'lightest.bike' 1.7kg 1000w mid drive

[nervagon]

I'll never own a sturmey again. I had the RK5 and it was shit. It needed constant re-indexing, would slip out of gear under force dangerously. I ended up with the a rohloff, and that was night and day. The rohloff was fantastic. I've read that only the 3 gear sturmeys are reliable, but my experience was so terrible I wouldn't want to risk it. The thing that makes rohloffs great with mid drives is how far out the cog is. Very easy to have an excellent chain line with motor like a BBSHD.

 

[nervagon]

Also with the CS-RK3 you'd have the combined weight of a cassette and an IGH. Ouch.

 

[nervagon]

Ah, bummer, Kindernay went bankrupt.

 

[raylo32]

I have never used an IGH but the idea is intriguing. What kind of shifter do you need for this Shimano or for a Rohloff? Do the hubs come with them?

Shimano makes a nexus 3 and the 36% gearing steps would work out like:

1st gear - 17.5mph 36% reduction <-- like dropping down to a 18.9" tire -
2nd gear - 24mph 100% - direct drive
3rd gear - 32.6mph 136% increase

.. and it's ~150g lighter ..

I think the Shimano is the right one.. you can run a >=14T rear gear and slow down the rate of chain/gear destruction that mid drives dish out. We're solving a typical mid drive problem and a gearing problem at the same time, not bad.

 

[nervagon]

I have never used an IGH but the idea is intriguing. What kind of shifter do you need for this Shimano or for a Rohloff? Do the hubs come with them?

Rohloff uses a special dual-cable twist shifter. It is a real pain to install actually, but you never have to fuss with it again after that.

In regards to the Bikee Lightest kit, it's important to remember that most IGH installs have very little slack in the chain, and the Lightest kit's internal torque sensor requires a slack chain. It hasn't been tested. The Rohloff tensioner that comes for bikes without sliding dropouts might actually be too stiff. I have one in a box somewhere but never used it because I had a single-speed frame. I had 14 speeds without a derailleur. It was glorious. I kind of regret selling that bike, especially at such a low price. It was going to be the recipient of the Lightest motor, but it was for the best I guess that I got rid of it before the motor finally shipped. It wouldn't have worked. The chain was taut and there was no hanger for a derailleur.

 

[neptronix]

I have never installed an IGH on an ebike before myself.
AFAIK it's typical to need to buy the shifter separate.

The downside of most IGH hubs is lower transmission efficiency than a chain. But if you do an ATF fill or some other lubrication upgrade, then you could bridge most of that gap and also likely improve the long term durability of the IGH beyond what it is stock.

Approximate cost of entry, all parts except wheel build included:
Shimano 3 speed: $133
Shimano Alfine 8: $275
Rohloff entry level model: $1250

I think i would be unhappy with a 3 speed drivetrain in the long run.

I think the ideal bang/buck case is the Alfine 8 with an ATF fill to reduce friction + a tune for the lightest controller that either adds a very slow soft start to the power ( if not present ), + a reduction in initial torque by intentionally running a lower battery amp : phase amp ratio.

Tuning down the initial torque would mean you need to use more of the gears to accelerate, but this would have a side effect of optimizing the efficiency you get out of the lightest and lowering the heat produced by the drive.. which would probably allow you to bump up the continuous power a little more.

In effect you would be making the motor feel a little laggy in trade for:
- Ideal gearing range to take advantage of the higher speeds
- Ideal efficiency
- Higher durability than a chain + derailleur setup ( unless you make a shifting boo-boo )

If i had an additional 20-35mm of length on the mid mount then i can just run oversized gears and forget about the extra cost of an IGH & all these concerns about making a non-bulletproof IGH durable enough to handle 1200W of power for at least 1000's of miles.

 

[nervagon]

I'll just add that when an IGH is installed as intended, you don't even need a narrow-wide chainring or chain guard. The chain can never drop. Chain bash guards on the chainstay are also a thing of the past.

 

[nervagon]

I have never installed an IGH on an ebike before myself.
AFAIK it's typical to need to buy the shifter separate.

The downside of most IGH hubs is lower transmission efficiency than a chain. But if you do an ATF fill or some other lubrication upgrade, then you could bridge most of that gap and also likely improve the long term durability of the IGH beyond what it is stock.

Approximate cost of entry, all parts except wheel build included:
Shimano 3 speed: $133
Shimano Alfine 8: $275
Rohloff entry level model: $1250

I think the ideal bang/buck case is the Alfine 8 with an ATF fill to reduce friction + a tune for the lightest controller that either adds a very slow soft start to the power ( if not present ), + a reduction in initial torque by intentionally running a lower battery amp : phase amp ratio.

Tuning down the initial torque would mean you need to use more of the gears to accelerate, but this would have a side effect of optimizing the efficiency you get out of the lightest and lowering the heat produced by the drive.. which would probably allow you to bump up the continuous power a little more.

In effect you would be making the motor feel a little laggy in trade for:
- Ideal gearing range to take advantage of the higher speeds
- Ideal efficiency
- Higher durability than a chain + derailleur setup ( unless you make a shifting boo-boo )

If i had an additional 20-35mm of length on the mid mount then i can just run oversized gears and forget about the extra cost of an IGH.

Rohloff comes with the rohloff twist shifter.

Also, I would have no problem buying a used rohloff because they run forever. But that also means that people rarely get rid of them!

But from experience with a sturmey RK5 and a Rohloff, neither gave me noticeably more top-end gearing than an 11T cassette cog would have. Are you sure you can get the gearing overdrive you need from a geared hub? I think you are going to be flushing money for loose-chained IGH setup that is still too slow.

 

[neptronix]

The only problem is an 11t gear will wear extremely fast.. and also wear the chain more.. at this power level you're probably replacing your drivetrain multiple times a year with regular riding.. i don't think that's acceptable. it's one of the main reasons i dislike mid drives.

The only way to overcome this is to run as large of gears in the front and rear as possible.

I wonder if adding soft start/ lowering initial torque would have a significant impact on the durability of a derailleur setup. I can't recall seeing anyone report on this.

..this would also improve the durability of the lightest.. especially since you and i both have 52v batteries and we're pushing it harder than stock in the first place.

But the question is how is how much room do we have to improve on the lightest' default tune?

I think you can answer that - with the throttle, does the motor have a nice 'kick' to it even if you're in the wrong gear for the situation?

 

[nervagon]

The only problem is an 11t gear will wear extremely fast.. and also wear the chain more.. at this power level you're probably replacing your drivetrain multiple times a year with regular riding.. i don't think that's acceptable. it's one of the main reasons i dislike mid drives.

This is why I like the Box Prime 9 series. They sell the top 4 gears separate for $16. Also, the Box prime 9 chain is 11-speed dimensions internally, and 9-speed externally. This means the chain plates are oversized/thicker.

I wonder if adding soft start/ lowering initial torque would have a significant impact on the durability of a derailleur setup. I can't recall seeing anyone report on this.

Let's call this "throttle thrash." An overlooked factor in this is hub engagement points. You want the chain and gogs to start spinning up immediately with the motor. My experience will differ from yours. Unfortunately the Bikee crank freewheel is kind of shit. Sloppy and low-engagement. My HXR chainring is very high engagement. There is no snap when the pawls catch up to the throttle.

But! I am pretty sure a White Industries premium freewheel is a drop-in replacement for the very low-end Chinese freewheel that comes with the Bikee Lightest kit:

E-Bike Freewheels – White Industries

That should be a noticable upgrade just based on how sloppy the Bikee freewheel is. I actually might be willing to go back to the Bikee bottom bracket/crankset if I do this upgrade to it.

..this would also improve the durability of the lightest.. especially since you and i both have 52v batteries and we're pushing it harder than stock in the first place.

I think higher voltage is almost always better for an IPM motor. I am NOT an expert in this, but increasing voltage while keeping the wattage limits the same is almost always better. Less ripple, less copper losses, less heat. If it's within spec for the FETs and controller, how is it putting more strain on the motor?

I think you can answer that - with the throttle, does the motor have a nice 'kick' to it even if you're in the wrong gear for the situation?

Wrong gear? There is no wrong gear on the 750W @ 52V. I can speed off from a dead stop in the 11T cog. For prolonged cruising and long climbs you want to be in a sensible gear, but if you live in a flat area you could probably stay in the highest gear most of the time depending on conditions. My testing was done in mild weather. I don't know if the motor would melt on a hot day.

 

[neptronix]

Interesting about the box drivetrain, $16 every few months makes this continual drivetrain wear problem more bearable.

I would guess we don't have an IPM motor but a typical brushless because this kit was once offered with an external ASI which doesn't support IPM.

To a point almost every motor on the market can benefit from higher voltage and therefore higher RPM. It's too bad we don't have a control group person running 48v so we can do efficiency/thermal comparisons but it's a pretty safe bet that we both improved continuous power with +/- .5% max efficiency difference.

Wrong gear? There is no wrong gear on the 750W @ 52V. I can speed off from a dead stop in the 11T cog. For prolonged cruising and long climbs you want to be in a sensible gear, but if you live in a flat area you could probably stay in the highest gear most of the time depending on conditions. My testing was done in mild weather. I don't know if the motor would melt on a hot day.

Nice, i imagine if you can do that then it's tuned for high initial torque which means we probably have room to detune for higher drivetrain reliability.

I have a thermal camera and my weather ranges from -10F to 110F yearly so whatever findings i have with thermals should be good.

Now if only my battery from grin would ship..

 

[rhauff]

To a point almost every motor on the market can benefit from higher voltage and therefore higher RPM. It's too bad we don't have a control group person running 48v so we can do efficiency/thermal comparisons but it's a pretty safe bet that we both improved continuous power with +/- .5% max efficiency difference.

My 48V battery is ready & waiting........and waiting..

 

[neptronix]

@Pilot Engineer when does our control group get his kit?

 

[nervagon]

@neptronix your calculations for 52V rpm seem a tad too high. Your estimates for loaded RPM actually perfectly match my unloaded numbers:


I know my max unloaded speed, so I can actually double check your calculations.

According to your numbers, the loaded RPM for the pinion gear @ 52V is 425rpm. That is actually the unloaded real-world RPM. So the real world 52V RPM under load is probably closer 372rpm to 380rpm in best case conditions, but see below for top gear, top speed RPM (this is the motors worst possible torque scenario).

And here Bikee is listing the max pinion RPM under nominal load.


Nominal load usually means under intended load conditions. So their marketing spec chart is listing a 48V loaded pinion RPM of 390 max.

@neptronix posted this a few pages ago:



My system is not quite achieving 450RPM unloaded at 52V, so I have no idea how they calculated 450rpm at 48V.
450RPM unloaded is probably without the motor attached to the drive train. That would account for my 5% lower real world unloaded RPM @ 52V of 425. Here is another way to calculate based on unloaded top speed:



And when I know my specific loaded top speed is 28mph to 29mph, we can calculate real-world loaded max RPM at top speed for a heavy rider:



This is so easy to calculate because the pinion gear is pulling the rear cog directly. I hope this helps. This is actual real world numbers. My numbers are considerably lower than your estimates.

And the most interesting part is I am losing 100rpm between unloaded and loaded operation at the top end of the cassette. My loaded performance (in high gear) is 76% of unloaded. Also keep in mind that the maximum torque possible in top gear is only 36Nm. The pinion gear has 33Nm of torque, and does not run through a reduction/torque boost before pulling on the cassette. So higher RPMs will be achievable in the lower gears.

More on torque in another post.

 

[neptronix]

It's very good to hear my math was off.
I took 15% off the unloaded speed and assumed that was the correct RPM given that this was the loaded RPM of a pretty similar motor. Definitely approximate!

I have all the tools to make proper measurements and will report back when i have ones from the real world.

100rpm difference between loaded and not loaded is within the range i'd expect.

 

[neptronix]

While i am waiting for the battery i was admiring the view of this drive today.

I can't believe how narrow i can get this drive with this 123mm bottom bracket. The pedals look incredibly narrow for a BB and i haven't even tightened down that right crank, so it goes in a few more MM.

Absolutely love it, i'm biomechanically most efficient with a very narrow stance and this is within a single digit % of perfect for me. it's a dream come true because i remember the BBS02 effing up my knee muscles in 2014 and i would never touch a mid drive until now, but always wanted one.



I measured my 123mm bottom bracket and it has an additional 3mm on the right side. This is approximately the amount of inset the freewheel needs. If you are using the bikee crank with this extra length then the addon PAS sensor should work. If you aren't, you probably lack the appropriate clearance.

The motor sprocket can also be flipped so it will line up perfectly with this narrower width.

I love that we can just use bike stuff!

 

[maiz]

I would guess we don't have an IPM motor but a typical brushless because this kit was once offered with an external ASI which doesn't support IPM.

Lightning rods are ipm and do great with ASI

 

[rivers]

I would guess we don't have an IPM motor but a typical brushless because this kit was once offered with an external ASI which doesn't support IPM.

I have seen photos of the Bikee motor. It is a conventional brushless inrunner.

 

[neptronix]

I have seen these pictures but will not share them out of interest of keeping the motor design outside the view of copycats for the manufacturer/designer. ( sorry )

This is a very well engineered kit and that motor looks super high quality with great heat shedding.

It has a smaller stator than Bafangs but easily double the pole count and therefore the torque will be higher and also the pitch of the sound will be higher.

Something constant about electric motor designs is that, the lower the RPM, the heavier, less efficient, but also quieter.

Human hearing is very strong in the 1000-8000hz range with a big peak at 3000hz and this motor is very likely loud not because it has a lot of sound energy but because it is in a high spot in our hearing range.

It's the same issue many geared motors have, especially if they don't have helical gears.



If we are running a sine/foc controller than we can expect the contribution of the motor to this sound is pretty low and the gears are the highest contributors to the sound energy.

This system has 2 gear reductions. The first stage is helical, the second looks straight cut and the interfacing is metal to metal.. so this is probably the loudest for sure.

Because this straight cut gear is bolted to the case, then this gear will transmit this extra sound through the case.
At least they made the highest RPM gear helical otherwise the gear noise would be unbearable.

The case is either magnesium or aluminum and these materials are great at conducing sound. The thinner, the better it is at conducting sound.
So are aluminum bikes, sound can transmit through the frame, the frame will act like a speaker. It won't amplify the sound, but give it more area to output the sound, increasing the distance and angles the sound can be heard from.

Here's some strategies for noise reduction i would consider trying.

General principle: use material to reduce the sound energy as close as possible to the source of that energy. It would be ideal to do this inside the case but it may be too hot inside to add rubber.

1. I think some rubber or other 'damping' material, if applied on the case near the output shaft, and placed under slight tension could absorb a good % of the noise close to the source. Gluing the material may not work. A wire wrap would be a cheap/easy way to do this for an experiment.
   
   
2. It may be possible to attach a damper or additional damper to the 2nd gear reduction's main gear. This is the best place you could absorb this noise!
   
   
3. More drastic: you could possibly put soundproofing material on half of the motor case ( where the output gear is ) and then add heatsinks on the opposite side to counteract the loss of air cooling and also concentrate the heat away from your soundproofing material ( rubber? ) which can be damaged from high heat.
   
   
4. You could design some metal piece that covers the motor and is lined with sound deadening material. This would work, but there has to be a nice airflow to the motor still.
   
   
5. Re-engineer the gear itself for these attributes: How to reduce Gear Noise | KHK Gears

Sound damping material, if applied correctly, should be very effective in the frequency range this motor produces.

 

[Golo]

Maybe adding some rubber between the frame and the motor and put it under a bit of pressure could reduce the noise. I will try that on mine.

Log into Facebook

Log into Facebook to start sharing and connecting with your friends, family, and people you know.

www.facebook.com

Watching this video on the owners group, and to me it sounds like this one is not bad at all.

I think and hope this motor will be great for me. Its good to hear that it is a inrunner and probably has good heat dissapeation. My current x1 stealth sometimes throttles down when running about 1000w for some minutes at steep climbing. This is a outrunner. With not the best heat transfer to the outer casing.

If this lightest 1000w motor has about the same power as a stock BBSHD as @nervagon describes I will have about the same power as I am used to also. With this little weight added to the bike.
I am going to build a 14s2p 21700 battery at about 2 kg so I will probably add only 4kgs total to my bike.

 

[nervagon]

Maybe adding some rubber between the frame and the motor and put it under a bit of pressure could reduce the noise. I will try that on mine.

Log into Facebook

Log into Facebook to start sharing and connecting with your friends, family, and people you know.

www.facebook.com

Watching this video on the owners group, and to me it sounds like this one is not bad at all.

I think and hope this motor will be great for me. Its good to hear that it is a inrunner and probably has good heat dissapeation. My current x1 stealth sometimes throttles down when running about 1000w for some minutes at steep climbing. This is a outrunner. With not the best heat transfer to the outer casing.

If this lightest 1000w motor has about the same power as a stock BBSHD as @nervagon describes I will have about the same power as I am used to also. With this little weight added to the bike.
I am going to build a 2s14p 21700 battery at about 2 kg so I will probably add only 4kgs total to my bike.

Based on my experience with the 750W, if you ordered the 1000W, it should be very comparable to the performance of the BBSHD at 1/3 the weight.

 

[neptronix]

The location of the controller is close to the motor and separated by a plate. Both have great contact with the case and use the case as a heatsink. There would be a great benefit in adding heatsinking on the part of the drive on the side opposite of the output gear.

Even more drastic would be to remove the controller, but the additional power we can output is limited by a plastic gear, so you may not be able to use much of this extra ( and significant ) heat reduction.

But you may not need to do any of this - unlike these other mid drives it seems the power rating lines up with it's thermal capabilities in stock form at least

 

[nervagon]

My unloaded RPM on the Lightest Kit is roughly 425.

85% of that would be around 360rpm for my target loaded RPM. My comfortable cadence while wearing regular street clothes is about 80rpm max. So that means I must be able to fit at least a 44T chainring to sync my body to the motor.

44T both syncs my cadence to the motor RPM and matches the motors real-world top speed.

And so that means either medium mount or dual chainring. But my current config is not optimal.

 

[neptronix]

I have a brand new 44T that fits and a medium mount if you want the easy way out

 

[nervagon]

@neptronix thanks for the offer, but I think I will be taking the Lightest kit off this bike. For me, it just doesn't match my vision for a stealthy city bike, so it's not worth the effort to go rearranging my install (and moving the motor further up the downtube with medium mounts will make it even more conspicuous). The Photon will better fill the stealth bike role.

But it could be an interesting candidate for my cargo bike:



It could replace the BBSHD and mount in-frame behind the seatstay using the dual chainring config. The chainstay is a monostay so I could fit a dual chainring of any size/configuration, and the motor ring could be well inboard and well clear of the rest of the drivetrain. The rear hub spacing is also boost, so I just have a ton of wiggle room to work with. Noise and simple PAS would be less of a concern on this particular bike. It's for hauling. It's a tank.