I created a purely mechanical Turbo Levo speed limit increase

[jd_3d]

------------------------------------
EDIT 2/21/2019 - Hey it's me JD from the future (2 months after this first post). I wanted to add a little info here on how you can obtain your own set of gears that is 3d printed by Shapeways so I created a quick Wordpress site with some info:

Guide to the LSS

In there is information about the latest design, links to purchasing, and installation instructions (or if you just want direct links, here is the LSS and LSSDT). Anyway, I still encourage you to read this entire thread as it provides a ton of information on the development, prototype feedback and reviews of the kit. Enjoy.
-----------------------------------

I've only had my 2019 Turbo Levo a few weeks but already felt the need for speed. I do a decent amount of road riding with my bike (at $7k, it's my do-it-all bike) and felt the 20 mph speed limit was a little too limiting. However, I don't want to mess with firmware changes that could brick my bike or cause warranty issues. I also had a number of requirements:

1) Must be 100% mechanical (no software changes)
2) Must be easily reversible (less than 5 min)
3) Must not have any re-wiring of the bike/sensors
4) Keep speed limit at around 28 mph.

In California Class 3 electric bikes w/ 28 mph speed limiters are legal on the roads, so If I made such a device I would be reasonably within the law when riding on the road. When riding on trails I do not need more than 20mph.

My wife thought I was crazy as I sat staring at my bike for like 30 minutes at a time lost in concentration for thinking about the best way to do it. Fortunately I'm good with CAD and have several 3d printers so once the inspiration for a planetary gear set came to mind, I was off designing...

A few days later (and several failed iterations) I had prototype 3 installed on the bike (ignore the ugly blue color, it will be stealth black eventually):

Little did I know at this point it would take 13 more prototypes before I got it working to my satisfaction. Since this post is getting sort of long I'll stop here and call it Part I. If you guys are interested I can do more posts and continue the detailed information on the design/troubleshooting and final outcome (spoiler: it's awesome).

Obligatory info:
Please read about the electric bike laws in your country: Electric bicycle laws - Wikipedia

 

[Maastricht]

You are a great engineer! Surprising sometimes how many smart people are walking around on this planet. I am very curious about the end result. Only pay attention to possible mud clogging as well.

 

[jd_3d]

You are a great engineer! Surprising sometimes how many smart people are walking around on this planet. I am very curious about the end result. Only pay attention to possible mud clogging as well.

Thanks! This was actually one of the most difficult things I've designed/created, even harder than a full ebike conversion I designed/built last year on my hard tail. I think I put 50 hours into it so far, which is a bit nutty for a few small gears I'll do another post tomorrow with more details.

 

[outerlimits]

What is the principle of it ? To make the sensor magnet go around slower ?

If so, could one not just relocate the magnet and pickup ?

 

[R120]

Wow, that is a serious bit of bodging - awesome!

 

[Krisj]

Very clever idea ?interested to see the progress

 

[James_MTB]

That's awesome

 

[R120]

Guessing it will need some sort of gasket/cover to keep grit and mud out?

 

[Stumpy]

Ingenious! - how does it work to increase the speed limit?

 

[Rob Rides EMTB]

That is awesome!

I guess the gearing slows the rotation of the magnet compared to wheel speed?

 

[Scoobaru]

That is awesome!

I guess the gearing slows the rotation of the magnet compared to wheel speed?

Yes, it's the opposite of what the crank on a Bosch bike does. On that you spin the crank one rev and the chainring spins 2 revs. This will spin the magnet at less revs than the wheel.

I agree this needs sort of gasket/cover for mud protection, in the UK anyway.

The teeth look very small, how tough is the material ?

 

[Stumpy]

Brilliant! - love things like this!

 

[Koo]

Well done, ingenious!

-Kooo

 

[Mcharza]

Brilliant, thumbs up!

 

[Sepp62]

Really cool !

Will you share the solution with the community ? (STL-file)

 

[2manytickets]

That’s cool. Good job. Oh and if you need any testers I’m in California. Ha ha

 

[jd_3d]

Part II

Wow, I'm amazed at all the feedback already. I wanted to first address some of the questions raised so far:

What is the principle of it ? To make the sensor magnet go around slower ?

Yes, the idea is the slow the rotation of the magnet relative to the wheel rotation. See below where I explain this in more detail.

If so, could one not just relocate the magnet and pickup ?

Moving the magnet and pickup to another spot on the wheel won't change the RPM and hence won't alter the perceived speed. You can relocate it to say your cranks so that it reads your crank RPM, but that is a more permanent change that wouldn't be easy to reverse if you need to take your bike in for service.

Guessing it will need some sort of gasket/cover to keep grit and mud out?

Yes, I have that in later designs.

I guess the gearing slows the rotation of the magnet compared to wheel speed?

Yes, exactly, I'll go into a little more detail below.

The teeth look very small, how tough is the material ?

The teeth are tiny (only about 1.5mm) but there is zero load on the entire gear set so strength is not too much of an issue.

For those interested calculating the gear ratio for a planetary gear set uses this formula (from Planetary gear ratio calculations):

Working out planetary gear turns ratios

Tr Turns of the ring gear
Ts Turns of the sun gear
Ty Turns of the planetary gear carrier
R Ring gear teeth
S Sun gear teeth
P Planet gear teeth
The turns ratio is as follows:
( R + S ) ×Ty = R × Tr + Ts × S

For my setup the gear ratio turns out to be approximately 1.5:1, or 1.5 rotations of the wheel will result in 1 rotation of the magnet.

So, for 20 mph bikes this will go to -> ~30 mph (a little less since you can't quite get to 20 mph in the first place)
For 25 kph bikes this would go to -> ~37.5 kph

Now, back to prototype 3, here is a video of it in action (sorry for the poor audio, my wife was washing dishes):


You can see the ring gear is mounted to the disc rotor so it rotates at the same speed as the wheel. The sun gear is mounted to the wheel hub and does not rotate. The planetary gear then rotates around the two.

Here is a photo of the parts before assembly:

This prototype had 2 major flaws:
1) The 3 mm magnet was too small to trigger the sensor
2) It required glue in the assembly process as well as gluing it to the rotor. Ok for a prototype, but not good for real-world use. I wanted zero gluing in the final design.

In Part III I will go into the redesigns I did and talk a little of the troubles I had with later prototypes.

 

[R120]

The video is quite mesmerising!

 

[Maastricht]

For a "newbie" very impressive, creative and thinking outside of the box! You must be a NASA engineer.

I assume Rob will reward you soon with a full membership.

Thanks for sharing and the planetary gear lecture!

 

[MattyB]

This is a great idea beautifully realised - if you can make it operate reliably even in sloppy conditions I suspect you are going to be overwhelmed by demand from around the world!

 

[oldfox]

That's just awesome, I love it!

Probably a good way to earn some money... Not just mine!

 

[jd_3d]

Part III

I realized I needed to go to a 10mm magnet (same diameter as the stock one) but this meant the entire gearset needed to be larger. The problem is there is so little clearance between the disc rotor and frame. You have about 7 mm of clearance in the inner area but once you move the where the sensor is you only have about 4 mm! This was the hardest part about the design. Take a look at the already tight clearance with the small magnet and imagine trying to make it 7 mm bigger:

Here is a small fraction of some of the prototypes (many of them I threw straight in the trash from pure frustration).

I finally had something working with a 10mm x 2mm thick magnet but then found out it wasn't reliably triggering the sensor so I had to move to a 10mm x 3mm thick magnet (which has been rock solid). But that was another 1mm of thickness I had to deal with.

Anyway, I found various ways to make the entire assembly thinner and then had to deal with how to attach the ring gear to the rotor. I went with a retaining ring that grabs onto the rotor with small clips. Once I got the tolerances right it snaps on really nice:

Next up in Part IV I'll show the final design (I still need to take some pictures/videos of it).

 

[outerlimits]

Thanks for answering my initial questions. You now have my attention.
Interesting you had to go down the road of a clip on mount. This would require you to use the rotor the clip mount was designed for ?
Would a clamp mount using the space behind the rotor be more of a universal fit ?

 

[TPP]

I've had an idea which would give a similar end result, but uses a different method. Unfortunately I'm lacking the knowledge of how actually make it.
It would satisfy the OPs requirements as I agree the need to not invalidate the warranty is vital.
The idea is to make an intermittent electromagnet which bolts onto the rotor in place of the current normal magnet, and switches off every other rotation. This would double the cut off speed. So for me in the UK 30mph. I imagine that it wouldn't be too difficult for someone to connect a phone accelerometer and or gyroscope chips to a small electro magnet and battery and then write the necessary code meaning that the magnet was switched on every 2nd rotation.
Do people who know a bit about electronics think this would be possible?

 

[R120]

I've had an idea which would give a similar end result, but uses a different method. Unfortunately I'm lacking the knowledge of how actually make it.
It would satisfy the OPs requirements as I agree the need to not invalidate the warranty is vital.
The idea is to make an intermittent electromagnet which bolts onto the rotor in place of the current normal magnet, and switches off every other rotation. This would double the cut off speed. So for me in the UK 30mph. I imagine that it wouldn't be too difficult for someone to connect a phone accelerometer and or gyroscope chips to a small electro magnet and battery and then write the necessary code meaning that the magnet was switched on every 2nd rotation.
Do people who know a bit about electronics think this would be possible?

This is how the Badassbox works

Badass eBikes Tuning für Bosch Brose Continental Fazua Kalkhoff Panasonic Pendix Shimano Yamaha

 

[TPP]

I've got a Specialized Kenevo with the magnet on the brake rotor. It looks like the badass box works with spoke mounted magnets. I don't think it a badass would work with a Kenevo though please correct me if I'm wrong.

 

[jd_3d]

I've had an idea which would give a similar end result, but uses a different method. Unfortunately I'm lacking the knowledge of how actually make it.
It would satisfy the OPs requirements as I agree the need to not invalidate the warranty is vital.
The idea is to make an intermittent electromagnet which bolts onto the rotor in place of the current normal magnet, and switches off every other rotation. This would double the cut off speed. So for me in the UK 30mph. I imagine that it wouldn't be too difficult for someone to connect a phone accelerometer and or gyroscope chips to a small electro magnet and battery and then write the necessary code meaning that the magnet was switched on every 2nd rotation.
Do people who know a bit about electronics think this would be possible?

Hi TPP, yes it's definitely possible to do it that way, but I'm weak in electronics so I would have a hard time designing it. Another challenge would be making it compact enough. I'm sure something like an Arduino could be coded to do what you want, but it's a bit large and then you have to worry about powering it with a battery. The clearances are very tight in the area of the sensor. Cost would also be a factor, as most of the electronic devices are over $100.

 

[jd_3d]

Thanks for answering my initial questions. You now have my attention.
Interesting you had to go down the road of a clip on mount. This would require you to use the rotor the clip mount was designed for ?
Would a clamp mount using the space behind the rotor be more of a universal fit ?

Yes, my current design requires you to be running the stock brake rotor. Do people switch these out often? Can you explain your idea for a clamp mount in more detail? What would it clamp onto, the rotor? If it's the rotor I'm not sure how you'd create a clamp that works for all types of rotors (i.e., the cutout patters are all different). I guess you could clamp to the spokes but that would also depend upon the wheel lacing, etc...

 

[jd_3d]

Part IV

I switched my brake rotor bolts to Titanium as the magnet was being attracted to the steel bolts and causing some noise. Throughout the design process I had noise issues where you could hear the gears making faint sounds. Finally I realized I could print the planetary gear in a soft TPU-like material which should quiet things down significantly. It ended up working great and now I cannot hear the gears at all when riding the bike. I think with the latest design the steel bolts would be fine.

The final prototype has a dust cover which also makes it more stealth. In my design the dust cover is slightly short because I still wanted to see the planetary gear so I knew it was working properly.

Final design w/ dust cover:

Final design with dust cover hidden. Note I realized the planetary gear would be kept in place by the magnetic attraction to the steel brake rotor, so that saved on some complexity compared to earlier designs:

A photo of the latest version on my bike:

A few photos showing the tight clearances. I'm right up to the limit in several areas:

Here's a short video showing the gear reduction. Notice the yellow piece of tape on the rotor and the small white gear in the gearbox aligned with it. After 1 rotation of the wheel you can see the planetary gear has only rotated about 2/3 of the way:


Here is a video of it in action:


Review
I've put about 50 miles on this latest prototype (V14) and the best thing I can say is I don't even know it's there. I don't hear it or feel it, but I can go faster and I think that's all you really want in a device like this. I live in California so I don't ride in muddy or wet conditions so I'm not sure how it would hold up, but I do think that area of the bike must be one of the least prone to get muddy (as all the mud is expelled away from the tires). Maybe someone who rides in the wet could answer, do you find that your brake rotor bolts are covered in mud after a ride?

Anyway, that's it for now. Maybe in Part V I could discuss options for building/selling these to some of you as I've already been PMed about that. If you are interested drop a quick note, that would help me understand if its worth pursing.

 

[outerlimits]

Yes, my current design requires you to be running the stock brake rotor. Do people switch these out often? Can you explain your idea for a clamp mount in more detail? What would it clamp onto, the rotor? If it's the rotor I'm not sure how you'd create a clamp that works for all types of rotors (i.e., the cutout patters are all different). I guess you could clamp to the spokes but that would also depend upon the wheel lacing, etc...

You will find people do infact run different brake rotors to the standard.

I’ll try and explain the clamp method.

Your design looks like a plate on the outer side of rotor is attached via clips to the rotor.
Introduce a plate to the inner side of the rotor that with the use of bolts/Screws attaches to your outer plate. Thus sandwiching the rotor in the middle.
The inner plate could be 2 pice for ease of figment white rotor attached. Or one piece requiring the removal of rotor for figment.