Bafang/8fun SYXD-01F internal photos, gear modifications

(I had incorrectly posted this in the wrong forum, I'm moving it over here. Sorry if you've seen it twice)

This is the hub used by the Faraday. I'm using one on my bike and selected it over other options due to the light weight (2kg vs 2.6kg or more for other Bafang alternatives) and the good disk mount (Cute has theirs 2mm out of spec, which won't work with my fork).

I cracked my hub open today to see how the internals compare to other Bafang/8Fun hubs and was surprised to see a double reduction planetary system. The clutch is also in a different place, it is mounted between the ring gear and the outside of the hub shell. This also means that the torque is transferred from the ring gear to the hub flanges through 6 M5 bolts (the ring gear is in the hub's "cap"). That is probably fine since 6 bolts of the same size also transfer all of the disk stopping power.

This photo is of the hub cap (where the disk is mounted) and ring gear:


This is the double reduction system. Note that the motor and larger gears are helical cut:


I'm curious if there might be higher RPM motors or gear sets available for this hub to increase the hub's maximum speed in a 700C wheel from 22mph (20mph realistically) to something higher. The native hub RPM is 250 (measured), 270rpm (spec) at 42 volts.

It is good to see that the hub has handled ~1000 miles, lots of it in the rain and mounted on bus racks, without any water intrusion. That was one of my concerns.

It would be remarkable if the Cute gears interchanged.
You might start by counting the # of teeth on the sun gear(out-put shaft).
The most recent Q100's have used both 9 T. and 11 T., while the older models used a much larger shaft, something like 15 T or 18 T. I have a couple sets of gears I could measure if there is an interest.
What you have now is a mid-speed motor, much like the Cute 260, the MXUS 260 and the Outrider standard and I can tell you from experience that that is the motor to use for 95% of riders(w/ 26" or 700 whl.)
If you want more speed, far and away, the way to prodceed is to up the Volts to 48 V for a top speed of 24 to 25 mph(w/ 700 whl.) Running a high-speed mini motor in a big wheel is only for the extreemly fit Velcro types.

I am a fairly fit cyclist, so the motor that I have only makes the bike about 5-10% faster than my other bikes for commuting in practice. It is close enough that I give myself the same allowance for ride times. That is why I'm interested in faster options.

Sadly my headlight (IQ Cyo Premium E) handles a max of 42V, so moving to a 48V battery is more expensive than considering hub modifications since I'd need a new battery and a new headlight or voltage converter. An option may be to get a 4S2P battery pack that is wired in serial with my existing pack to boost it to 48V, but allowing me to pull 36V off of the base pack for my lights.

I couldn't get the planet carrier off so I wasn't able to count the teeth on the sun gear. I didn't try too hard, but it appears to have been press fit to some level onto the axle. Do the Cute gears also have a helical cut (it looks like they don't from photos)? It would be really interesting if they did interchange, it did strike me that the internals of this hub are similar to the Cute Q100. What is the diameter and how many teeth are on the output side of the Cute gears?

I'm considering making a new fork for the bike that would have a 2mm out of spec disk mount to match the 2mm out of spec mount on the Q100 hub.

If that motor is constructed like the Q100, the part you refered to as the planet carrier is also a half-shaft. It comes off after removing the three machine screws. They are best removed using a hand held impact driver, otherwise, the slots are almost garrentied to get buggered.
Yes the larger gear is helical with the Cutes. You can count the teeth on the replacement gear-set pictures. I did once and determined they are for the 328 standard(not H) motor. Since they are the only replacement gears available, it's irelevent what the other gear-sets are. They would be your only gear-set available without buying a complete motor.
That's the second time you have mentioned "Disc mount spec" being off and frankly, I don't know what you are talking about.
I've mounted three different Q100's on two different bikes and the disc always is centered well enough to the caliper(if that is what you are refering to)that the adjustable carrier could deal with it. Nobody else has ever had a problem in the dozens of Q100 install accounts here at the Sphere.
I have run into that problem with other motors and 2 m/m isn't anything that a strategicly placed thin flat washer won't solve. One on the axle inside the drop-out will move the disc inward 2 m/m, or two sm. ones between the caliper and the carrier will move the carrier inwards. Final adjustment is done with a flat file.
It seems to me, that if you want a high-speed mini, you need to go with the Q100 328

I removed the three screws and the plate didn't come off easily. I was running short on time and didn't explore much further. They did require an impact driver to remove.

The drawings that I've seen for the Q100 show the disk at 8mm from the locknut on the axle instead of the standard 10mm. My fork (it's my fault, I built the fork, dropout, and disk mount) has the disk mount 4mm too close to the dropout. There is no way to use spacers to get things farther apart from each other. 4mm too close (from my disk mount) just barely works with a BB7, but 6mm too close (2mm from the hub, 4mm from the disk mount) will not. Someday I'll make a new fork.

I just checked the Q100 drawings and either I mis-read it in the past or they've fixed it, now they show 10.5mm (0.5mm extra is fine for almost all brake calipers).

How does the torque on the Q100 compare to the Bafang motors? There is a new higher torque Q100H, but it doesn't come in 328rpm. I wonder if it is the same motor, just geared down to 260rpm (328/260 = 1.26, which is pretty close to their 30% increase in torque). I'm okay losing 30% torque from my Bafang motor, but I wouldn't want to lose 50% of it.

If I get a chance soon I'll pop the cover off again and count teeth. On the Q100 I assume that both the planet gears and the sun gear are changed to go between high RPM and low RPM hubs? BMSBattery has the Q100 planet gears, but not the sun gear.

I realized that I was confusing the Q100 disk rotor placement with the rotor placement on the similar Outrider motor that Grin sells:
http://www.ebikes.ca/documents/MOutriderF.pdf

That one has the disk rotor mount at 8mm from the locknut, which is 2mm out of spec. The Q100 looks fine.

Sadly it looks like you can get the Q100 with high spec motor or with 328rpm, but not both . I've also seen a lot of photos of Q100 motors with internal rust, where the 8Fun SYX is pretty well sealed.

Well, it has to be a half-shaft, remember, the sun gear is spinning the actual motor rpm underneith. I suspect it's tight on the dowels and a sharp rap with a hammer will loosen it.

How does the torque on the Q100 compare to the Bafang motors?

Click to expand...

As this is the first 2Kg. mini from Bafang, which nobody here has used, it's hard to say.

There is a new higher torque Q100H, but it doesn't come in 328rpm. I wonder if it is the same motor, just geared down to 260rpm (328/260 = 1.26, which is pretty close to their 30% increase in torque)

Click to expand...

Calculating "torque" of an electric motor, or even defining it, Is not a simple thing and assigning tq values as a function of a linear rpm scale is not the reality. By your reasoning, a Q100H 201 would have 30% more "torque" than a Q100H 260, which it certainly does not.
I ride with a 201 standard and a 201H every day(on the same bikeand I would judge the H to be 10% stronger on identical batteries and controllers. Since the laminations appear to be the same, I can only assume the H uses higher quality magnets. The 328 CST, with it's narrower stator is probably down additional 10%. It is the weakest of the Cutes.
Notice that I do not use the term "wind" when describing motor speeds, and here is the reason. A while back I counted the gear-sets and calculated the mechanical ratios of 4 variations of Cutes and found an exact correlation between the gear ratios and the no-load motor speeds. This would lead one to believe that the motor winds are all the same and the differences are solely a function of mechanical leverage. Yet, some how, I don't believe this is wholely accurate and it remains a "mystery of the East".
So it's hard to quantify the performance between versions, but I have used enough Q100 standard 328's to exatrapolate how a Q100 frt. in a 700 cc w/ a fit rider on board.
But first, some generalizations;
pretty much all hub motor's no-load speed will be determined by the motor's max. rpm.
The exception being high-speed mini's. They become current limited, pulling every Amp the controller willprovide and still wanting more.
This results in two distinque power curve forms past the no-load point.
The rpm limited motor falls away fast and after a couple mph, has dropped far enough the the rider can't provide any additional speed.
The current limited motor on the other hand, extends as a plateau and the rider can add several mph's.
This can be seen on the sim. @ Ebike CA.(It's more obvious at higher Voltages)
Looks like a 328 in a 700cc whl, with 36V bejind it will reach 25 mph.
But the slightest incline or headwind will quickly start pulling it down, so any rider input at top speed will be mostly maintaining said top speed.
But under more favorble conditions, I would think the rider could add 3mph's and with some help in the form of a tail wind or decline 28-29 mph could be reached.
The real downside to this set-up is hills. If you can't pedal up it on a road bike, don't expect much help. If the climbing speed falls to half the no-load speed, it will be time to get off and push.
Effecicency will be down somewhat, but will be off set by and large by rider input.
Resist the temptation to add Amps, 17 should be about the limit.
I have found that using 328's in big wheels is akin, to use a gas motor analogy, is like " lugging" the engine, which in turn, heats up the controller.
I use a larger 9-FET controller from ELB and while not addressing the problem directly, the additional internal airspace allows them to run much cooler than the typical cig. pack sized 6-FET

motomech said:

The exception being high-speed mini's. They become current limited, pulling every Amp the controller will provide and still wanting more.
This results in two distinque power curve forms past the no-load point.
The rpm limited motor falls away fast and after a couple mph, has dropped far enough the the rider can't provide any additional speed.
The current limited motor on the other hand, extends as a plateau and the rider can add several mph's.
This can be seen on the sim. @ Ebike CA.(It's more obvious at higher Voltages)
Looks like a 328 in a 700cc whl, with 36V bejind it will reach 25 mph.
But the slightest incline or headwind will quickly start pulling it down, so any rider input at top speed will be mostly maintaining said top speed.
But under more favorble conditions, I would think the rider could add 3mph's and with some help in the form of a tail wind or decline 28-29 mph could be reached.
The real downside to this set-up is hills. If you can't pedal up it on a road bike, don't expect much help. If the climbing speed falls to half the no-load speed, it will be time to get off and push.

Click to expand...

The ebikes simulator hasn't been very useful to me because it doesn't allow you to account for any human input. I'm mostly riding my ebike with the motor providing a 60% wattage boost based on my measured power from the torque sensing BB. There is no throttle. It is more like the motor provides a 3-5mph increase in speed than the human does that. I've never ridden my bike with a throttle and don't know what the motor is capable of on it's own. It would be nice if the simulator allowed you to specify input power from the human, just like you can adjust the throttle. It would also be nice to put in lower vehicle weights.

328 in 700C going 25mph makes sense. The circumference of a 38-622 tire is 218cm, or 0.001355 miles. At 328rpm that is 26.7 mph.

I'm using the 6-FET 15A Infineon Controller that Grin sells and have never felt it get warm, even after climbing at 7-10% grade (variable) for a mile at 16-18mph (so 3.5 minutes). That climb is a normal part of my riding routine. That is about the only time that I ever see the controller consistently put out 500 watts. The simulator would have the speed at 6.4mph for comparison (using an Outrider mini hub since they don't list the 8Fun mini).

I don't understand this "The rpm limited motor falls away fast and after a couple mph, has dropped far enough the the rider can't provide any additional speed."

Why can't the rider just outride the motor's maximum speed, causing the motor to freewheel? That is what is happening on my bike pretty much anytime I'm going downhill (even on a slight grade). I see the CAv3 reporting 15W consumption when I'm riding over 21mph (hub RPM is over 250 then, which is it's max), which is what I assume it takes to disengage the clutch and allow the motor to freewheel. I'm providing all of the power in that case.

alex