Hub mid-drive dual motor combos

If you used that thing as a primary motor at 52v, you'd burn it out fast.
Motors are rated primarily on their thermal properties as a single motor configuration.

I found out that motor (AKM 75) uses a 15 amp peak (7 amp continuous) KT controller that is available in 24V/36V and 36V/48V variants but the company is only offering it with the 24V/36V variant.This despite the 36V/48V version of the controller costing the same amount of money.

Apparently the reason for this is because the motor uses .5mm laminations.
 
.5mm laminations, lol.

That's why i'm thinking a modern dual reduction geared mtor with typical 0.35mm lams is just the ticket for this app..
..even poorer difference between power inputted vs ability to shed that power, but you're only going to be blasting the throttle on that thing intermittently.
 
.5mm laminations, lol.

Yeah that doesn't seem like a very good specification for a motor with a very high reduction at either 21:1 (201 rpm version) or 14.2: 1 (328 rpm version).

At 52v we could expect RPMs to increase 44.4% over 36v which would bump max RPMs to 290 for the 21:1 reduction and 474 RPMs for the 14.2:1 reduction version.

290 rpm at 21:1 gear reduction would mean the motor is turning 6090 rpm and 474 rpm at 14.2: reduction would mean the motor is spinning 6730 rpm.

That is a lot of RPM and I wonder how much gain with something like .2mm laminations.
 
Last edited:
For those worried about side covers breaking l am coming up on 600 miles of running exclusively on the 14t cog. To take some pressure off the side cover you don't need a 7 speed freewheel a 5 speed would work just as well which would take a lot of the leverage off the cover from using 7th gear . Next time I have my rear wheel off I will put the 5 speed on, and as a bonus that will give me room to mount a torque arm inside the frame.
 
Here is a thread where a person ordered a Cute Q100H (i.e. Aikema AKM 100H) motor customized with .35mm laminations:


Results shown below is the 201 rpm version of the Q100H which has 12.6:1 reduction ratio ( AKM-100H 36V350W EBike Rear Driving Hub Motor [AKM-100H rear motor] - $83.70 : Zen Cart!, The Art of E-commerce) at 38.49V:

1726312287459.png
 
Last edited:
Below is a comparison of the AKM 75 hub motor using either the 328 rpm version (i.e. 14.2:1 gear reduction) with 36V or the 201 rpm version (i.e. 21:1 gear reduction) with 52V:


1726363455523.png



And here is the result if I lower the amps on the 52V from 15 to 10 in order to equalize torque:


1726363548582.png


Obviously the motor power for flat ground cruising in the first and last scenario is not sustainable though. But maybe with .2mm laminations and perhaps some additional gear reduction the last scenario could be?

P.S. Ebikes.ca only tested the 328 rpm (i.e. 14.2:1 gear reduction) version of the AKM 75 so I had to simulate the 201 rpm (i.e. 21:1 gear reduction) version of the motor by converting from "hub motor" to "mid drive" and changing the chainring from 42T to 30T while adding an additional 1.48:1 reduction. (re: 14.2:1 x 1.48:1 = 21:1)
 
Last edited:
Also AliExpress is selling the 800 Watt version with a cassette.

You are right. According to this thread back in Oct 2020 they released a 800W version for the cassette:


Not sure how it got the power upgrade?

I do know back in 2015 the AKM 128H freewheel was rated at 800w despite the fact it had .50mm laminations. Could it be the cassette version of the AKM 128H finally got .35mm laminations like the Q100CST (i.e. the cassette version of the Q100) did back in 2015:


Currently, the Q100 CST shipping from BMSB already has 0.35 laminations. This is also why it's such a good performer even though the motor is tiny. If you want to exceed the Q100 CST's power levels, the easiest solutions I see is to offer a Q100H in 0.35mm or 0.2mm if possible as well as a Q128 CST in 0.35mm or 0.2mm.

Some posts discussing why the Q100CST hub motor's stator is narrower than the Q100H motor's stator thus making it a smaller motor. The last post also speculates that the .35mm laminations were used on Q100CST (i.e. the cassette version of the Q100) to offset the reduced width of the stator:


The problem with a cassette is that it will reduce the motor length by almost 10mm. This is because the ratcheting mechanism has been moved from inside the gear cluster to inside the freehub body.


I would actually love to see the CST version everything as freewheels are all but dead in the US and many other countries. Unfortunately, on these hub motors, the clutch mechanism required for a cassette is built into the hub motor cover, reducing the motor stack height by 10mm.


I would love to switch to cassettes, but the cassette ratchet mechanism takes up room in the casing, causing them to use a smaller (weaker motor).



I am thinking they switched to 0.35 lams in the CST to help offset the shorter motor stack.
 
Last edited:
So at 6 pounds is that the lightest for that much power.?

It is AFAIK.

So now you are probably asking why do not we see more big inrunner hub motor designs like this?

Well, I am not sure why we don't.

One guess I have is that whenever you have a powerful inrunner motor like the 800w AKM 128 you have an increased chance of throwing the magnets off the rotor. The because the 1.) faster the wheel speed 2.) the larger the rotor and 3.) the higher the gear reduction..... the greater the magnet speed and thus the greater the centrifugal force acting on the magnets attached to the rotor which can result in the magnets detaching from the rotor. Ebikes.ca even described this problem on the little Bafang G311/G310 motors --> G310-G311 - Grin Kits - Product Info

1726439961315.png

In fact, BMS battery to this day only sells the AKM 128 800w motors with the 201 rpm (reduction 10.5:1) option and not the 328 rpm option (reduction 8.1:1) they offer with the 500W version of the same motor. Though other sellers like Top Bike Kit do sell the 800w with the 328 rpm option. Not sure why this difference exists? Could it be that Top Bike kit AKM 800w uses a better adhesive?

P.S. I'm assuming another factor that might contribute to magnets flying off is increased magnet weight (i.e thicker magnets which would be used to increase magnet strength). As you see in the picture below Sur-ron (which makes a very powerful inrunner motor) actually uses screws (and probably adhesive too) to attach the magnets to their rotor:

1726441205538.png
 
Here's a thought if you're still on the fence; dual shengyi sx2's:

Motor Simulator - Tools


View attachment 358097

The two motors will come out to 12lbs and scale like a billy goat..
..buy one with regen to help on the way down, and the other one freewheeled.. :)

vs..

bbs02 @ 8lbs + 27mm 9C ( smallish DD ) @ 12lbs = 20lbs total

..the only problem with geared motors is that off road can give the gears a real beating, but if you have a full suspension bike with some air let out of the tires, they might survive.. & if one motor were to go, at least you have another :)

Maybe on a dyno those Shengyi's are efficient but what about when actually installed on a bike that actually makes turns and thus tests the lateral stiffness of the front wheel?

That Shengyi motor has a very poor spoke tension balance between drive side and non drive side and only 38mm distance between non drive side and non drive side spoke flange. In fact the spoke tension imbalance is so bad the best I can do is a 69:31 ratio by making the non drive side spokes "elbows out":

1726861706729.png
In contrast, a regular Shimano disc hub (Shimano Deore HB-M525A 100mm Front hub dimensions | Freespoke) would have 55.5mm between spoke flanges and a spoke tension imbalance of 57:43:


1726862199898.png
So while the Shengyi might be very energy efficient if only riding the bike in a straight line once a person needs to turn the reduced lateral stiffness will no doubt force them to bleed off more speed than a front wheel using a conventional disc brake hub.

In fact, I shudder at the idea of putting a hub like this on any bike with a more forward weight distribution. However, a downhill bike going down a hill would be in the same category though (re: going down a hill shifts more weight to the front and thus increases the size of the contact patch of the front tire. The greater the size of the contact patch the greater the traction and thus the greater the potential stress applied to front wheel during a turn. Think "Taco" when stresses get too high relative to the spoke bracing angle.)

This hub motor design needs to improve.
 
Last edited:
Like all other hub motors there's been basically no long term reported issues with the specifics of the wheel dish etc for any given motor.

This seems to be primarily your concern. My time is best spent not arguing about these specifics.
 
Last edited:
Where’d the 21:1 reduction figure derive from for the Aikema 75 motor?

The SX version uses a 21-33 nylon gear, so I can’t imagine any sensible sun and ring gears that could achieve 21:1.

A 6T sun gear doesn’t count as sensible, if that’s what it uses, but assuming it does, then 6 into 33 = 5.5:1, meaning another 4:1 reduction is required to achieve 21:1. I.e. ring gear would need to be 84T (4 x 21T). Would 84T even fit in that hub shell? They’d need to so fine pitched that they’d be prone to destruction under any kind of shock load.

I’m not saying it’s an impossible gear reduction, obviously, the numbers are plausible, but I have serious concerns about the robustness of any gearbox that implemented it in this way.
 
Having read many of the replies on this issue, most seem to be related to the choice of 1 motor or 2 when building a bike, and making choices based on the intended use. I for one have 2 motors, a geared hub and a BBSHD with 46t front and 14/28 rear. I will say I am perfectly happy. This bike was factory made with a single hub motor, so the choices I made were based purely on economics and what I could do myself with a conversion kit. I am 6'0 and 220 lbs, my bike from the facory weighed 78 lbs, and came with a 1000w hub drive. My primary interest in riding is 20-40 miles in the backwoods, on forestry service roads, which involve a lot of 4 plus miles of long steep grades in excess of 6%. It didn't take me long to burnout 2 30a controllers, which in retrospect the simulator predicted after about 2 miles.

After researching this forum on adding another motor, I found a lot of good information (Some good- some bad) and I decided to plunge in and install the BBSHD myself. Having an electrical engineering background helped a lot, though most of what I did previously didn't involve electric motors, but I had the basic concepts. The BBS installation turned out to be a piece of cake, including some cusomized electrical configurations, like kill switches on both motors, left brake kill bbs, right brake kill hub, BBS PAS and Throttle only only with manpower pedal on the hub. My Handlebar and extra bar look like a motorcycle dash with 2 lcd's, phone/nav app, camera, kill swithes, throttles, etc.

So How do I mitigate the potential torque problems many writers claim to exist with the hub casing? The simple answer is don't over torque it. I follow a couple of basic rules: I never use the BBS alone! I aways start up from a stop or very low speed with the hub throttle while in a middle or low gear; I always ride with the hub motor unless I want to go faster than 20 mph and my watts get upward of 700w or so, then I kick in the BBS PAS with the appropriate gear and power level to split the load as evenly as I can. Essentially that is how I also ride the mountains; if I am on a long steep grade I am operating on a 46t/26t BBS PAS, and control throttle on the hub to split the load as much as possible, rarely going over 700-800w on either motor. Checking motor and controller temps after a couple of climbs of 5-6% running for a long period of time the temps are always less that 35-40C barely warm to the touch. I have a 52v 20ah battery running the BBS and a 48V 21ah bat on the hub, I also have a 3rd battery I carry on times that I expect to go on extra long excursions This third bat is fed to a blender with the other 48v battery to the hub which gets most of the use. All batteries have on/off switches and common connectors so I can reconfigure the batteries to the motors if needed. I can ride the mountains for much longer than my body can handle.

So, in conclusion, I will say that this has been a very successful conversion. I have a high level of redunancy, which is very important to me riding alone in the backwoods, and I am able to tackle any and all hills I have attempted in recent months. I do have a 36t front chainwheel, but so far I haven't found any reason to put it on, nor do I see a reason to add more bigger cogs to the rear. I didn't mention that this bike is a fat tire, and I plan to convert an exiting mountain/hybrid bike this winter. This convertion is aluminum frame with 42mm tires. I will probably keep it simple and use the BBSHD and no hub motor. Field testing will determine where I from there.
 
So How do I mitigate the potential torque problems many writers claim to exist with the hub casing?

Easy. Put your second motor in the other wheel. Not jamming superhuman forces through the rear hub side cover.
 
Like all other hub motors there's been basically no long term reported issues with the specifics of the wheel dish etc for any given motor.
1.) That's because there are not many people in this forum using that motor. (Example: This is all I get from a forum search of the terms Shengyi SX1 --> Search results for query: Shengyi SX1 )

2.) In the event we actually get people using the SX1 how much experience do they have comparing the SX1 as a wheel to other hubs both hub motor and analog? This assuming the rim diameter and spokes are the same. And what bike are they riding? At what speeds are they riding at? That is going to be very important for objectively determining how much cornering performance loss is actually occuring. Even better would be a controlled test though.

2.) Even Justin admits in situations where the spoke is almost vertical it has very little ability to provide triangulation strength to prevent the rim from buckling:


Now granted Justin flipped the spoke in the video from "elbows in" to "elbows out" but the resulting spoke bracing angle is still relatively vertical with 700c rim and the Shengyi SX1 non drive side bracing angle with a 700C rim is even worse than the Clyte H-UFO direct drive hub example used by Justin in the video. This because while the non drive side distance of the Clyte spoke flange is 9mm from centerline compared to 10mm non drive side distance from Centerline on the Shengyi SX1 the DD hub benefits from a larger PCD (232mm vs. 141mm) which helps spoke bracing angle. This (9mm distance to spoke flange + 232mm PCD) results in the DD hub motor having a 2.7 degree bracing angle compared to the Shengyi SX1's 2.4 degree bracing angle which comes as a result of 10mm distance from centerline to spoke flange and 141mm PCD.

Here is the "hybrid hub" (i.e non drive side Justin's DD hub motor and non drive side Shengyi SX1) I made up in freespoke to show the difference in spoke bracing angles for a 232mm PCD with a spoke flange distance of 9mm from centerline and for a 141mm PCD with 10mm spoke flange distance from centerline:

1726877745394.png

I came up with the 9mm distance from centerline for the Clyte H-UFO (used by Justin as example) because it has a 4mm dishing offset coupled to a 26mm distance between spoke flanges. 4mm offset means the non drive side distance is 9mm from centerline and the drive side distance is 17mm from centerline (i.e non drive side is 4mm shorter and drive side is 4mm longer than what would be if the hub were non dished which would be 13mm non drive side and 13mm drive distance from centerline). Likewise the shengyi SX1 with its 10mm distance from centerline is the result of having a 9mm dishing offset coupled to a 38mm distance between spoke flanges (i.e non drive side is 9mm shorter and the drive side is 9mm longer from centerline than what it would be if the wheel were not dished)

3.) Justin admits that a spoke tension balance of 70:30 results in a more likelihood of spoke breakage:


....and we are at a nearly as bad ratio of 69:31 with the Shengyi SX1 even under the best conditions (which is with non drive side elbows out)--> Hub mid-drive dual motor combos

4.) Not all front hub motors have such poor spoke bracing angles like the two examples (Clyte H-UFO and the Sjengyi SX1) given. Take for example the Aikema AKM75 you used as an example earlier in this thread:

1726954313971.png

1726955784049.png

5.3 degree bracing angle on the non drive side is way better than the 2.7 degree bracing angle for the Clyte H-UFO non drive side and the 2.4 degree bracing angle on the non drive side for the Shengyi SX1.

Also notice how the spoke tension balance on the Aikema AKM 75 is massively better than the Shengyi SX1 which has the worst spoke tension imbalance of any front hub in the ebikes.ca spoke calculator--> Online Spoke Calculator for Hubmotors and Ebikes

My prediction is that it in a controlled test you would get way better cornering performance with such a hub.
 
Last edited:
Having read many of the replies on this issue, most seem to be related to the choice of 1 motor or 2 when building a bike, and making choices based on the intended use. I for one have 2 motors, a geared hub and a BBSHD with 46t front and 14/28 rear. I will say I am perfectly happy. This bike was factory made with a single hub motor, so the choices I made were based purely on economics and what I could do myself with a conversion kit. I am 6'0 and 220 lbs, my bike from the facory weighed 78 lbs, and came with a 1000w hub drive. My primary interest in riding is 20-40 miles in the backwoods, on forestry service roads, which involve a lot of 4 plus miles of long steep grades in excess of 6%. It didn't take me long to burnout 2 30a controllers, which in retrospect the simulator predicted after about 2 miles.

After researching this forum on adding another motor, I found a lot of good information (Some good- some bad) and I decided to plunge in and install the BBSHD myself. Having an electrical engineering background helped a lot, though most of what I did previously didn't involve electric motors, but I had the basic concepts. The BBS installation turned out to be a piece of cake, including some cusomized electrical configurations, like kill switches on both motors, left brake kill bbs, right brake kill hub, BBS PAS and Throttle only only with manpower pedal on the hub. My Handlebar and extra bar look like a motorcycle dash with 2 lcd's, phone/nav app, camera, kill swithes, throttles, etc.

So How do I mitigate the potential torque problems many writers claim to exist with the hub casing? The simple answer is don't over torque it. I follow a couple of basic rules: I never use the BBS alone! I aways start up from a stop or very low speed with the hub throttle while in a middle or low gear; I always ride with the hub motor unless I want to go faster than 20 mph and my watts get upward of 700w or so, then I kick in the BBS PAS with the appropriate gear and power level to split the load as evenly as I can. Essentially that is how I also ride the mountains; if I am on a long steep grade I am operating on a 46t/26t BBS PAS, and control throttle on the hub to split the load as much as possible, rarely going over 700-800w on either motor. Checking motor and controller temps after a couple of climbs of 5-6% running for a long period of time the temps are always less that 35-40C barely warm to the touch. I have a 52v 20ah battery running the BBS and a 48V 21ah bat on the hub, I also have a 3rd battery I carry on times that I expect to go on extra long excursions This third bat is fed to a blender with the other 48v battery to the hub which gets most of the use. All batteries have on/off switches and common connectors so I can reconfigure the batteries to the motors if needed. I can ride the mountains for much longer than my body can handle.

So, in conclusion, I will say that this has been a very successful conversion. I have a high level of redunancy, which is very important to me riding alone in the backwoods, and I am able to tackle any and all hills I have attempted in recent months. I do have a 36t front chainwheel, but so far I haven't found any reason to put it on, nor do I see a reason to add more bigger cogs to the rear. I didn't mention that this bike is a fat tire, and I plan to convert an exiting mountain/hybrid bike this winter. This convertion is aluminum frame with 42mm tires. I will probably keep it simple and use the BBSHD and no hub motor. Field testing will determine where I from there.
You could change the 7 speed freewheel for a 5 speed and take some leverage off the side cover. I plan on doing this next time I have my wheel off just as a little extra insurance on not stressing things more than nessisary.
 
Where’d the 21:1 reduction figure derive from for the Aikema 75 motor?

It comes from Top Bike Kit:



1726957519608.png
 
Last edited:
:)

21:1 for the 328 RPM
14.2:1 for the 201 RPM

Something’s clearly amiss right there.

Just for fun … 328 X 21 = 6,888 RPM

Around 8000 RPM unloaded.

AFAIK, the 14.2 reduction for the 201 version is correct. 21:1 is obviously a mistake.
 
:)

21:1 for the 328 RPM
14.2:1 for the 201 RPM

Something’s clearly amiss right there.

Just for fun … 328 X 21 = 6,888 RPM

Around 8000 RPM unloaded.

AFAIK, the 14.2 reduction for the 201 version is correct. 21:1 is obviously a mistake.

I agree it's a typo, but if this 21:1 reduction ratio actually exists it would be for the "lower speed, high torque, suitable for hill" application which is the 201 rpm version. Meanwhile the 14.2:1 reduction ratio would be for the "lower torque, higher speed, suitable for smooth road" application which is the 328 rpm version.

But honestly I am having a hard time believing this 21:1 gear reduction actually exists wnen Top Bike kit themselves lists both versions of the AKM-75SX using the same 33-21 planet gears:


1727048007139.png
 
I think I know what happened now ..

According to an email BMS battery sent me today the Cute Q75 (which is an Aikema AKM-75SX) is 12:1 for the 328 rpm version and 14:1 for the 201 rpm version.

So if Top Bike kits motors are the same as BMS battery someone at Top Bike kits got "12" reversed to "21" and thus 21:1 was inputted on the webpage listing rather than 12:1.
 
Thank you, everyone, for the discussion about these motors. There wasn't much more information available on the web. I would like to get your feedback on whether the AKM 75SX low RPM version is suitable for someone who wants to ride without motor power most of the time without feeling resistance, but then use it at full power to get back home.

Also has anybody use it with 24v? I have a great 8s bms with no use at the moment
 
Back
Top