New Bafang Crank-Drives

[Warren]

nioko,

"It's not about 45kph for the whole commute...I currently do reach the 31kph where the bionX stops supporting me"

Sounds like you won't be disappointed. The big difference with a motor running through the gears, and not electronically limited to 20 mph, is that your average speed will be higher. My system is functionally very similar to the Bafang. I'm an old guy and I can average 20 mph for many miles @ 11 Wh/mile . My best average on my hilly, 7.2 mile loop, from my front door, was 25.2 mph pedaling at maximum effort @ 12.3 Wh/mile. I don't think I saw over 650 watts on the CA on that ride. I was probably adding from 150 to 250 watts throughout the ride.

To see a higher wattage reading I would be pedaling at a ridiculously low and inefficient rpm. Now, on an upright bike, where you can stand to climb, you could do it standing, at a much lower rpm, but the motor would be at a very inefficient speed. Most of that high wattage would be going to overheating your motor.

 

[Warren]

On second thought. My system is "36" volt, 25 amp, where the 750 watt Bafang is a "48" volt, 25 amp system, and somebody said it spins at about 110 rpm under load, so you might be able to pull 900 watts at an 80-90 rpm cadence.

 

[nioko]

Warren, thanks for your input.
what motor are you using?

nobody here with a BBS and similar usage profile as I have?

 

[Kepler]

I've received today a BBS01, 250w, 36V. Bought as an experimental 'legal' alternative to my other 'illegal' project. Now I'm thinking hard about which bike to put it on (first!)

However I have an 'engineering' question. I note that although the freewheel on the crank clearly works, there seems to be more 'drag' involved in the freewheel action than I've encountered with a conventional rear screw-on sprocket freewheel or with a cassette-type freehub. If I hold the unit horizontally with a crank arm on the shaft and give it a decent spin in the 'freewheel' direction (ie; backwards), it stops within one revolution.

I presume that the BBS01 crank 'freewheel' is more akin to what we might call a Sprag one-way bearing. Is this correct? And if so, is the apparently greater drag normal for this type of freewheel?

Thanks,

Savvas.

The freewheel on the cranks is a conventional ratchet type. Very smooth and so far seems quite strong.



The freewheel the motor gear reduction system is a sprag bearing which is located in the white intermediate gear. This means the white intermediate gear and the gear attached to the motor's rotor do not spin when pedal without power. Again a nice smooth system with minimal drag and no reports of failure or lockup as yet.

 

[jk1]

The 750W unit on PAS only pulls about 550W. Subtract 100W to 150W for parasitic drag

Say at 150w/750w means the total efficiency of the motor and gear reduction to the cranks must be 80% in total if you get only 150w losses. i.e total 20% losses. Did you measure 150w losses ? its probably a lot more then that becuase I was reading in another thread here: http://endless-sphere.com/forums/viewtopic.php?f=28&t=56957
That the gearbox design by itself has 33% losses as its 3 stage helical, so how can total losses be only 20%? what about motor losses ?


"A typical single-stage spur gearhead, for example, is about 90% efficient whereas a two-stage model is about 85% efficient. Most planetary gearheads are roughly 97% and 94% efficient for one and two-stage units, respectively. "
http://machinedesign.com/archive/spur-v ... ervomotors

If each stage of spur gear head is 90% efficient for strait cut,then helical will be at most 89%.

So for example comparing to another popular mid drive the Bafang BBs02 has a 3 stage helical NON planetary reduction, each stage is 89% efficient STg 1 2 and 3,11% less per stage then the total efficiency multiplied = 33% total loses.

Now compare this to the cyclone 2 stage planetary, stage one is 97% 9.331 planetary strait cut and stage 2 is a chain reduction of 97% efficiency. so 3% *3% = 9% total efficiency loss. so that's a difference of over 24% ! "

 

[Warren]

nioko,

"what motor are you using?"

http://www.evalbum.com/4525

 

[zzmatias]

Has anyone tried the BBS02 with a set of 22.5V Zippy batteries? I've heard the built-in controller only gives you the option of 36V and 48V, but I was wondering what were the low/high voltage cut limits and if they were configurable.

 

[Kepler]

The 750W unit on PAS only pulls about 550W. Subtract 100W to 150W for parasitic drag

Say at 150w/750w means the total efficiency of the motor and gear reduction to the cranks must be 80% in total if you get only 150w losses. i.e total 20% losses. Did you measure 150w losses ? its probably a lot more then that becuase I was reading in another thread here: http://endless-sphere.com/forums/viewtopic.php?f=28&t=56957
That the gearbox design by itself has 33% losses as its 3 stage helical, so how can total losses be only 20%? what about motor losses ?


"A typical single-stage spur gearhead, for example, is about 90% efficient whereas a two-stage model is about 85% efficient. Most planetary gearheads are roughly 97% and 94% efficient for one and two-stage units, respectively. "
http://machinedesign.com/archive/spur-v ... ervomotors

If each stage of spur gear head is 90% efficient for strait cut,then helical will be at most 89%.

So for example comparing to another popular mid drive the Bafang BBs02 has a 3 stage helical NON planetary reduction, each stage is 89% efficient STg 1 2 and 3,11% less per stage then the total efficiency multiplied = 33% total loses.

Now compare this to the cyclone 2 stage planetary, stage one is 97% 9.331 planetary strait cut and stage 2 is a chain reduction of 97% efficiency. so 3% *3% = 9% total efficiency loss. so that's a difference of over 24% ! "

BBS02 doesn't have a 3 stage gear reduction, it has a 2 stage gear reduction.

My real world tests showed the motor needed 120W to turn the output shaft with cranks and pedals (no chain) at 100% throttle.

It would be interesting to see what a Cyclone setup would draw under the same test conditions.

 

[jk1]

Kepler

it has a 2 stage gear reduction.

ahh ok if it's a 2 stage helical then that would be 22% losses just for the geartrain . So 150w losses still sounds too low because thats only 20% and no factor for motor/ controller losses ?

So you got 120w full throttle no load test with no chain ? At 750w you are at full load so the losses would be a lot more then at no load hence a lot more then 120w.

 

[Kepler]

I really don't think the drive train losses are worse than any good drive setup. My Bafang geared hubs pull about the same no load Watts and give me similar Whr/km under similar conditions.

 

[Avigo]

Also factor in the quietness of helical gear, l like my middy more than my hub drive because it is very stealth, you can not hear the motor. Viewing from the stern it looks like a normal bike.

 

[ErikDK]

The freewheel the motor gear reduction system is a sprag bearing which is located in the white intermediate gear. This means the white intermediate gear and the gear attached to the motor's rotor do not spin when pedal without power. Again a nice smooth system with minimal drag and no reports of failure or lockup as yet.

Is the large gear rotating during pedal without power and thus spinning the small gear (pinion)?

 

[crossbreak]

ahh ok if it's a 2 stage helical then that would be 22% losses just for the geartrain

You can not sum up loss in percentage points. Math is a bit more complicated here. You have to multiply. For example if one stage has 11% loss, you calc 0.89 x 0.89 which leads to 0.7911, which means a total loss of 20 .79% in such a drivetrain

Still, where did you get the "11% loss per stage" figure? that is nonsense.

Additionally, the second stage is not helical as you can clearly see in the picture above. Helical gear reductions get around 95%, straight ones 97%... so this drive should get around 92.15% efficiency.

As a matter of fact, most of the loss is induced by the motor. It may get 85%, so: 78.3% of the energy gets to the chain wheel. Another 4% are gone after passing the chain... so 75,1% get to the rear sprocket. From here, another 7% may be lost in the SRAM dual drive if you have one. Then there are still 72.2% of your initial energy% left.

the cyclone 2 stage planetary, stage one is 97%

I highly doubt that. Maybe if you put through 150 or 200W. At 500W+ a lot will be lost since gears+bearing are overloaded. Just have a look at how tiny this gearbox is, then have a look into the datasheets of planetary gearbox suppliers. http://www.harmonicdrive.de/cms/upload/pdf/GK_2011_2012/de_en/planetengetriebe/hpg-126-143.pdf
As you can see, they claim 90% efficiency for a 2-stage drive of the same size as the cyclone and 11:1 reduction, at only 45Nm at the output. With a single stage, this figure wont be better at this torque level, since the gears+bearings are over stressed.
Companies like Harmonic drive cannot afford to lie and loose their reputation. Most of their customers test their products by efficiency and reliability means. Cyclone can tell as much BS as they want, since there are plenty of customers out there that believe their "truth".

It would be interesting to see what a Cyclone setup would draw under the same test conditions.

Really, it would be. Sadly no one will ever try both on the same dyno. We can talk a lot, only such a test would tell us the truth.

 

[Kepler]

The freewheel the motor gear reduction system is a sprag bearing which is located in the white intermediate gear. This means the white intermediate gear and the gear attached to the motor's rotor do not spin when pedal without power. Again a nice smooth system with minimal drag and no reports of failure or lockup as yet.

Is the large gear rotating during pedal without power and thus spinning the small gear (pinion)?

Yes it is. The shaft on the small gear then goes inside the sprag bearing located in the white Nylon intermediate gear.

 

[GUI13]

for information ...

controller


and new connectors

 

[Kepler]

Nice one. Thanks for posting

 

[Kepler]

ahh ok if it's a 2 stage helical then that would be 22% losses just for the geartrain

You can not sum up loss in percentage points. Math is a bit more complicated here. You have to multiply. For example if one stage has 11% loss, you calc 0.89 x 0.89 which leads to 0.7911, which means a total loss of 20 .79% in such a drivetrain

Still, where did you get the "11% loss per stage" figure? that is nonsense.

Additionally, the second stage is not helical as you can clearly see in the picture above. Helical gear reductions get around 95%, straight ones 97%... so this drive should get around 92.15% efficiency.

As a matter of fact, most of the loss is induced by the motor. It may get 85%, so: 78.3% of the energy gets to the chain wheel. Another 4% are gone after passing the chain... so 75,1% get to the rear sprocket. From here, another 7% may be lost in the SRAM dual drive if you have one. Then there are still 72.2% of your initial energy% left.

the cyclone 2 stage planetary, stage one is 97%

I highly doubt that. Maybe if you put through 150 or 200W. At 500W+ a lot will be lost since gears+bearing are overloaded. Just have a look at how tiny this gearbox is, then have a look into the datasheets of planetary gearbox suppliers. http://www.harmonicdrive.de/cms/upload/pdf/GK_2011_2012/de_en/planetengetriebe/hpg-126-143.pdf
As you can see, they claim 90% efficiency for a 2-stage drive of the same size as the cyclone and 11:1 reduction, at only 45Nm at the output. With a single stage, this figure wont be better at this torque level, since the gears+bearings are over stressed.
Companies like Harmonic drive cannot afford to lie and loose their reputation. Most of their customers test their products by efficiency and reliability means. Cyclone can tell as much BS as they want, since there are plenty of customers out there that believe their "truth".

It would be interesting to see what a Cyclone setup would draw under the same test conditions.

Really, it would be. Sadly no one will ever try both on the same dyno. We can talk a lot, only such a test would tell us the truth.

With you all the way crossbreak except for one minor detail.... The second stage is helical.

 

[Gab]

crossbreak

Additionally, the second stage is not helical as you can clearly see in the picture above. Helical gear reductions get around 95%, straight ones 97%... so this drive should get around 92.15% efficiency.

That is not correct, the Bafang is Dual stage helical spur gears as kepler pointed out. Helical gears are not 95% efficient, That type of efficiency can only be achieved with planetary gearboxes. " Most planetary gearheads are roughly 97% and 94% efficient for one and two-stage units, respectively. I had posted on another thread this link :

http://machinedesign.com/archive/spur-versus-planetary-gearheads-dc-servomotors


You can see their that strait cut spur gears are 90% efficient for single stage and 2 stage is 85% efficient . So helical especially dual helical is even worse then 85%, probably around 80% efficient hence about 20% losses just in the drivetrain. Their is a reason Optibike and Cyclone motors use strait cut planetary gears, they are a lot more efficient then helical non planetary designs. It's probably around 10 to 12% more efficient then a dual stage helical reduction.

So i guess in this design with the Bafang they have sacrificed efficiency for lower noise, this would explain the low speed people get for the power used, mid drive should be more efficient and faster then a hubmotor with the same power not the same, but if the gear systems uses 10% more power then this advantage can get lost it seems and it would explain why Kepler you have the same consumption with a middrive Bafang and Hubmotor geared Bafang.

 

[Kepler]

This is a bit of strange argument or should I say discussion I personally had no interest in mid drives before the Bafang became available and never had expectations that a mid drive would be more efficient than a geared hub motor as a commuter. I must say though, this drive has pleasantly surprised me. It has amazing power for its size, it is incredibly quiet, and without doubt, the stealthiest setup I have ever used.

I am more than happy to take a bit of an efficiency hit for the many advantages picked up using this drive. My god, Bafang should be sending me a sales commission :lol:

 

[crossbreak]

ok, it really looks straight in the above picture

The helical angle was chosen quite small to keep thrust loads low.

Gab, your source is highly questionable. it stands against all I have learned at university. no one bothers about higher efficiency of straight cut gears, since it is absolutely negligible. They are simply cheaper, which is the main reason to use them. This article is close to the truth: http://automotivethinker.com/transmission/straight-cut-gears-vs-helical/

 

[Gab]

That link i sent is from a servo motor manufacturer who compares actual values of efficiency with different gear heads on real data, what you learn in university is generalizations and not the real world data.

Its a known fact in car raceing circles that strait cut gears are more efficient, this cannot be denied. So you think its because they are cheap that strait cut gears are used in F1 gearboxes ? No, its because of efficiency ! You can read here:
"All the gears are straight-cut to minimise power losses"
http://www.formula1-dictionary.net/gearbox.html


Planetary gear designs are not a cheap option either ! they are one of the best gear reduction designs available for efficiency and compact power handling ability. But the issue is 2 fold with the Bafang, not only is it helical but it is not a planetary design, which are also inherently more efficient. So 2 different hits to efficiency that multiply as its 2 stage. If they had decided to chose helical for quietness then at least they should have made the gear reduction into a more efficient planetary design to balance out the efficiency losses. Yes the Bafang is a nice kit and quiet going buy this thread but one thing it is not is efficient compared to other common mid drives.

 

[Miles]

Ref: http://machinedesign.com/archive/spur-versus-planetary-gearheads-dc-servomotors

A typical single-stage spur gearhead, for example, is about 90% efficient whereas a two-stage model is about 85% efficient. Most planetary gearheads are roughly 97% and 94% efficient for one and two-stage units, respectively.

Well, I have to say that I was somewhat surprised by these figures.

 

[spinningmagnets]

I can't speak for everyone, but I would gladly swap several percentage points of efficiency for a very quiet drive. Perhaps I can just carry 5% more battery to achieve the same range?

The efficiency bump when using a mid drive on an uphill (the only reason I remain such a strong enthusiast concerning E-bikes is hills), by downshifting and keeping the motor RPM's up far out weighs the efficiency of a quiet direct drive hub. It's also that there are mid-drives with straight-cut gears, so...the buying public has a choice.

 

[crossbreak]

what you learn in university is generalizations and not the real world data.

One word: No. We do a lot of real world tests on the benches as well as in drive cycles, not only generic ones... efficiency figures must alsways be measured on a bench and can not be guessed by "real world experience".

As said, efficiency depends mostly on the motor rather than the gearbox. Claiming the cyclone to be more efficient than the Bafang middrive is absolutely pointless without real world data... no let me correct myself, without any data

The figures you quote are also no real world data, the article lacks of measurements or even literature reproval. I can't find any figures in the formula1 article. still it's completely new to me that F1 cars are built for outstanding efficiency. I'm still waiting for the 100mpg F1 car

 

[Cinelab]

Once we're on "porn"pictures, Could someone kindly post some of BB and crank axel.

Would Like to know if there's any way or modification that would make it fit on a Fat bike 100mm BB.

THX