inch wheels.
9C clone type motors do best in 18-24inch.
Motors with a more tall aspect ratio and higher pole count like the Magic pie are best above 24 inches.

.. if you want to take advantage of a smaller wheel size.. i can't recommend looking into 22 inch BMX rims enough for upright bike use.
 
neptronix said:
The truth about hubs is that they have an optimal wheel size they should be operated at. AKA they have sweet spots in the RPM ranges where iron/cogging losses are at their lowest, and you see the highest efficiency. They also have a sweet spot where power output is at it's highest... which tends to happen in smaller wheel sizes on higher voltages.

Each of these points depends on stator design. The ebikes.ca simulator will help you find these two sweet spots per motor... :)
Pole count is a big determining factor.

But i'll give you some clues..

MACs seem to do best in 24 inch wheels and larger.
The larger of the shengyi geared motors probably like being in 20-24 inch wheels.
9C clone type motors do best in 18-24inch.
Motors with a more tall aspect ratio and higher pole count like the Magic pie are best above 24 inches.

.. if you want to take advantage of a smaller wheel size.. i can't recommend looking into 22 inch BMX rims enough for upright bike use.

Very true...I can tell you a 12T MAC tends to over heat when you run a 29x2.6" tire/wheel with a 52v battery. I wanted a higher top speed so I went to the largest diameter tire that would fit in my frame but going to a lower turn count/faster rpm motor like a 10T MAC with a 26" wheel/tire would have been a better choice and would work a LOT better from both the efficiency and the acceleration standpoints :D .

The only drawback that comes to mind for a smaller diameter wheel/tire is it doesn't roll over obstacles as easily. My personal preference is a 26" or larger diameter wheel but nothing wrong with the smaller diameters if you like them and as Neptronix pointed out, the smaller diameters have some advantages :wink: .
 
Yes, the ideal motor for a 29 inch wheel or larger tends to be a geared motor or a super tall statored, high pole count motor such as the magic pie or the RH212 that grin is selling.

Some dual reduction geared hubs also work best in this big wheel size.
The ebikes.ca simulator is a great tool to guide you through the best motor configuration possible.

You know, both of my recumbent bikes have 20 inch wheels and they ride like a luxury car. A large wheel diameter does help with bump compliance, but suspension does a 100x better job and also has the benefit of keeping the wheel planted, rather than launched into the air when you hit a bump. Once you start going faster, design details from cars and motorcycles start making more sense.

And when you want to extract the most power out of a hub motor possible, the smaller the wheel and the more volts the better... an electric motor makes more power the faster it spins.. not unlike an internal combustion engine.
 
neptronix said:
Yes, the ideal motor for a 29 inch wheel or larger tends to be a geared motor or a super tall statored, high pole count motor such as the magic pie or the RH212 that grin is selling.

Some dual reduction geared hubs also work best in this big wheel size.
The ebikes.ca simulator is a great tool to guide you through the best motor configuration possible.

You know, both of my recumbent bikes have 20 inch wheels and they ride like a luxury car. A large wheel diameter does help with bump compliance, but suspension does a 100x better job and also has the benefit of keeping the wheel planted, rather than launched into the air when you hit a bump. Once you start going faster, design details from cars and motorcycles start making more sense.

And when you want to extract the most power out of a hub motor possible, the smaller the wheel and the more volts the better... an electric motor makes more power the faster it spins.. not unlike an internal combustion engine.

I agree...larger diameter wheels are a very minor effect as far as smoothing out bumps but a suspension makes a very large improvement :wink: .
 
Yes, I stopped riding my semi recumbent, but could have rode it more with even a primitive rear shock.

Re the mac or similar motors overheating. I did a lot of testing on motor overheating for E bikekit.com. Jason had no interest in selling more powerful kits that are typically legal in the USA. So our testing revolved around 26" wheel, and the loads it could handle. While 500w rated geared motors might get nice and hot, we rarely had to warranty replace a motor if the total load was under 300 pounds, and grades not over 8%. But at 400 pounds load, which I got with a bob trailer full of sandbags, I killed the geared motor in about 20 min. The 9 c type dd motor survived the 2000 foot vertical test hill to the top at 400 pounds load, but was very hot in the 7t. In 10t, it was not seriously hot at the top. All tests were done at 100 F plus ambient temps btw.

Anyway, you can melt motors real fast with over wattage, but IMO the real culprit for melting typical ebike hub motors in the 500w rated type, is fat people, and trailers. A high wattage geared motor would be real nice for some types of bikes. Stuff like the tadpole trike with a 200 pound freezer for ice cream on it. But instead, mid drives are now improving a lot.
 
dogman dan said:
Re the mac or similar motors overheating. I did a lot of testing on motor overheating for E bikekit.com. Jason had no interest in selling more powerful kits that are typically legal in the USA. So our testing revolved around 26" wheel, and the loads it could handle. While 500w rated geared motors might get nice and hot, we rarely had to warranty replace a motor if the total load was under 300 pounds, and grades not over 8%. But at 400 pounds load, which I got with a bob trailer full of sandbags, I killed the geared motor in about 20 min. The 9 c type dd motor survived the 2000 foot vertical test hill to the top at 400 pounds load, but was very hot in the 7t. In 10t, it was not seriously hot at the top. All tests were done at 100 F plus ambient temps btw.

Anyway, you can melt motors real fast with over wattage, but IMO the real culprit for melting typical ebike hub motors in the 500w rated type, is fat people, and trailers. A high wattage geared motor would be real nice for some types of bikes. Stuff like the tadpole trike with a 200 pound freezer for ice cream on it. But instead, mid drives are now improving a lot.

Like you said in the last sentence "mid drives are now improving a lot.", but one thing they can't do is Regen.

Unfortunately hub motors can't do Regen well when using a single battery due to limitations of charging imposed by the BMS.
 
ZeroEm said:
Unfortunately hub motors can't do Regen well when using a single battery due to limitations of charging imposed by the BMS.

Do you have link to this information? hub motors regen's poor performance is limited by BMS's.

I think I may be wrong as Regen current goes back into the battery via the discharge port rather than the charge port.

If so, that makes the capacity of the battery and the type of cells the limiting factor with a 1000 wh battery able to take twice the Regen current as a 500 wh battery assuming they both have the same cells and charge level.

So let's say I wanted a peak Regen of 2000 watts (for heavy weight on steep hill) that means I would need a 2000 wh battery if looking for 1C Regen charge rate.
 
ebike4healthandfitness said:
So let's say I wanted a peak Regen of 2000 watts (for heavy weight on steep hill) that means I would need a 2000 wh battery if looking for 1C Regen charge rate.

You could stick your foot in the spokes for braking, too. But it works better if you use your feet to push pedals, your motor to propel the bike forward, and brakes for braking.
 
I got a MAC mid drive geared motor.


Triketech said:
With just about any electric motor pole switching speed plays a huge part of efficiency. It starts at 0 RPM where efficiency is zero.

BLDC motors can't pack enough poles to be efficient in a hubmotor envelope to be very efficient below 150 RPM or so, and even

Thsi makes me wonder why the Mars motrs and teh Motenergy BLDC motors are only 8 poles. Lol. YOu would think they would lse resolution / RPM with the halls. I guess it does not matter in a 5KRPM motor. They are quite ecxpensibvly popular.
 
Chalo said:
ebike4healthandfitness said:
So let's say I wanted a peak Regen of 2000 watts (for heavy weight on steep hill) that means I would need a 2000 wh battery if looking for 1C Regen charge rate.

You could stick your foot in the spokes for braking, too. But it works better if you use your feet to push pedals, your motor to propel the bike forward, and brakes for braking.

But then I'd have to change the pads and rotors more often and I wouldn't get the Regen.

P.S. Electric cars have huge batteries for very long range and all use single speed plus Regen over multi speed transmission without Regen. Same goes for electric motorcycles like Zero and HD LiveWire.
 
ebike4healthandfitness said:
Chalo said:
ebike4healthandfitness said:
So let's say I wanted a peak Regen of 2000 watts (for heavy weight on steep hill) that means I would need a 2000 wh battery if looking for 1C Regen charge rate.

You could stick your foot in the spokes for braking, too. But it works better if you use your feet to push pedals, your motor to propel the bike forward, and brakes for braking.

But then I'd have to change the pads and rotors more often and I wouldn't get the Regen.

P.S. Electric cars have huge batteries for very long range and all use single speed plus Regen over multi speed transmission without Regen. Same goes for electric motorcycles like Zero and HD LiveWire.

You should borrow a bike with regen and descend a steep hill in order to see how it works in the real world. If your bike can climb the "steep" hill, it won't have any issues using regen during the descent. If your route starts on a hill, with the battery fully charged, then that's the only time you'll have an issue, due to overcharging, not overcurrent.
 
If your BMS is the issue, the wiring carrying the regen current can bypass it.

Some cells / chemistries are fine with higher than 2C charge rate

but they do need to be depleted enough to accept it.

LTO can go to 8+C no worries, full refill in 6min.

But they are low density, as I mentioned maybe just use big braking resistors for drag brakes.

However always have trad braking available for emergency stops

do not rely on regen except for drag braking on long descents, especially heavy loads
 
Back
Top