Full suspension mountain bike Friction Drive

falcongsr said:
Nice work! Are you charging at work for the ride home with that supply on your desk? Are you using any kind of balance circuit?

Yes, charge back up for the ride home. No balance circuit, just using a cheap boost converter connected to a laptop supply. current tapers off to zero as the pack reaches full charge. I only balance charge if I really need to. Check the balance after most charges at home though.
 
Boost regulator looks like this one: https://www.amazon.com/WINGONEER-10-60V-12-80V-Step-up-Converter/dp/B01MG4211N?keywords=dc+boost+converter&qid=1539178377&s=Electronics&sr=1-7&ref=sr_1_7

Nice that both the voltage and output current are adjustable.
 
Kepler said:
However, all is not lost and perhaps we can do this with a 50kv motor on 24V and single reduction of around 10:1.

So we now have 50kv x 24V = 1200 rpm. then a 10:1 single reduction with 120 rpm now at the crank. The Vesc has no problems at a 50% duty cycle or even a bit lower so no problems tuning the crank at 50 rpm.

And now the engineering.
So the mount is easy and can be mounted up nice tight in a similar position to my latest bracket but using the ISCG-05 mount.
In regards to the drive system, I would consider using a #25 chain and sprocket combo. This will keep the sprocket diameter quite compact and should be plenty strong enough to handle the power output of this motor combo.

Keeping it simple,the motor would be connected direct drive (via reduction) so the cranks will turn on there own with the drive is activated and when free pedaling, the motor will be driven. Not sure how much drag this will place on bike when pedaling but it may well be quite minimal. However, if the drag is too great, I would consider using a sprag clutch on the motor side. This wont stop the pedals turning under power but would remove the drag on the system when free pedaling.

So it looks like there might be another project in my future. :D Best I finish this one first.
So I built almost exactly this about three months ago on a Giant Reign, except a 125kv motor and a 4S pack.
https://endless-sphere.com/forums/viewtopic.php?f=28&t=96769&p=1417783#p1417783
Going to direct drive eliminated the freewheel and square taper crank which saved quite a bit of weight, and as a bonus the motor picks up seamlessly since it is already spinning. I really don’t notice the motor drag, and I ride a lot without assist.

By the way, I received your Friction Drive mount a few months ago. Very Nice! It looks like I can easily adapt it to my Giant Trance swingarm.

I am curious about the life of your Mountain Bike tire, as I run a Specialized Slaughter in the rear, and I would prefer not to burn them up. (They are not cheap!)

Also, have you done any steep (15-20% grade) climbs on dusty dry dirt? Was the friction drive effective?
 
That's nice setup FZbob.
I'm still looking at using a 60KV motor with a 5S battery . John suggests 24V (7S),can I ask why so high with the voltage if we are trying to keep the rpms down. I know these motors run most efficiently at around much higher rpms,so can anyone shed light on what would happen if I run a 60KV motor @18.5V and then limit the rpm further via the vesc ? One other thing ,when the Vesc (assumingly) limits the V's to slow the rpms down does it also limit the current at the same time ? Or,does it increase the current proportionally to achieve the same power output?
Could this be a problem with too much current getting pushed as a result of a drop in voltage? Electronics 101 continues....
I'm thinking of a similar setup as FZbob but using a HTD toothed belt and pulley attached to the inside of the crank. I think this may be quieter and I hear that they are capable of handling the torque. Just trying to see if I can source one in aluminium rather than steel for the weight benefits.
Fzbob,how did you overcome the 185rpm at the cranks? Or is much less when under load? I'd like to try this with less reduction if possible.
 
FZbob,

That is an interesting build. Basically a working version of what we have been discussing. Nicely done.

The Specialized tire you mentioned wont work with a friction drive. Needs to have a slick centre. One of the drawbacks of friction drive. The tire i am using works suprizingly well.

I have done probably 20% climbs. With grip tape, no problems. It will climb the same as your RC mid drive setup.
 
urga said:
That's nice setup FZbob.
I'm still looking at using a 60KV motor with a 5S battery . John suggests 24V (7S),can I ask why so high with the voltage if we are trying to keep the rpms down. I know these motors run most efficiently at around much higher rpms,so can anyone shed light on what would happen if I run a 60KV motor @18.5V and then limit the rpm further via the vesc ? One other thing ,when the Vesc (assumingly) limits the V's to slow the rpms down does it also limit the current at the same time ? Or,does it increase the current proportionally to achieve the same power output?
Could this be a problem with too much current getting pushed as a result of a drop in voltage? Electronics 101 continues....
I'm thinking of a similar setup as FZbob but using a HTD toothed belt and pulley attached to the inside of the crank. I think this may be quieter and I hear that they are capable of handling the torque. Just trying to see if I can source one in aluminium rather than steel for the weight benefits.
Fzbob,how did you overcome the 185rpm at the cranks? Or is much less when under load? I'd like to try this with less reduction if possible.

I suggested 7S to keep amps down. The 60kv motor really opens up your options with a single reduction.
 
A few updates on the friction drive setup.

I have increased the size of the battery bottle. Trying to squeeze everything into a standard bottle was getting ugly and i could see some reliability issues possably creeping in. Also tidied up all the cabling and added a 3 pin plug for the motor connection.

Also added regen to the system activated by a second button. Regen is relativily low at 20 amps but on a low powered setup like this, it is suprizingly effective. I returned 40 Whrs to the pack on the first full test. On a high powered bike, 40 Whrs is nothing however, when using 120 Whr battery, 40 Whrs is huge.
 

Attachments

  • 20181012_180452.jpg
    20181012_180452.jpg
    287 KB · Views: 3,550
That looks amazingly clean,John.
I understand why you recommended the 7S,just wondering what the repercussions of using a 5S would be,since I would like to keep rpm as low as possible. Not too worried about getting a big amount of power out of the system .
Any good websites out there regarding using BLDC motors that you would recommend?
 
Kepler said:
FZbob,
I have done probably 20% climbs. With grip tape, no problems. It will climb the same as your RC mid drive setup.

That's good news! I saw that you used 120 grit belt sander belt material. How did you secure it? Did you do a straight cut joint parallel to the axis of the motor, or a diagonal cut to prevent peeling? Have you ever tried the black self adhesive grip tape material which is used on stairs and ramps?

I'm still hoping that my Slaughter tire might work, as the center "knobs" are very low and close together (and half worn out)...
 
FZbob,
You really need to use belt sander material as the cloth backing gives you some longevity.The grip tape is something like 80 and will chew your rubber up very quickly. I use 180,but I would have went finer if I could have sourced it quickly.I initially tried 240 wet and dry paper to see what difference it would make, and it was night and day. Obviously that paper didn't last so when the belt sander paper arrived I attached that.It had the same perceived grip, only a little noiser on the road bike. Won't matter on a mtb. I used 3m double sided tape and stuck the paper the full width of the can(for more contact area) and it hasn't shown any signs of lifting after multiple wet rides. I cut mine straight across ,but cutting on the shear makes sense.Although you can't perceive the joint via the tyre.
I have used mine through very wet conditions and it works great, albeit a little noiser and obviously more wear is too be expected.
Over here in Scotland the drive was proving almost useless if there was any signs of a moisture....and that's almost all the time here :)
 
urga said:
Fzbob,how did you overcome the 185rpm at the cranks? Or is much less when under load? I'd like to try this with less reduction if possible.
I pedaled really really fast! :D Just kidding!

I typically spin 40-80 RPM, mostly around 50 or so. I originally ran it on 3S, but the torque curve was like a diesel. At lower cadence there was a lot of torque, but as i spun faster the power dropped off. So instead of downshifting and spinning to climb as I normally do, I had to let the pedal RPM drop and then mash pedals. About 370 watts maximum.

I switched to 4S and the torque curve is very flat. This feels very natural to ride! About 470 watts maximum.

This graph may be helpful... (hopefully I got it right...)

BLDC Torque Curve c.jpg
 
That's really helpful,thanks
So are you running 4S2P on 18650's or are you running a Lipo pack? If you are what assist time are you getting?
I like to use the assist very little ,and like you,I like to spin up climbs.I've bought some 26650 cells which are rated at 5000mAh which I plan to fit inside a waterbottle along with the min vesc etc. I would like to make it as small as possible(less capacity) but don't want leave it at 5000mAh and up the V's because of the no load rpm. So would rather go up to 10000mAh and drop the V's .Hence the low KV 63mm motor.You can tell I don't know what I'm talking about....John's been pretty patient with me,but I think I'm wearing him down :)
 
urga said:
That's really helpful,thanks
So are you running 4S2P on 18650's or are you running a Lipo pack? If you are what assist time are you getting?
I like to use the assist very little ,and like you,I like to spin up climbs.I've bought some 26650 cells which are rated at 5000mAh which I plan to fit inside a waterbottle along with the min vesc etc. I would like to make it as small as possible(less capacity) but don't want leave it at 5000mAh and up the V's because of the no load rpm. So would rather go up to 10000mAh and drop the V's .Hence the low KV 63mm motor.You can tell I don't know what I'm talking about....John's been pretty patient with me,but I think I'm wearing him down :)
I'm running a Turnigy 4S 5800 Lipo pack. Yesterday I did a 12 mile ride with 1500 feet climbing, and I reached the low voltage cutoff at the last hill at the end of the ride. I'm thinking about an 8000 pack.
(By the way, I'm a bit concerned that I'm hijacking John's thread. Can you ping me over on my thread?)
 
Kepler said:
Also added regen to the system activated by a second button. Regen is relativily low at 20 amps but on a low powered setup like this, it is suprizingly effective. I returned 40 Whrs to the pack on the first full test. On a high powered bike, 40 Whrs is nothing however, when using 120 Whr battery, 40 Whrs is huge.

I just caught that you now have the motor fixed against the tire! This is very exciting for me, as I was concerned about motor bounce on bumpy trails. Also, the regen option is very cool. At 20 amp regen, how much braking force do you get? Lightly dragging the brakes, or almost skidding?

How far is your motor pressed into the tire? Do you have any feel for more pressure creating more drag or not?
 
Yes the motor is constantly in contact with the tire and adjusted not to slip with up to around 350 Watts. At that tension the drag is minimal when free wheeling. I then back off the power to 250 Watts which for me is all the assist I need. Anymore and I feel like I am getting pushed into corners. With the constant contact on the tire, the assist is instant but subtle. Also zero bump or rattle with this setup which on a mountain bike is particularly important I think.

I am still experimenting with the amount of regen I want to use but -20A feels pretty good so far. I have also programmed a ramp in the Arduino interface so that when regen is activated, it ramps from 0% to 100% over a one second period. The regen at this level certainly slows you down and feels like you are lightly activating the brakes. Again for me, its all about subtle intervention.

The regen has huge potential on this setup when used on tight technical single track where you are using the brakes constantly. Not so much in a commuting role especially if your commute isn't particularly hilly. One thing I did find is that to get the most out of the regen, it must be activated by your brake lever. I have it on a second button and found I was constantly missing opportunities to use the regen simply because I was too busy concentrating on the decent. Even with the reduced opportunity to use the regen, I still managed to increase my range by 15% and I recon if I regened every time I touched the brakes, that would be more like 30% or 40%

Now working out how I am going to attach the regen to the brake levers. I don't want to use micro switches as they can be unreliable and are not water proof. That leaves Hall sensors as the next option. Complicates things a little as these need a 5V, Gnd, and signal wire rather then a simple on/off like a micro-switch. I plan to move the Ardunio interface to the handlebars and mount in a small junction box under head stem.

Another project to keep me entertained :)
 
urga said:
Yeah,no worries. Sorry John

No worries, its actually been an interesting and useful comparison. Concepts have a bunch of similarities

A quick summary

Friction drive Pros
--Low mechanical complication
--Very stealthy.
--Quite operation
--Minimal impact on the way the bike rides when not under assist.
--Regen capabilities and subsequent extended range
--Ability to keep the bike under power while not pedaling.

Friction drive Cons
--Low powered application only
--Can not use an aggressive rear mountain bike tire
--Tire pressure needs to be relatively high for reliable tire contact.
--Accelerated tire wear. (double best case)
--Loose a little bit of ground clearance
--Could clog up in wet muddy conditions

single reduction RC bottom bracket drive Pros
--Minimal impact on the way the bike rides when not under assist.
--Can use tire of choice
--Can run what ever tire pressure suits you.
--No accelerated tire wear
--Can run in muddy conditions

Single reduction RC bottom bracket drive Cons.
--Not very stealthy in appearance
--Added mechanical complications
--Possibly a little noisey due to the chain reduction
--No Regen capabilities and subsequent extended range
--No ability to keep the bike under power while not pedaling.
--Reduced ground clearance due to the 90 tooth reduction gear
 
Some bb mid drives can be stealthy, just not completely. This is what I'd do for a reasonable sized bb motor.
Find/Build the largest crank protector you can, preferably with lots of arms spreading out. Perhaps a 60T size protector.
You'd have to eyeball it for the look you want in camouflaging the motor, see the pic below.
Or just all black - Pedals, Arms, Gears, Protector, Motor, Chain, Wiring, Housings.
For the other side of the motor, you'd have to paint it some kind eye fooling pattern.
Cant get away with that rc noise though.
 
Kepler.

I'm hoping you might be able to help.

I've got a flipsky 6,6 vesc running the latest firmware, in my E-bike (DD hub) using 2 hall sensor throttles on ADC 1 & 2 for brakes and throttle. (No reverse). I'm using the flipsky nrf51xxx Bluetooth adapter wired into Rx and TX.

I'm trying to connect the android version of the vesc tool (I think it's v0.95) I see the bluetooth adapter and it initally connects but then gives the error "could not read firmware version..."

I've tried changing the baudrate from the default to 9600. And tried switching the Rx & TX cables on both baudrates encase I got that wrong. The error was the same.

Have I missed something. Other than not being able to use bluetooth for real time monitoring and changing setting. The vesc is working flawlessly.

Cheers
 
John,

Isn't that taken care of by buying a crankset with a freewheel ?
not ever having ridden or even seen a bike with a freewheeling crankset it really is a question for me.
Since I ride for my primary way of transportation, ( no operating car at this time ) I often ride without peddling . So I need a drive that will operate sometimes with only the throttle to give my legs and cardiovascular system a very much , Needed, rest .

Kepler said:
A quick summary

--No ability to keep the bike under power while not pedaling.
 
' if I regened every time I touched the brakes,'
Seems logical,but what if freewheeling on a long descent ? Would you have to feather the brakes to activate it? Not ideal as you are wasting speed,pads and regen energy by applying the brakes,even lightly. Or are you planning on having the sensor activate before the actual biting point? Even so,it could be a balancing act if the terrain was a little lumpy.
That's a good idea moving the Ardunio to the stem similar to the Di2 systems.
On my gravel bike,I'm planing on mounting the motor hard up to tyre as well,but was thinking it would be nice to engineer a way of backing it off from the tyre,when say you wanted to freewheel or if the battery died etc. I think that one of the good things about your previous setup was it could be setup to disengage when not under power. Of course you would lose out on the new kid on the block...regenerative braking.
Thinking back to your earlier designs ,did you not have a cable actuated method of engaging and disengaging the drive? If that could be made to work better, you would have the best of both worlds.
 
Kepler said:
Friction drive Pros
--Low mechanical complication
--Very stealthy.
--Quite operation
--Minimal impact on the way the bike rides when not under assist.
--Regen capabilities and subsequent extended range
--Ability to keep the bike under power while not pedaling.

Friction drive Cons
--Low powered application only
--Can not use an aggressive rear mountain bike tire
--Tire pressure needs to be relatively high for reliable tire contact.
--Accelerated tire wear. (double best case)
--Loose a little bit of ground clearance
--Could clog up in wet muddy conditions

single reduction RC bottom bracket drive Pros
--Minimal impact on the way the bike rides when not under assist.
--Can use tire of choice
--Can run what ever tire pressure suits you.
--No accelerated tire wear
--Can run in muddy conditions

Single reduction RC bottom bracket drive Cons.
--Not very stealthy in appearance
--Added mechanical complications
--Possibly a little noisey due to the chain reduction
--No Regen capabilities and subsequent extended range
--No ability to keep the bike under power while not pedaling.
--Reduced ground clearance due to the 90 tooth reduction gear

I would also add to the Friction Drive pro's - saving 350 grams of chain and sprockets. If Regen is used, saving another 200 grams or so of batteries.

I think that single reduction mid drive may also be limited to low (ish) power, as I needed the 1/4 inch pitch #25 chain to get a reasonable driven sprocket diameter. I have run it up to 470 watts, but only briefly. At over about 350 watts, the chain starts to sound distressed. I generally try to operate around 80-200 watts, which was my goal. It works very well at that level. There is definitely a bit of chain noise, which is somewhat proportional to watts.

I did run an experiment for a few weeks with the freewheel configured for no pedaling. For my purposes it was way under powered (using 3S), and just worked much better as an assist, so I shelved it. This experiment was what turned me on to the possiblilty of running direct drive with no freewheel.

I think that not disengaging the motor may be a bit of a game changer. For Friction Drive, it allows for Regen, and that is big. For mid drive it just gets simpler and lighter. It sounds like in both cases, the engagement is much smoother and more natural.

I was saving your Friction Drive kit as a future project for my other bike, but I think I may convert my current E-bike much sooner. The electronics are already in place, so I just need to mount your kit to my swingarm.
 
urga said:
' if I regened every time I touched the brakes,'
Seems logical,but what if freewheeling on a long descent ? Would you have to feather the brakes to activate it? Not ideal as you are wasting speed,pads and regen energy by applying the brakes,even lightly. Or are you planning on having the sensor activate before the actual biting point? Even so,it could be a balancing act if the terrain was a little lumpy.
That's a good idea moving the Ardunio to the stem similar to the Di2 systems.
On my gravel bike,I'm planing on mounting the motor hard up to tyre as well,but was thinking it would be nice to engineer a way of backing it off from the tyre,when say you wanted to freewheel or if the battery died etc. I think that one of the good things about your previous setup was it could be setup to disengage when not under power. Of course you would lose out on the new kid on the block...regenerative braking.
Thinking back to your earlier designs ,did you not have a cable actuated method of engaging and disengaging the drive? If that could be made to work better, you would have the best of both worlds.

The regen will will activate by the free movement in the brake lever before the brakes pads touch the brake disc. I don't think it will be too hard to feather on rough terrain. I am going to go with sensors on each brake lever with each leaver giving 50% of the available regen. This will give me 2 stages of regen braking. I think I will be able to increase the amount of regen to -30A if I have it setup with 2 stages.

Programming and software testing is done. Working on the hardware and mounting now.

20181015_072506 (2).jpg

I have built the drive with cable activation. It just ended up being an unnecessary complication. I have designed this mount to be quick removal if needed. it is held on by one 8mm bolt. Less then 20 seconds to completely remove the drive. I think the only time I would do this is if I smashed the drive up against a rock.
 
Back
Top