Trial riders need their clutch?

macribs

10 MW
Joined
Jul 22, 2014
Messages
3,702
Been thinking of a conversation last week for the past days. A co worker tried my bike and returned without the e-grin. He "kicked the tires" walking around and finally said electric bikes are fun but they lack any serious punch. First I laughed because I thought he was making a joke, turns out he was dead serious.

I told him hub motor like on my bike was not the best motor for trial riding, and that a mid motor would most likely give him what he needed. I told him about astro builds from Matt. Then about the gearbox astro from tangentdave, the LMX and the Neematic, and even hub motors as mid drives with jack shafts.

Few days later he had been reading up on the matter. Watched videos, visited websites and still couldn't grasp what e-bikers are doing things this way. I was thinking he meant he couldn't get enough torque to the wheel, but the torque was not enough he said. Because of the "spin up time", comparing electrical motors to turbo chargers. Then I remember the first comment as he was done with the test ride, the lack of punch.

He was actually missing the clutch. When I asked him he said he used to be a trial rider. And for his style of riding clutch control as well as throttle control is what makes vertical climbs possible.

Has anyone ever tried to do a clutch on a mid drive? Will a clutch make it easier to ride trial style over big roots, jump canals, climb almost vertical rocks and whatever else crazy stuff trial riders do every weekend?

Would a rekluse style clutch do the trick you think? The more I have thought about this the past days the more it seems like a worthy project to test. Not that I would ever ride crazy hard core trial style, but a little more punch at take off could be a good thing also for oldtimers.
 
For this train of thoughts, lets not focus on whatever losses a clutch with add. If a clutch eats away too much power, that can be solved with a bigger motor, or twin or triple motors. The added cost and added complexity is not a concern at this point. Now it is all about what can be achieved with the use of clutch for a mid drive, a left side mid drive with proper sprockets and chains.

Lets focus on what would be the benefits of adding a clutch to a mid drive. And would any style clutch work or would "clutchless clutches" like rekluse be worth considering.
 
I dont think clutches are needed for the ebike...i grew up riding atv and dirt bikes and they are extremely fun with gas vehicles, but i just cant see where i would want to ever use one on my bikes. Ive been ebiking for almost a decade and i find myself not really enjoying hub motor bikes anymore most of the time. I think he just needs a mid drive with alot more power with a properly tuned controller. They can climb anything and you wont need to downshift or clutch. How would the clutch benefit a bike with no gears?
 
macribs said:
but the torque was not enough he said. Because of the "spin up time", comparing electrical motors to turbo chargers

this sounds like the controller is stangling phase amps too much, Adaptto controllers and many other controller tend to do that. The old style EB3 infineon controller suffer much less from lack of punch down low, mainly because they don't really limit phase current.

No clutch here and certainly no problem tackling proper trials terrain.

[youtube]DaIY1je1rok[/youtube]
 
It's not the clutching/declutching that gives the "torque" the trials rider is speaking of, it's the ability to temporarily spin the rotating mass (flywheel/crankshaft) faster than the output shaft, therefore storing some energy.

This energy can then be released at will by engaging the clutch, dumping that stored energy into the output shaft.

But I agree with the comment above, if there's a "turbo" effect that causes the torque to increase as the motor speeds up, that's the controller limiting current at low speed. The nature of an electric motor is the exact opposite of that, maximum torque at zero(ish) RPM, tapering as it speeds up. All else being equal, this "curve" should be pretty linear until near the maximum speed.
 
I recall one incident where I needed the clutch to dump the ICE power instantly, and the throttle was not fast enough. An e motorcycle would have had the same problem.

For situations like this with e power it might be important to connect a clutch lever to an ebrake that would apply reverse power to the motor to cancel the differential momentum of the motor nearly instantly but not actually brake to a stop. Essentially emulate a clutch's behavior or something similar and even better. It requires taking just the right amount of energy out of the system, but this can be done, and very quickly.

Combine this with strong variable throttle controlled deceleration (which might be enough by itself). The goal is to make the electric machine more precisely controllable than the gas trial bike. It would require some special engineering but perhaps it has already been done by the trials bike manufacturers, or they will soon get there.

Interesting ideas for these applications.
 
For trials riding, it is important to have the ability to "launch" hard from a complete stop, to get the front wheel up and get as much speed into the bike as possible in half a bike length. What a clutch allows with an IC engine is to rev the engine with the clutch disengaged, store energy in the crankshaft and flywheel of the engine, then convert that stored energy into "instant" forward motion by engaging the clutch aggressively. The torque applied to the back wheel may be several times the maximum that can be applied by engine torque alone, albeit for a very short time.

No matter how much torque (within reason) an electric motor has, it cannot offer that "instant snap" in quite the same way.. that is probably what your ex-trials rider is missing.

However, advances in motors and controllers are narrowing the gap, and a French rider on one of those EM5.7 trials bikes has won a national championship...

Probably a slightly different riding style is needed for an electric trials bike.. the lack of clutch may be a difference rather than a disadvantage.

I think I saw a thread years ago where somebody (Gwhy maybe???) was experimenting with a clutch on a home-built or modified trials bike with an electric motor, but I can't remember details. I think it might be worth a try.
 
Drum said:
No matter how much torque (within reason) an electric motor has, it cannot offer that "instant snap" in quite the same way.. that is probably what your ex-trials rider is missing.
It could, if the electric motor is massive like the engine/crankshaft (like some of the motorcycle hubmotors, if used in the frame as a middrive, or if a flywheel were added to the motor output prior to the clutch. Overspeed the motor then pop the clutch....


Or without the clutch, if it were powered by a controller with no ramp-up, just instant pour-on-phase-current, that is capable of handling that without popping FETs or IGBTs or whatever it uses. And that the battery/BMS can handle the current without sagging in voltage too much, or overheating the cells/interconnects/etc.
 
It would be interesting to calculate the stored energy in the rotating mass of the gas engine and compare to the torque that an electric motor designed for the purpose could develop in a time similar to the clutch's ability to transfer the power. It would probably require a very high peak current controller, but it might be practical to produce even greater impulse acceleration than the clutch could transfer. One would have to go through the math to figure it out.

In any case, there are different requirements for different applications.
 
No clutch needed with the electric powerband, your bike just doesn't have enough power.
 
I'll just note that a rekluse clutch won't be of any benefit here. I had one on my DRZ. It's a centrifugal clutch system, so as the RPMs of the motor get higher and move towards the set point, it's like letting go of the clutch lever as the rekluse system engages the clutch pack for you.

More technically, they used to use ball bearings on a ramp, and now they use wedges. So the wedges are stuck between 2 rings, as these ring spins faster the wedges move outwards (centrifugal force), pushing the 2 rings apart, as this ring separates it puts pressure on the clutch pack (metal rings and friction rings) engaging the motor to the transmission. So there is no need to feather the clutch lever, it's all done with engine RPM. When the engine is idling, it's technically in neutral, when you rev the engine, it goes into gear (unless you are actually in neutral).

So then you just tune the system so that when the clutch lever isn't pulled, the engine doesn't stall, but is ever so slightly engaged and slips a bit. Which also lets you shift without pulling in the clutch, but you still have to let off the throttle so it goes back into "neutral".

Suffice it to say, you lose the ability to do things you would do by popping the clutch.

To mimic popping the clutch, you need a speed throttle type of system. So when you crack the throttle, the controller allows the motor to pull all the current it needs. With a sufficiently large motor, it shouldn't be all that much different, but it would probably be much heavier.
 
Alan B said:
I recall one incident where I needed the clutch to dump the ICE power instantly, and the throttle was not fast enough. An e motorcycle would have had the same problem.

For situations like this with e power it might be important to connect a clutch lever to an ebrake that would apply reverse power to the motor to cancel the differential momentum of the motor nearly instantly but not actually brake to a stop. Essentially emulate a clutch's behavior or something similar and even better. It requires taking just the right amount of energy out of the system, but this can be done, and very quickly.

Combine this with strong variable throttle controlled deceleration (which might be enough by itself). The goal is to make the electric machine more precisely controllable than the gas trial bike. It would require some special engineering but perhaps it has already been done by the trials bike manufacturers, or they will soon get there.

Interesting ideas for these applications.

I tip my hat to you sir, that might very well be one way to achieve it without using a clutch.


No matter how much torque (within reason) an electric motor has, it cannot offer that "instant snap" in quite the same way.. that is probably what your ex-trials rider is missing.

Hence the clutch. Let the electric motor "spool up", or excel from standstill into an RPM range where torque is already maximum. Dumping the clutch should give you all possible torque and power instantly.

I dont think clutches are needed for the ebike...
As a general rule of thumb I will agree, for specific usage to demanding riders clutch or some clever magic in software might just be what is needed to turn even the most dedicated petrol heads.

No clutch here in the video, and certainly no problem tackling proper trials terrain
I've seen a few of the EM 5.7 videos and if you watch closely you can hear riders are climbing differently then when doing the same on ICE bikes. They EM 5.7 riders keep the momentum going by not letting motor get to a halt, thereby having greater power off the line. But that might not be doable on all tracks and across all obstacles. I will take my sweet time watching those videos this weekend, and I will write down at what part you can see rider is working to keep momentum in order to clear obstacle.
 
flat tire said:
No clutch needed with the electric powerband, your bike just doesn't have enough power.

I know. But then again, my bike is not a trial bike. It is an everyday bike with just enough power to make you smile.
 
atarijedi said:
I'll just note that a rekluse clutch won't be of any benefit here. I had one on my DRZ. It's a centrifugal clutch system, so as the RPMs of the motor get higher and move towards the set point, it's like letting go of the clutch lever as the rekluse system engages the clutch pack for you.

More technically, they used to use ball bearings on a ramp, and now they use wedges. So the wedges are stuck between 2 rings, as these ring spins faster the wedges move outwards (centrifugal force), pushing the 2 rings apart, as this ring separates it puts pressure on the clutch pack (metal rings and friction rings) engaging the motor to the transmission. So there is no need to feather the clutch lever, it's all done with engine RPM. When the engine is idling, it's technically in neutral, when you rev the engine, it goes into gear (unless you are actually in neutral).

So then you just tune the system so that when the clutch lever isn't pulled, the engine doesn't stall, but is ever so slightly engaged and slips a bit. Which also lets you shift without pulling in the clutch, but you still have to let off the throttle so it goes back into "neutral".

Suffice it to say, you lose the ability to do things you would do by popping the clutch.

To mimic popping the clutch, you need a speed throttle type of system. So when you crack the throttle, the controller allows the motor to pull all the current it needs. With a sufficiently large motor, it shouldn't be all that much different, but it would probably be much heavier.

I think there could be benefits of having a clutch on e-bikes for some riders. Most of the time the clutch will just be there, then use it when you need it. Wonder how much a clutch would affect the performance hit, and how much less efficient an ebike with clutch would be over a similar bike without clutch.

AFAIK the rekluse now kind of works like a CVT. Which means you can alter take off RPM. You can get a user friendly take off, or if you do the re work of the rekluse correct you can a very direct take off. Years back we spend hours and hours finding the perfect CVT setup for out snowmobiles. Correct setup with hig RPM take off, you had to hang on to the bars not be stranded behind the accelerating snowmobile. Of course there are primary and secondary clutches on snowmobiles making it harder to fine tune things. And to make things even more complicated, you can alter the cogging in transfer case too. Until you pulled out all your hair. By then you hopefully got a setup sharper then anyone else. And then you discover your friends have gotten turbos....... :lol:

Tuning the clutches correctly you will accelerate faster then your friends with similar machines, you will climb higher when hi marking, and most likely they won't out run you on the lakes either. I've yet to take a rekluse apart but from what I've been told about them it seems to me they might be worth looking into. Maybe they have flywheel weights, springs or other things that can be manipulated. A little light sanding here, a stiffer spring here and before you know it bob is your uncle.
 
The clutch being required or not depends alot on the kind of riding you do, and your skill level. For 95% of riders, they can get along without it, as long as the motor is snappy.
However you're taking away an input for fine control, and if you have someone used to having that control, they will miss it. All Trials riders would fall in that category. some singletrack log hopping guys will also miss it for the same reason.
 
macribs said:
I know. But then again, my bike is not a trial bike. It is an everyday bike with just enough power to make you smile.
Yes, well that's the answer for your friend and the /thread, amidst all kinds of other irrelevant discussion. How much power does your bike have?
 
We're putting our electric bikes up against some pretty stiff competition here too guys.


Trials bikes are extremely light for the amount of power they have, 125cc 2 stroke motocross engines are a highly evolved animal, 30+ HP in a very tiny package along with a six speed gearbox and a multiplate wet clutch.

It's not going to be easy to match their power/weight ratio with electrics, not saying it's not possible but we're not talking about toy dirtbikes here. These machines have a multi decade and multi $Billion R&D head start.
 
flat tire said:
macribs said:
I know. But then again, my bike is not a trial bike. It is an everyday bike with just enough power to make you smile.
Yes, well that's the answer for your friend and the /thread, amidst all kinds of other irrelevant discussion. How much power does your bike have?

No it is not. Because power is plenty when the motor first gets some RPM's. I got a 205 in a 17" maxing out about 15 kw peak. But from dead stop it is "slow". Tried various controllers and was gonna swap to the mobipus but he went OEM :(

I am not comparing my pig heavy bike against a trail or mx bike. I'm just saying that power delivery is not perfect. And that was the point of the former trail rider. You can't ride technical if you can't trust instant power. Any delay will affect your riding. Even unlocked adapttos pushing 16+ kw peak is like that. A small delay.
 
macribs said:
flat tire said:
macribs said:
I know. But then again, my bike is not a trial bike. It is an everyday bike with just enough power to make you smile.
Yes, well that's the answer for your friend and the /thread, amidst all kinds of other irrelevant discussion. How much power does your bike have?

No it is not. Because power is plenty when the motor first gets some RPM's. I got a 205 in a 17" maxing out about 15 kw peak. But from dead stop it is "slow". Tried various controllers and was gonna swap to the mobipus but he went OEM :(

I am not comparing my pig heavy bike against a trail or mx bike. I'm just saying that power delivery is not perfect. And that was the point of the former trail rider. You can't ride technical if you can't trust instant power. Any delay will affect your riding. Even unlocked adapttos pushing 16+ kw peak is like that. A small delay.

That throttle mapping is one of the largest issues EVs have. too snappy and some people cant ride it. too slow and it sucks for those that want performance. and too often, the throttle response and feel is overlooked.
 
What you need is controller logic designed for your application.

I suspect trapezoidal controllers are better at this. In industrial controls we would use motors with real encoders on them and FOC to minimize the torque ripple.

The only delay that is inherent is from the ADC reading the throttle and the motor inductance affecting current ramp-up. These are far below (faster than) human response time. Reading the rotor position from hall sensors takes even less time, and these CPUs are running at more than 10,000 instructions per millisecond. Applying voltage for max torque can occur in far less than a millisecond. The inductive time constant for a Cromotor as an example is about 1.3 milliseconds, so full current at low speed could be forced in a millisecond since we have so much voltage surplus with zero back EMF. Even waiting 5 Tau would be under 7 mS.

These delays you are experiencing are from the software, it is not designed to these requirements.
 
I think this is why companies like Electric Motion and Oset use mid drives and put a lot of effort in setting up their controllers and drive train. I believe the Oset 24.0 has a 1.4kW continuous drive system but gives 8kW peak power for that clutchless "snap". It sure looks a lot of fun and I wish I could ride like the pro's:
[youtube]CHhlyMPs2CA[/youtube]

[youtube]REeeYtp-8qE[/youtube]
 
Urban hooligan riding

[youtube]i4AjRnqVqfU[/youtube]


Even hubmotors do jump

[youtube]TuwESVujaHg[/youtube]

[youtube]40KnynInACQ[/youtube]
 
The videos demonstrate the importance of the rider capability versus bike capability. Riders can have multiple bikes for each task but real skill is evident in one rider performing on a single bike. See the road racing bikes used for trials type jumps and balancing on youtube.

In MotoGP rider adaptability is critical in adapting their style to new machines with different engine, clutch, electronics and chassis characteristics.

A clutch is beneficial in specific situations but if power and torque are equal rider skill is much more important, and adaption of rider skill is the single most important aspect.

The calibration of gas bikes has advanced with electronic control. If we or OEMs emulate this calibration a clutch is not the issue.
 
Trials isn't racing it's like a stunt / obstacle course and speed in unimportant. The bikes just need to be light and very agile, don't need to go very fast, and I have complete faith that a electric drive system with sufficient power density can provide the instant torque that trials courses need, but without a clutch.

Also, yes most cheap bike controllers have a ass-ton of throttle lag HOWEVER this doesn't have to be the case: as noted above throttle response can theoretically be almost instant.
 
Back
Top