Full suspension mountain bike Friction Drive

How much contact width are you actually getting/need?
I'm thinking of the rollers which run between the tread blocks and how little could be gotten away with.
That's exactly the one I'm curious about. Possibly if it could be combined with the cogged roller idea, but with enough space that the cogs would only make contact if it started slipping.
Ride infrequently. Still rocking a Cannondale F700 from time-to-time. That's the model Cannondale loaded with as much '90s tech as was available: Headshok, Grip Shift X-Ray shifters, Onza pedals with the rubber inserts, and a subpar proprietary cable pivot for the cantilever brakes that was dropped for a Tri-Dangle which wasn't exactly an improvement.

Haven't looked at bikes in 20 years. What are the options for used, full suspension https://bikesbros.com/full-suspension-mountain-bikes/, titanium mountain bikes? Something I can find on Craigslist around a grand. Found a few hard tails in what appeared to be decent shape for less so am assuming that's a reasonable target.

Litespeed used to be good. Dean used to be good but more boutique. XC/all-around suspension rear suspension meaning it doesn't need to have huge travel (but more than a Unicoi). A good rear suspension design would be nice. Horst link? Something better than the unified rear triangles that were new 20 years ago. Titanium for no other reason than the cool factor.

Grantmac said:
That's exactly the one I'm curious about. Possibly if it could be combined with the cogged roller idea, but with enough space that the cogs would only make contact if it started slipping.

Each side of my roller is 4mm wide, so 8mm total. The actual contact area would be slightly more because it's angled.

There are no problems with slippage when the pressure is dialed in right, but I don’t yet know if the wear is acceptable - I rode about 70km of mixed road and mtb with almost no noticeable wear. But then I accidentally let the motor mount get crooked (maybe due to heat, or me constantly messing with it) and it munched almost through to the casing threads in about 20km. Just like a car with crooked wheel alignment.

The new mount I'm making should keep a straight even contact much better, and I'll post again when I’ve done a longer period of testing with it.

I also tried a "makes contact with casing around knobs" roller with Specialized Slaughter, and Maxxis Forekaster tyres but had the same result as Kepler.

There is always a tiny bit of slip with this system because rubber. As a result the voids in the roller always end up scootching up to the edge of the tread blocks. Even though it's not pushing on them hard, it's still striking each knob as it rolls around, which makes for the jet engine noise Kepler aptly described.

However I did find it difficult to correctly mesh the roller to the tread knobs over the curved tyre profile as the circumference changes from centre to edge. A nicely meshing volute gear profile may be more successful.

Here is the one I tried with the Slaughter tyre:
It sounds as though slippage is just a fact of life. Interesting to hear that 8mm of contact can work. Although it seems like a rigid mount becomes very important with the narrow contact patch.

I wonder if you could create a roller which makes contact with the sides of the knobs or if you would have the same noise issue?
Has anyone on ES successfully driven the rim? Stumbled on this the other day, seems like this concept could solve some of these problems as well as I’m sure opening up a bunch of new ones :). Thought I’d share, may spark some ideas for those that haven’t seen it.

this may be it usefull as idea?
Finally found a true MTB tire that works with a friction drive. I am not sure if the tire is new model but was very pleased to find it.

Tire is a Schwalbe Smart Sam. 2.6" wide 29'er. It has a thread section down the middle of the tire that is closely packed enough to run a roller over without sounding you have pegged a playing card in your spokes.

Fitted the tire to my current friction drive MTB and added an 8mm wide drive ring to motor to pick up the centre tread. Only just started playing around with this configuration but it is looking very promising. The drive ring is just made from aluminium at the moment but if wear is an issue, I will make it from steel. Also keen to play around with knurling the ring for better grip. So far I have put 200W through it with relatively light tension without slip in dry conditions.



I have now fitted the tire to my 2019 Specialized Carbon Comp Stump Jumper to see how it performs out on the tracks. Really pleased with how it has performed so far. Found the grip to be more then adequate compared to the stock tire even in quite muddy conditions.


So I can see a new Friction Drive project in my future. Hope to bring all my learnings together on this one.

The Plan:

2019 Specialized Carbon Comp Stump Jumper donor bike.
Internally mounted and fully hidden batteries. (200 Whr) (Bike has SWAT internal storage system)
Internally mounted and fully hidden speed controller and associated electronics
Pivoting friction drive motor that will mechanically decouple from the tire when assist is not required
Regen capabilities
Range extender bottle battery (200Whr) compatible.
150W typical assist with 300W boost when required.
Will only add 1.5kg to the stock bike's weight making the bike a full 2kg lighter then the new benchmark Specialized S Works Levo SL.

Stay tuned. New thread to begin shortly. :)
Slow but steady on this project. Lots of protyping but finally making headway on the Specialized Stumpjumper Carbon Comp.

Mounting to an oddly shaped carbon fibre swing arm has been a challenge but with a combination of hand carved machined plastic to meet the curves and bumps, milled aluminium, and trusty cable ties, a solid motor mount has now been achieved.

Mechanically I am close. The motor engages and disengages nicely on and off the tire using a cable system activated by a fork lockout lever. Grip pressure is good for around 150 Watts at this stage. Nothing to write home about but on bike that that has not been compromised by weight or drag, 150W out on the tracks is gold. I think perhaps 250W of peak power will be achievable with knurled drive ring on the motor.

Importantly, there is no rattle or sound from the drive when disengaged. The bike feels no different to stock standard. Engaged there there is a bit of tire contact noise but no more then knobby tire noise on a sealed road.

Next job is to sort out the drive activation method. I am thinking a hall sensor on the swingarm so I don't need to run any extra cabling to the handle bars.

Ground clearance has been reduced slightly with the drive 10mm lower then the the 32 tooth chainring. Not ideal bit hopefully an acceptable compromise.



In terms of grip and knobbly tyres, might it be worth exploring the opposite end of the scale to sandpaper?.. I wonder if a flat-surfaced low-shore silicone ring moulded around the motor cylinder - and having, say, a 10-15mm thickness - might impress into the harder rubber of the tyre, when engaged, and act as a sort of soft cog? Some of these two-part mould-making silicones have high resistances to tearing. The silicone ring might be further strengthened with (tyre-clearing) disc walls either side.
Skateboard wheel on hybrid road/MTB tyre no slippage 1500w brushless motor 62kmh on Diamondback dual suspension mtb
I tried using a knurled steel roller ,I could wear a brand new tyre to canvas in half hour,using skateboard wheel works great for me 2000km already and still half tread left using 1500 W 62kmh
Also tried silicone over the roller it will tear off anything you try to put no matter how good you clamp it
mark06082013 said:
Also tried silicone over the roller it will tear off anything you try to put no matter how good you clamp it

OK, scratch "silicone" from my previous comment and substitute "polyurethane" in its place. As skateboard wheels are made from polyurethane, setting a 10mm ring to bond around a motor's cylinder might achieve a similar level of grip without the bulk of adding an extra wheel drive to the mechanism.
I have tried so many methods of getting reliable drive from the motor to the tire. Been down the polyurethane track and it does improve grip over a smooth metal roller but any moisture on the tire, and we are back to square one.

This is a mountain bike so you are going to be very limited to when you can ride if the track is damp.

At this stage the knobby tire has been placed on the backburner due to noise and replaced with the latest version of the Schwalbe Hurricane. the tire needs a bit of rework with a finger belt sander to make the rolling surface completely smooth, but once you do this, the tire is as quiet as any tire I have tested. This latest tire in 2.4" wide and offers a surprising amount of grip out on on the tracks even in damp condition. Trick is to run its minimum pressure of 23psi to get the side knobs more engaged and team it with a really aggressive front tire.


To make the roller work in wet weather, I have gone back to belt sander material but applied it with better moisture protection for durability. What I previously found was that the moisture would get under the edges of the grip tape causing it to de laminate in a short period of time in wet conditions. I have now added adhesive heavy duty heat shink on each side of the roller to stop moisture getting to the edge of the grip tape. With this setup, the grip tape is the part of the roller in contact with the tire.


I have riden this in both damp and really muddy conditions. No surprise, deep mud sucked however packed wet and damp conditions we not a problem at all.

As previous mentioned, this is a low powered setup with 100W used most of the time and boosting to 150W when I really need it. This might not sound like much, and it isn't however it is just enough to make the bike feel light to pedal. Riding with my regular non assisted friends, we usually pretty even in overall speed however with just 100W extra, they had no chance of staying with me through the hard pedaling sections of the ride

One small drawback with the belt sander material is that dirt can stick to it under some conditions but not enough to cause any major issues.

Internal batteries are now completed. These are tiny. 2 x 6S1P 18650 for 3ah per pack. With the 2 packs, this gives me around 120Whrs which gets boosted to around 160Whrs with regen. This is good for around 30km of assisted riding which is plenty for me especially with knowledge that if when completely flat, the drive is fully disengaged from the bike.




Weight penalty is negligible. The drive system including internal batteries adds 1200 grams to the bike. The Schwalbe tire I am now using is 300 grams lighter then the original tire I had on the bike and I have installed a light weight rear cassette for another 300 grams savings. So total added weight to bike is a mere 600 grams which is the the weight of a water bottle. Needless to say their is no weight penalty that you feel on the bike.

An added bonus is that the bike rolls even better then before due to the change of back tire.

I have bought a knurling tool and am keen to still give a knurled roller a go. I am hoping with only 100W, the tire will handle the load.
mark06082013 said:
Skateboard wheel on hybrid road/MTB tyre no slippage 1500w brushless motor 62kmh on Diamondback dual suspension mtb

Would love to some pictures of this setup. Sounds interesting
Added a bash guard for a bit of extra protection for the motor.



One of the minor problems I found was that the assist is so subtle, sometimes you forget it is on. So I have added 2 LEDs. One LED is low power (100W.) 2 LEDs is boost (150W) Both flashing is regen. Went for high intensity LED's. OMG they are bright. Might have been a bit of overkill :) . I will wait till I ride the bike out on the tracks before I decide if they are too bright.


Tested my new knurling tool. Works a treat on aluminium. I have no idea how well the knurling on an aluminium ring will stand up.
Time to do some testing.




  • 20200705_134601.jpg
    479.7 KB · Views: 1,666
Experimenting with knurling the motor can directly. No problems cutting the knurl on to the motor can and the grip on the tire is excellent. Time to do some testing. I am not concerned with accelerated tire wear with the small amount of power I am using. More interested in how well the knurl wears.

Interestingly, the knurl I did on the aluminium ring on my other bike stayed effective for around 100km. Hopefully the knurl into steel will have decent longevity.


  • 20200710_172237.jpg
    299.9 KB · Views: 1,613
Finally running a serious MTB tire with a friction drive. Feels like the last piece to the puzzle.

Selected a Specialised Butcher Grid 2.6" 29er. It's actually meant to be a front tire which isn't a bad thing as front tires tend to have a more aggressive tread pattern than rear tires. This tire has about a 5mm grove in the centre of the tire so using and electric finger file, I cut a 10mm centre groove in the tire to accommodate a smooth surface for the motor roller.


What I didn't like about this was that cutting the groove weakened the tire significantly down the centre and felt it would fail quite quickly once a friction roller was applied to it.

Solution was to add a 10mm friction band to the centre of tire. This was made from a worn 700C road tire. Basically I just cut off the sides of the road tire with dressmakers scissors leaving just the centre of the tire 10mm wide. This band has stretch in it so deflating the MTB tire, fitting the friction band, and re inflating the tire made for a nice snug fit.



Next I needed to machine up a 10mm roller for the motor to match the 10mm gove on the tire. Based on the excellent feedback on this thread, decided to use a skateboard wheel and machine it suit. Got to say, results are mixed. It's not a bad solution for 250W and dry conditions which is fine for our Australian summer however I do want more so stay tuned for more experimentation using a knurled steel ring.



Anyway, the bike has now done around 100km on serious single track and I have got to say I am loving this setup. The additional 1500 grams placed on the bike is insignificant and the power, although subtle makes the bike feel like 9kg cross country bike while climbing without compromising the bike's "normal 150mm MTB" feel when descending.

I am now running 6S3P using 18 3ahr 18650 cells which fits nicely inside the frame. This gives me 25km of full time assist on single track which is heaps considering how hard I am going. (I am rarely overtaken)



  • 20201115_171339.jpg
    668.6 KB · Views: 1,149