Observed trials bike designing

[Josh K.]

Hi Folks, Very interesting thread.

I am a fan of the idea of using a clutch on the light electric motorcycles. I recently went to Moab Utah, and rode the slick rock
trail again, but with a 350cc motorcycle instead of a traditional pedal mountain bike. One interesting thing of course is to see
the E bikes being used out there too. I saw a pair of Segways, and can imagine the Sir Ron crowd totally digging the miles of skate
board park like area to ride.

But, what I learned is that the motorcycle guys are using the clutch to offer some slip, when climbing. When we have about 40 hp, and the throttle is at about 1/3rd, you can climb walls out there. The traction is amazing, especially when tire pressure is down
at around 14 lbs. But, sometimes we just did not have enough hp to keep it going up the hill, and if one were to add throttle, the front wheel would come up, and scare you. So we learned to feather the clutch up top, and it smoothed out the climb. It was nothing short of amazing. I loved it.

I kind of think of it as a slipping gear reduction, that gave us more control. I feel that it could also be very useful in ramping up
the torque, as we transition from very flat, to very steep, and need to launch up onto the hill climb. Which brings me to hill climbs.
I think that really steep, big hill climbs could be really fun for the big boy motors, where we could really let them run, full tilt.
I used to run an E tek, at 50 volts, and with a 500 amp controller. But could never get the traction right for a full launch. We would leave 80 feet of rubber, before hooking up completely, or would spin the rim in the tire...But, I would love to build up an old school dirt bike with and extended rear triangle, for hill climbing. And, I bet it would be advantageous to be able to feather the slip, when digging up a monster hill. Something that would empower the rider to be able to modulate the power a little more.

Additionally, It is my belief that the full on road motorcycles are going to grow up, and use a 6 speed with a clutch as well. I was one of the first to buy a Zero MX motorcycle, back in 2008. And it was a great minibike. Yeah, I said it, minibike. Fun for around mom's house type of thing, for the teens. It did not have awesome take off, or top end. Sure, it was really fun from 15 mph to 45 mph, where it was in it's sweet spot. But it totally could have used at least a two speed, with a clutch to dump the power hard, and to really throw rocks. People love the electric motor, and it's raw power on take off, but we could have more, where we let the motor spin up a little bit, then dump both the clutch and the throttle for the best power, ideally in a lower gear, then up through the range, for ultimate speed.

I run a 1,500 watt mid drive through my bicycle drive train, and love to use all of my gears. Sweet hill climbing, super efficient when I want to be, and nice high end for passing roadies. Sure, I use a 9 speed, but that is because it's easily obtainable in the bicycle world. My new bike is using an 8 speed internal hub, with a 1,500 watt Astro flight build. And, I can say that after 10 years of using this system, if I were to be building up a motorcycle, I would totally want a 6 speed or so, with a clutch, to blast the power through the system.

Soo, I am tinkering with the idea of building up a trials like motorcycle, to ride in Moab, with an E tek and am looking for a gearbox clutch combo to set up between my motor, and rear wheel. Ideas? Thanks so much, Josh K.

 

[bikerpete]

MountainClimber2M.JPGHi Folks, Very interesting thread.

I am a fan of the idea of using a clutch on the light electric motorcycles. I recently went to Moab Utah, and rode the slick rock
trail again, but with a 350cc motorcycle instead of a traditional pedal mountain bike. One interesting thing of course is to see
the E bikes being used out there too. I saw a pair of Segways, and can imagine the Sir Ron crowd totally digging the miles of skate
board park like area to ride.

But, what I learned is that the motorcycle guys are using the clutch to offer some slip, when climbing. When we have about 40 hp, and the throttle is at about 1/3rd, you can climb walls out there. The traction is amazing, especially when tire pressure is down
at around 14 lbs. But, sometimes we just did not have enough hp to keep it going up the hill, and if one were to add throttle, the front wheel would come up, and scare you. So we learned to feather the clutch up top, and it smoothed out the climb. It was nothing short of amazing. I loved it.

I kind of think of it as a slipping gear reduction, that gave us more control. I feel that it could also be very useful in ramping up
the torque, as we transition from very flat, to very steep, and need to launch up onto the hill climb. Which brings me to hill climbs.
I think that really steep, big hill climbs could be really fun for the big boy motors, where we could really let them run, full tilt.
I used to run an E tek, at 50 volts, and with a 500 amp controller. But could never get the traction right for a full launch. We would leave 80 feet of rubber, before hooking up completely, or would spin the rim in the tire...But, I would love to build up an old school dirt bike with and extended rear triangle, for hill climbing. And, I bet it would be advantageous to be able to feather the slip, when digging up a monster hill. Something that would empower the rider to be able to modulate the power a little more.

Additionally, It is my belief that the full on road motorcycles are going to grow up, and use a 6 speed with a clutch as well. I was one of the first to buy a Zero MX motorcycle, back in 2008. And it was a great minibike. Yeah, I said it, minibike. Fun for around mom's house type of thing, for the teens. It did not have awesome take off, or top end. Sure, it was really fun from 15 mph to 45 mph, where it was in it's sweet spot. But it totally could have used at least a two speed, with a clutch to dump the power hard, and to really throw rocks. People love the electric motor, and it's raw power on take off, but we could have more, where we let the motor spin up a little bit, then dump both the clutch and the throttle for the best power, ideally in a lower gear, then up through the range, for ultimate speed.

I run a 1,500 watt mid drive through my bicycle drive train, and love to use all of my gears. Sweet hill climbing, super efficient when I want to be, and nice high end for passing roadies. Sure, I use a 9 speed, but that is because it's easily obtainable in the bicycle world. My new bike is using an 8 speed internal hub, with a 1,500 watt Astro flight build. And, I can say that after 10 years of using this system, if I were to be building up a motorcycle, I would totally want a 6 speed or so, with a clutch, to blast the power through the system.

Soo, I am tinkering with the idea of building up a trials like motorcycle, to ride in Moab, with an E tek and am looking for a gearbox clutch combo to set up between my motor, and rear wheel. Ideas? Thanks so much, Josh K.

There's not much in the way of gearbox/clutch available off the shelf that I'm aware of. I suspect the simplest route is do what GasGas did with their e-trials bike - cut the crankcase off the gearbox and graft in your eTek.

I really doubt that you need more than two or three gears tops for almost any e-moto. You're not nearly so constrained by the narrow torque & power rpm range.
I also suspect you'll find that you don't need the clutch slip on climbs anywhere near as much. I think the main reason you'd want it on an e-moto is just because it's so much faster/more responsive and yet more stable than twisting a clunky old throttle tube while you move over the bike to stay balanced.
The clutch would certainly go a long way to solving the spinning rear on take off - once the clutch is fully slipping engine RPM makes virtually zero difference to transmitted power. It's ONLY the pressure modulation to the plates that adjusts power delivery. So you could wind up the motor a bit then feed clutch and it'll not matter in the least if you send the motor RPM to the limit, power will remain almost exactly the same until you feed out more clutch. It just demands a finely calibrated left index finger.

I'd suggest considering just getting an ICE trials bike with seized engine or something. They're relatively lightweight (65-70kg), have a gearbox & clutch, heaps of traction with tubeless rears running 3-5 psi ("down at around 14 lbs" - my pressure gauge doesn't even go that high!). If you plan on a seat (sacrilege on a modern trials bike) then you'll have room for a decent battery, by the time you lose the cylinder & carby there's room for a fair size battery already.
Of course you could add an extended swingarm if you want to, but it's pretty wild what a stock trials bike can climb with a bit of technique.

 

[owhite]

There's not much in the way of gearbox/clutch available off the shelf that I'm aware of. I suspect the simplest route is do what GasGas did with their e-trials bike - cut the crankcase off the gearbox and graft in your eTek.

this may not be relevant but there's another very long discussion of the merits of gearboxes over here.

 

[motodojo]

I have 2 stroke and 4 stroke trials bikes and an Oset 24 and have been riding trials for over 40 years. I've been riding an electric mountain bike for over a year and one design feature from it that should be incorporated into an electric trials bike is the dropper post. Almost all the trials brands have a version with a seat but for pure trials the seat gets in the way. A design like the Mecatecno Dragonfly could have incorporated a dropper post.

Another avenue that hasn't been explored is an electric trials bicycle. If you watch Chris Akrigg videos you will see him doing trials on an eMTB but us mortals would need to ditch the derailleur and have a motor guard.

I wish my Oset had a clutch and this DOB Bad Sheep from China looks like it would be the next generation of the Oset 24 size platform. More power and a clutch/flywheel with the lightness of using MTB suspension. Hopefully they will be imported into the US at a reasonable price and not like the Dragonfly and EM which start at $10K

There doesn't seem to be much info about the DOB Bad Sheep available but the image on the DOB bikes Australia has the best looking one yet with the downhill forks, hefty skidplate, clutch cover and motorcycle tires. Some of the videos out there have some sketchy details.
https://youtu.be/8d4baUKBUDs

This one seems to have a rattly dry clutch.
https://youtu.be/0H9HWXMIR6g

 

[bikerpete]

I have 2 stroke and 4 stroke trials bikes and an Oset 24 and have been riding trials for over 40 years. I've been riding an electric mountain bike for over a year and one design feature from it that should be incorporated into an electric trials bike is the dropper post. Almost all the trials brands have a version with a seat but for pure trials the seat gets in the way. A design like the Mecatecno Dragonfly could have incorporated a dropper post.

Another avenue that hasn't been explored is an electric trials bicycle. If you watch Chris Akrigg videos you will see him doing trials on an eMTB but us mortals would need to ditch the derailleur and have a motor guard.

I wish my Oset had a clutch and this DOB Bad Sheep from China looks like it would be the next generation of the Oset 24 size platform. More power and a clutch/flywheel with the lightness of using MTB suspension. Hopefully they will be imported into the US at a reasonable price and not like the Dragonfly and EM which start at $10K

There doesn't seem to be much info about the DOB Bad Sheep available but the image on the DOB bikes Australia has the best looking one yet with the downhill forks, hefty skidplate, clutch cover and motorcycle tires. Some of the videos out there have some sketchy details.
https://youtu.be/8d4baUKBUDs

This one seems to have a rattly dry clutch.
https://youtu.be/0H9HWXMIR6g

The Kuberg Ranger has a small moto style seat (more like a saddle than a seat really) that hinges off the front. Lift it up and it becomes a cantilevered seat, or drop it and it blends into the trials style cutaway seat area. Pretty neat.
Personally I can't be bothered with a seat of any sort. I made an easily removeable seat for a bike a few years ago but only used it a couple of times - I certainly don't miss it. The DOB have optional removeable seats - again, I never bother. Just more weight and clutter as far as I'm concerned.

I've also considered building an electric trials bicycle - that could be a very cool toy! I hadn't decided if it would have pedals or not. Potentially ditch the pedals and have a super low geared motor (maybe 10kmh top speed) in place of the bottom bracket, a bit forward so the pegs can be in the correct place. Trying to combine a freewheel and high torque, on/off style motor use seems unlikely to end well.
Very small & light battery. Plan on it having about the same range as a comp trials bike - ie almost none. Just sufficient to spend an hour or so doing pure trials moves.
Such a bike couldprobably get away without a clutch due to the extremely low gearing and light weight. Potentially you could setup the rear brake lever so it had a switch at fully out which engaged whatever throttle setting you've got - like the e-clutch on the EM's. You could lock the back wheel like a trials bicycle and hop etc, then with a quick release of the brake it would give a flick of power for gaps etc. The controller would have to have virtually zero ramping, which would make smooth throttle control challenging, but that might not be such an issue for bike type trials where you aren't really travelling any distance.
No suspension so the chain could be fairly tight, which would allow a beefy rear disc to stop it almost instantly.
A low inertia motor might be an advantage.

FWIW I am the owner of DOB Australia.

And how about this one for some e-mtb trials skills https://yewtu.be/watch?v=B04WDoFc1p0
I just read the comments to this video - surprised me that no pne who commented seems to know that Toni was junior world MTB trials champion before moving to moto.

 

[motodojo]

Here's some impressive eMTB trials riding.
https://youtu.be/GG0cpyrkXjg

Tutorial on riding steps.
https://youtu.be/otb7IQvqpnA

 

[motodojo]

Here's my street legal trials bike with removable seat. It comes off in seconds with dzus fastener and weighs almost nothing. I ride a few miles on the street to my riding area then take it off and use the same seat on my Oset. With an electric motorcycle since there is no exhaust or airbox it would be possible to have a dropper post. I like to sit when trail riding to rest my legs.

 

[bikerpete]

Well after a lot of contemplation, discussion and consideration I have finally been getting time up on an e-trials with clutch.
It pretty much confirms a lot of what I had anticipated, which is mildly gratifying.

The new dob "Bad Sheep" has a 3kW continuous rated motor, a flywheel and multi-plate clutch (dry). It also idles in order to totally eliminate any startup delay and to give instant access to a base level power at all times.
I'm really very impressed with what they've built.
But the topic here is about the clutched experience.

Basically it is transformational compared to riding un-clutched (of course it's easy to compare - just don't use the clutch), much as expected.
The speed & control with which you can engage and disengage power is vastly improved, which enables some significant techniques.
There's a bit of learning to be done to work out which situations really call for clutch and which can as well or better be tackled on throttle alone. With the quick throttle response there are some things that would call for clutch on an ICE bike that can be performed perfectly well on throttle on the e-trials.

One thing that it highlights to me is that to really get the response you need for trials the gearing has to be low. It's not totally about gearing down for more torque, but as much about having the acceleration within the useful range. If the gearing is too high then it's far too easy to accelerate (off clutch or throttle) to a much quicker speed than you want. Often you want to accelerate, but from say 10kmh to 12kmh. You want the increase in torque and power, but you don't really want to pick up much speed. With higher gearing that can be really hard to manage, especially on a light bike such as this.
If you accelerate hard, the bike can pick up too much speed very quickly, then you roll off the throttle and suddenly the power is gone. What you want is a large increase in torque, for say a step in amongst a bunch of rocks, and a slight increase in speed.

The other thing that is quite evident, and expected, is that high gearing just destroys much of the benefit of the flywheel. Either it can't deliver enough instant torque, or more commonly all the inertia gets used in the initial acceleration and there's nothing left to drive the wheel up the obstacle. Lower gearing consumes less energy in the initial acceleration so there's something left for later.

These two also combine - sometimes you drop the clutch at high RPM and the bike just rockets away. It can be useful sometimes, but other times the back wheel just picks up too much speed and hits obstacles traveling too fast, causing it to bounce or slam in un-controlled etc.
In trials you mostly want really fast acceleration within a quite narrow speed range.

Frankly it's a madness to think that putting gearing that can pull 50-60kmh+ on a single gear trials bike is going to work. It doesn't as far as I'm concerned.
It's like riding an ICE trials bike permanently in 4th or 5th gear - you can do it no problem using the clutch, but it's not going to work anywhere near as well for most activities as 1st or 2nd.
I believe the EM falls victim to this, and it's probably the main reason many top level riders find it just a bit underwhelming for really generating big lift. If it was geared down it would probably splat and zap almost like the ICE bikes, but then it would run out of puff at the top end, exacerbated by the typical characteristic of electric motors to have max torque at low RPM, reducing as RPM rises.
Gearboxes are good inventions, alongside flywheels. It's disappointing Dragonfly decided to ditch the gearbox for their production bike.

Personally I'd prefer a bit more inertia in the flywheel, something to carry the bike a bit more when you cut the throttle. Sudden changes in motor speed can make staying balanced pretty tough sometimes. With a relatively low inertia flywheel and fairly high gearing the flywheel can't smooth this out enough.
I've yet to put my Nucular on this bike, but I'm anticipating that having some control of the RPM rise and fall rate might make a decent difference in this area. It will somewhat simulate aspects of a flywheel, while still allowing a real flywheel to do it's thing.

I think a flywheel and clutch, combined with a controller that can control RPM rise and fall rates (actual RPM control) separately from torque ramping would make a pretty wicked combination. Set the RPM ramps fairly slow and the torque ramp really fast, let the flywheel and clutch provide instant controlled delivery & do a bit of smoothing and things would be pretty sweet.

Going back to riding clutchless is like going back to a kids toy compared to a real bike!
I'm sure there are plenty of people who will have absolutely no need or use for a clutch. I'm just not one of them!

Maybe sometime there'll be a controller with the smarts to emulate a mechanical clutch and flywheel, but it's a pretty niche bit of gear. Or else it'll be a something that's been designed to have ultimate flexibility, probably really complex to program and get all the factors tuned.
Until then I'm totally sold on a flywheel/clutch combo.

 

[speedmd]

Niche programing for certain. Torque curves would need some perfect ramps dialed in to benefit much. One thing, a huge benny is, you could just add a clutch in line and use it or not without the dreaded ice motor stall. Thinking, as long as the motor is geared low enough to match or exceed the pulse the flywheel-clutched ice setup is delivering, with a reasonable amount of current control, the clutch would be optional. Would be interesting to see a few different motor types used for comparisons on how they would be to control on spin up. Traction control? 8) Some of the high end axial type motor designs may be worth a look for O. trials apps.

 

[bikerpete]

Niche programing for certain. Torque curves would need some perfect ramps dialed in to benefit much.

What I've been experiencing on this bike is that the torque curve is considerably less relevant than the RPM curve once you've got a clutch.
I think you could have extremely high torque slew rates as long as the rpm slew rates were slow enough that you can respond to the accelerations.
When dropping the clutch for dynamic moves you really want instant torque changes, but you don't want big instant speed changes. Same in many technical sections.
That kind of fits with the torque & power maps out of ICE engines, but with benefits of a much flatter curve.

One thing, a huge benny is, you could just add a clutch in line and use it or not without the dreaded ice motor stall. Thinking, as long as the motor is geared low enough to match or exceed the pulse the flywheel-clutched ice setup is delivering, with a reasonable amount of current control, the clutch would be optional.

I'm not quite sure what you're suggesting here. It seems like you're suggesting that with sufficient current control you could do away with the clutch?
That's just not true, it is absolutely physically impossible to do what a clutch enables on a twist grip alone. Even an electronic "clutch" is very far from a real clutch.

Would be interesting to see a few different motor types used for comparisons on how they would be to control on spin up. Traction control? 8) Some of the high end axial type motor designs may be worth a look for O. trials apps.

Could be interesting, but I feel like the motor isn't a big deal. Even a pretty ordinary motor can probably provide all the power and control you'd need on a trials bike.
I really think it all comes down to getting a really good understanding of the finer details of trials power delivery needs, then setting up a controller to match those needs.

It would be awesome if some engineering PhD student did a research project on it

 

[speedmd]

One of these is most likely way overkill. https://emrax.com/e-motors/emrax-208/

140 nm peak and if geared 10 : 1, would be some serious instant torque. Not sure how your clutch grabs, but my relatively good condition sherco would be no match for these torque levels from my experience or going by my math. Control? Not so certain how to go about setting it all up. Most likely, you would not need to rely on clutching for a much larger range and portion of segments.

 

[bikerpete]

One of these is most likely way overkill. https://emrax.com/e-motors/emrax-208/

140 nm peak and if geared 10 : 1, would be some serious instant torque. Not sure how your clutch grabs, but my relatively good condition sherco would be no match for these torque levels from my experience or going by my math. Control? Not so certain how to go about setting it all up. Most likely, you would not need to rely on clutching for a much larger range and portion of segments.

That's certainly some impressive specs on the emrax - no doubt an equally impressive price too!

I'm not sure you're following my argument about torque vs control, or perhaps you don't agree, which is fine of course.
It's completely pointless having massive torque if you haven't got it under total and virtually instant control. Twisting a throttle tube just doesn't cut it - it's slow, it's impossible to control accurately as the bike moves rapidly into all sorts of exotic positions under you and it's often impossible to instantly cut power both when things are in control and also when things get out of control. The idea of whiskey throttling with a motor like that Emco and no clutch just doesn't bear thinking about!

There is also the separation between torque and RPM to consider. This is something I've become much more aware of with the clutched dob bike. It's really struck me how often I want an instant increase in torque, but not much actual acceleration. A clutch and low gearing gives you that quite nicely.

I do think there's possibly a difference in riding styles at play here too.
Out of interest @speedmd, how comfortable are you twisting the Sherco throttle to the stop, waiting for it to wind out until it 4-strokes, probably around 9,000 rpm or so, then dropping the clutch - from stationary or just creeping along?

That's the sort instant on & off impulse I'm talking about and aiming for. I am confident that trying to emulate that without a physical clutch is going to end very badly - Firstly trying to snap the throttle open fast enough that the bike doesn't accelerate forward too much as the RPM builds, then just the inertia in the motor itself is going to carry the bike waaay past vertical, let alone trying to snap the throttle closed instantly right as the bike explodes forward and up and you rotate behind it.
I just can't see that ending well at all! To say the least.

With a clutch this little move is certainly exciting, but it's totally manageable and gets used frequently.

 

[speedmd]

I totally get what your saying about the instant slip out of the clutch when landing from a hop or max power needed to get up some huge steps from a stop. Something like the axial I noted Would need some sort of finger or slide lever to control max current allowed by throttle. The ele motor, as great as it is at a stall in generating torque, has virtually no power at zero rpm. It needs to wind up to generate some hp. If too slow to wind up, then a clutch would certainly be a good way to accomplish this if needed.

The current limiting control could default to a reasonable current (torque) level and have a half dozen or so steps with increasing max current settings. In the default setting, just enough torque is made to lift the front controllably for slow speed maneuvering without darting away as you can do now with clutch. Thinking the throttle would be good to have a zero power detent position and also a braking - regen region unwinding it 15 -20 degrees as well as the standard function when winding it on. Have to admit, something like the axial super motor example is something that would be interesting to see develop a bit further. Interested in the new motor koenigsegg has come up with also. Smaller and cheaper also. https://www.autoevolution.com/news/koenigsegg-explains-how-it-created-the-raxial-flux-motor-180689.html

BTW, I love doing 3 and 4th gear burnouts with the sherco. Got the clutch feeling much much better than when I first got it. Still playing with the jetting a bit but much better than stock setup now. A 250 would be a better size for me.

 

[bikerpete]

I totally get what your saying about the instant slip out of the clutch when landing from a hop or max power needed to get up some huge steps from a stop.

I know I use the big power moves as examples, but the same applies to much smaller moves - it's the level of control that's key.

Something like the axial I noted Would need some sort of finger or slide lever to control max current allowed by throttle. The ele motor, as great as it is at a stall in generating torque, has virtually no power at zero rpm. It needs to wind up to generate some hp. If too slow to wind up, then a clutch would certainly be a good way to accomplish this if needed.
The current limiting control could default to a reasonable current (torque) level and have a half dozen or so steps with increasing max current settings. In the default setting, just enough torque is made to lift the front controllably for slow speed maneuvering without darting away as you can do now with clutch. Thinking the throttle would be good to have a zero power detent position and also a braking - regen region unwinding it 15 -20 degrees as well as the standard function when winding it on.

I think we just have to agree that we have quite different perspectives on how we want to ride.
What you're talking about here is exactly what I'm trying to get away from.

The idea that we need to simply manage current is very far from what I want to achieve.
After spending time riding what is really quite a competent clutched trials bike, I'm even more aware that controlling RPM is at least as important as controlling torque, if not more so.
One of the absolute worst characteristics of all the electric bikes I've ridden is that as soon as the tyre looses traction the motor instantly spins up, making a bad situation worse. Yes, you could try to develop traction control, but I've no clue how you develop an automated traction control that can discriminate between the very fast wheel acceleration from eg deliberately doing a radical wheelie (the high rpm-drop the clutch style) and the wheel slipping at low power settings, particularly when so often the front wheel is in the air, so no use as a reference.
This is a "feature" of controlling current alone, rather than current and rpm.

Having a braking region past throttle zero would be a nightmare! I sometimes rotate the entire throttle assembly on the bars when extending really strongly up through the bars and trying to lift the bike with me, on vertical steps for instance, even with a pretty securely bolted Domino throttle. The throttle tubes usually die because the idle stop gets broken from being slammed into.
Having the bike go into regen when that happened would be a disaster.
I think regen is probably best managed with maybe a thumb lever, or perhaps the first couple of mm of travel in the rear brake. I've even wondered about using a pressure transducer in the rear brake line, but I think I'd prefer to be able to use regen without any mechanical braking involved at times eg long downhills.

Having virtually any delay in motor spin-up/down is not ideal (without a clutch) - good riders are varying the power to the wheel continuously at well less than 0.1 second intervals - that's actual delivered power. Varying the power at the motor does not translate directly to varying the power at the wheel due to inertia of the motor itself - in light motors it's close certainly I accept.
What you mention about developing some RPM in order to develop power is one of the reasons the better e-trials bikes now idle and use a clutch.

Have to admit, something like the axial super motor example is something that would be interesting to see develop a bit further. Interested in the new motor koenigsegg has come up with also. Smaller and cheaper also. https://www.autoevolution.com/news/koenigsegg-explains-how-it-created-the-raxial-flux-motor-180689.html

BTW, I love doing 3 and 4th gear burnouts with the sherco. Got the clutch feeling much much better than when I first got it. Still playing with the jetting a bit but much better than stock setup now. A 250 would be a better size for me.

A fancy high power motor would certainly be an advantage - more power & torque, less weight, smaller ... But it still comes down to the level of control that will make it work or not.

 

[speedmd]

I do get your issues with the added complexities - range of the twist throttle venturing into regen or hard braking when looking for a safe landing position. Not sure if the twist throttle setup is best for standing - while holding on for dear life and controlling power delivery at the same time. Don't recall ever getting into the whiskey throttle effect while riding the snowmobile. Easy to do on the twist throttle when one makes the rookie mistake of gabbing the throttle with a bit too high a wrist position. Traction control could "look" at ground speed as it does on some car systems. At some point it all gets much too complex and become much like a F1 car with hundreds of setup numbers to switch in and out. Finding a relatively simple optimized electronic control setup that is adaptable for most riders is key to pushing this forward.

 

[bikerpete]

I do get your issues with the added complexities - range of the twist throttle venturing into regen or hard braking when looking for a safe landing position. Not sure if the twist throttle setup is best for standing - while holding on for dear life and controlling power delivery at the same time.

If you consider the throttle as "Available power" & clutch as "Delivered power" then a twist grip is fine, and allows for all sorts of weird and wonderful position relationships between rider an machine, far more than a thumb throttle.

Traction control could "look" at ground speed as it does on some car systems. At some point it all gets much too complex and become much like a F1 car with hundreds of setup numbers to switch in and out. Finding a relatively simple optimized electronic control setup that is adaptable for most riders is key to pushing this forward.

I can't even start to imagine a system that could deal with a vehicle that is intentionally rotated well past vertical, endoed up onto the front wheel, hopped along on rear alone, that gets ridden in mud, through rivers, on surfaces that move (sand, loose gravel, rocks that pull out of the ground) etc.
Lack of imagination on my part I guess.

One of the attractions of trials is that it's technically difficult. I'd probably lose interest if the bike handled traction control, could self balance on the rear and so on. That's the fun of it - developing these skills and getting the satisfaction of learning.
I know I'm riding a modern bike, not a pre-'65, because they help, but fundamentally if the rider puts in a bad input they get punished. I like that.
Just my perspective.

So I'm not interested in traction control, on-the-fly setting adjustments & so on.
I just want a machine that provides power & control in a predictable, versatile manner in a wide variety of surfaces, terrain & techniques.

Let's say I'm not a believer in adding technological "improvements" just because you can. Simple is good, but can be deceptively complex.

 

[speedmd]

Making a excellent functioning machine is easiest to do by continually improving it and adding features- ("complexity") as their needs and function develop. Re thinking it to a ideal "simplified" form is the more difficult part for most.

A good friend suggests often that "doing things the hard way is often the easiest". :lol:

No substitute for doing the skill work until you get it. In trials, if your not into crashing reasonably often, you will not develop expert skills. Not quickly at least. With that said, making improvements often make things much worse initially.

As far as active traction control goes, it would go in the pie in the sky (potential long term benny) thinking category. In the earlier days of trials, common knowledge was to add flywheel mass and de tune the motors to add better traction for slick conditions. Somewhat passive traction control. Even if the electric motor is better in most situations, proper "Mapping" / throttle control characteristics need to be well dialed in to take advantage. Eliminating the clutch may never really be viable in this app. I think that if you can setup a simple machine that can ride more segments or sections of segments with less reliance of constant clutching, that alone will be a significant improvement. May never happen but it is pie in the sky none the less for me at the moment.

As far as the revvy ele motor syndrome ( motor - wheel speed running away after traction breaks), we discussed earlier in setting up a speed throttle to help control this. It may be that simple, if you can add simple variable current limiting to it. The twist would control your rpm-limit - ramps and the current limiting would control, your torque limits - ramps. Each would have their slow to fast adjustments and power limiting functions. As you mentioned before, a 125cc is much easier for most to develop skills on. These setting could be used to dial it all back to more novice friendly settings when approaching - learning seemingly impossible segments.

 

[ecstatic_subjectivity]

One of the absolute worst characteristics of all the electric bikes I've ridden is that as soon as the tyre looses traction the motor instantly spins up, making a bad situation worse.

Hey all, appreciate all the conversation and back and forth here.

I've had the luck to own a number of electric bikes - OSET 24.0, KTM FreeRide-E, and an Electric Motion Epure Race. I've got a pre-order in on a Stark Varg, that I'm likely to cancel, because it doesn't have a clutch. I've also ridden most of the electrics out there Zeros, BMW C-Max, Alta, SurRon, etc). One of my friends was an early adopter of the Brammo Empulse, as well, and I've got a pretty good spread of experience between road-racing, offroad, onroad, and a bunch of time building my own bikes.

These days, most of my riding is currently Hard Enduro / Trials type riding, and after riding the ePure for awhile, I'm convinced there's no way to actually replicate the hard enduro bike quality - thusly my desire to cancel the Stark. What bikerpete said above is exactly the thing that made me think that electrics were going to struggle to be at the top of the game, and what I eventually realized is that for my style of riding, there was no way for things to work out the way I want them to without the combination of a large amount of flywheel mass, a clutch, and multiple gears. While it's true, I could get up anything that I can on the ePure with the FreeRide, it required significantly more work and energy to balance the lack of additional controls enabled by the clutch. It also created an issue where once you broke traction, you spent significantly more time with the wheel spinning, even if you closed the throttle immediately once it started to spin - the whole rotating being so much lighter means you dig deeper holes, take more time to regain traction, and have less reaction time. Yes, you could install some kind of traction control, but sometimes the potential of that degree of spin is an acceptable tradeoff, or a split second of high spin is acceptable right before you mash the bike into an obstacle where it will hook.

As I've thought about it more and more, I've slowly come to the conclusion that I don't believe it is possible to replace two controls with a single control - the behaviors that are bad in one context are good in another, so without the ability to proactively tell the machine what I want, I sometimes need massive wheelspin, sometimes I want it to hook, and sometimes I want to balance the two actively, and there's no way to do that with just one input to the bike's controller. Even if you could simulate a clutch perfectly with an "electronic clutch", you can't change the fact that the power delivery of an electric motor accelerating and a flywheel decelerating as it drives a rear wheel have fundamentally different power delivery characteristics which are impossible to modulate off a throttle alone, no matter how fast you can move your wrist.

Combined with the delays to the Stark, I'm shifting focus a bit, and I'm going to build a bike based around a 300 2T dirtbike frame, with a QS138 90, EM 260 controller, with the electric motor driving a rotating assembly, clutch, and transmission of a traditional 2 stroke, minus the piston and conrod. I'll be moving shortly, but after that move, I'm hoping to have it knocked out in prototype form relatively quickly. The goal is a bike that can be used for evening rides with my friends up the nastiest stuff Washington has to offer without being too disruptive, while giving me all of the capabilities of my ePure race plus more. Eventually, I'll probably machine out the crank assembly, mount a flywheel directly to the driveshaft of the motor, and have that drive the clutch assembly directly, but that will be significantly more work than simply building a sprocket mount for the end of the crank that the electric motor can drive.

Thanks to everyone in this thread for the lively discussion - I'll start a thread when I start the build!

 

[bikerpete]

I take it you're meaning you're going to keep the crank and flywheel of the ICE engine and just run a chain to that? You're going to keep the flywheel I presume?
That'll be interesting. I think it could work out pretty well if the weight penalty isn't too big a concern, for hard enduro type riding it probably isn't too big a deal.

That's almost what GasGas did with their TXT-e, which was possibly the best of the e-trials bikes, if you could get one. They just went direct to your next step and replaced the ICE motor with an e motor.

I can't imagine how scary a Varg would be with all that power and no clutch! That sounds terrifying.

I think you're spot on the money that in dirt riding wheel slip/spin is not always a bad thing at all. And that it's an exercise in futility to handicap a rider with just a single control - we've got 4 appendages, might as well use them for more than just standing and hanging on!

Your comment about the fundamental difference between a flywheel decelerating and a motor accelerating is a good way to say exactly what I keep banging on about. I'll be using that.

I'll be interested to hear what you settle on for primary reduction ratio. Probably depends a fair bit on the donor bike and it's original purpose and RPM range.
Besides your straight speed requirements, if the bike was originally a high revving two-stroke then the flywheel might be pretty anemic for the much lower revving QS, it just wont store enough energy to really be useful unless you use a step-up ratio on the primary. I'd be trying to find a bike that was originally relatively low revving. Well, actually I'd just be using a trials bike, but there aren't any other bikes as far as I'm concerned. LOL.

One of the absolute worst characteristics of all the electric bikes I've ridden is that as soon as the tyre looses traction the motor instantly spins up, making a bad situation worse.

Hey all, appreciate all the conversation and back and forth here.

I've had the luck to own a number of electric bikes - OSET 24.0, KTM FreeRide-E, and an Electric Motion Epure Race. I've got a pre-order in on a Stark Varg, that I'm likely to cancel, because it doesn't have a clutch. I've also ridden most of the electrics out there Zeros, BMW C-Max, Alta, SurRon, etc). One of my friends was an early adopter of the Brammo Empulse, as well, and I've got a pretty good spread of experience between road-racing, offroad, onroad, and a bunch of time building my own bikes.

These days, most of my riding is currently Hard Enduro / Trials type riding, and after riding the ePure for awhile, I'm convinced there's no way to actually replicate the hard enduro bike quality - thusly my desire to cancel the Stark. What bikerpete said above is exactly the thing that made me think that electrics were going to struggle to be at the top of the game, and what I eventually realized is that for my style of riding, there was no way for things to work out the way I want them to without the combination of a large amount of flywheel mass, a clutch, and multiple gears. While it's true, I could get up anything that I can on the ePure with the FreeRide, it required significantly more work and energy to balance the lack of additional controls enabled by the clutch. It also created an issue where once you broke traction, you spent significantly more time with the wheel spinning, even if you closed the throttle immediately once it started to spin - the whole rotating being so much lighter means you dig deeper holes, take more time to regain traction, and have less reaction time. Yes, you could install some kind of traction control, but sometimes the potential of that degree of spin is an acceptable tradeoff, or a split second of high spin is acceptable right before you mash the bike into an obstacle where it will hook.

As I've thought about it more and more, I've slowly come to the conclusion that I don't believe it is possible to replace two controls with a single control - the behaviors that are bad in one context are good in another, so without the ability to proactively tell the machine what I want, I sometimes need massive wheelspin, sometimes I want it to hook, and sometimes I want to balance the two actively, and there's no way to do that with just one input to the bike's controller. Even if you could simulate a clutch perfectly with an "electronic clutch", you can't change the fact that the power delivery of an electric motor accelerating and a flywheel decelerating as it drives a rear wheel have fundamentally different power delivery characteristics which are impossible to modulate off a throttle alone, no matter how fast you can move your wrist.

Combined with the delays to the Stark, I'm shifting focus a bit, and I'm going to build a bike based around a 300 2T dirtbike frame, with a QS138 90, EM 260 controller, with the electric motor driving a rotating assembly, clutch, and transmission of a traditional 2 stroke, minus the piston and conrod. I'll be moving shortly, but after that move, I'm hoping to have it knocked out in prototype form relatively quickly. The goal is a bike that can be used for evening rides with my friends up the nastiest stuff Washington has to offer without being too disruptive, while giving me all of the capabilities of my ePure race plus more. Eventually, I'll probably machine out the crank assembly, mount a flywheel directly to the driveshaft of the motor, and have that drive the clutch assembly directly, but that will be significantly more work than simply building a sprocket mount for the end of the crank that the electric motor can drive.

Thanks to everyone in this thread for the lively discussion - I'll start a thread when I start the build!

 

[ecstatic_subjectivity]

Yes, your understanding of the design is correct. I may have to do some weighting of the rod pin to keep the crank roughly balanced, but given that it is captured between two bearings, I don't think it'll be a huge deal.

I'm hoping the weight penalty will be sub-30 pounds to start, as the head, piston, conrod, water pump, radiator, etc, will all be removed. I'll have a bottom end, clutch assembly, crank, and transmission / shift linkages. I'm looking at pre-counterbalanced bikes to minimize weight. As it's a two stroke with roller bearings, I don't need a high pressure oil pump, splash lubrication is fine.

The nice thing with doing direct drive to the crank, is it appears that the QS138 90 with flux weakening will easily spin to ~6k RPM, which combined with a slight reduction on the motor output shaft to crank gearing would allow me to spin the crank to ~7k or so, which is within the usable torque range of the stock bike. I might even be able to do a 1/1 motor to crank drive, and simply lengthen the final drive ratio. The other advantage of feeding the bike through a transmission is that you end up with a pretty significant reduction ratio out of the box (due to the crank to clutch gearing), as well as not having to worry about the packaging issues of sprocket location changing the suspension characteristics.

I have a feeling that the Varg would be amazing in their designed use case - MX, where your good riders are always trying to put the power down, and you've got a lot of grip, and the amateur riders aren't going to gripe about that. I think it'd also be great in supermoto, where you're spending almost all of your time hooked up in high traction situations. In the slower, more technical stuff, though, I think it'd be a handful, although if they deliver on the promised adjustability, you could tune it to your riding, but it's always going to constrain you to one map at time, and the value of the clutch/flywheel is that it allows you to modify how that map is applied to the ground, live.

On the power characteristics front, the 300 2ts are tractors, all power biased towards the bottom, so they match the power delivery characteristics of an electric about as much as an ICE can. The nice thing about KTM is they make wide (woods/enduro) and tight ratio (motocross) gearboxes, so a wide ratio gearbox will give me the ability to maximize the available power on tap for a given situation, while also giving me the ability to kick it up to 6th and do transitions / fire roads at up to 65mph. I'm also pretty curious if I can get more range by running the motor consistently in its highest efficiency areas. I'm sure a close ratio gearbox would still pay a lot of dividends, but I'd probably end up mostly using it as a 2 or 3 speed, whereas the wide ratio opens up a bit more flexibility. Honestly, a 3 speed would probably be fine (technical sections, faster hillclimbs, transition roads), but the modern stuff is more gears with a trivial weight penalty, so not a huge deal.

End goal would be to figure out what the ideal dimensions / packaging would be to have the motor mounted and driving the clutch basket directly, with adjustable flywheel weights on the motor output. My initial thoughts are the ideal is being able to mount the motor on the top of the transmission, moving a bit of weight backwards, and letting me drop the battery assembly as low in the frame as possible.

The last thing I'll note that folks often don't consider is how much the large amount of gyroscopic mass of that an ICE has stabilizes the chassis of a bike. 4 strokes have a lot of spinning mass, enough that it's very notable in hard enduro, which is why everyone favors the two strokes. I found the Alta and FreeRide deflected a lot off of bumps, especially on jumps - whereas a 4 stroke might slide very progressively, a 2 stroke might be a bit snappier, the electric bikes with very little rotating mass would break loose or deflect fast enough that it took some active management. The lack of inertia means electrics also come back in line pretty easily, but having the ability to tune the rotating assembly to suit your style can be a huge boost in confidence, especially for riders who aren't pro level / as comfortable riding around those characteristics. For me, stability in a chassis is a factor you want to be able to balance. The more stable, the more comfortable it is on fireroads and when you're doing big slides, but the more you have to work to get it to change direction in the tight stuff. The two strokes are, IMO, about the sweet spot of stability vs agility, so the goal would be to tune it to replicate those characteristics. Electrics have a huge advantage in throttle responsiveness over IC motors, so I'm hoping to have a very aggressive throttle tune that gets moderated through the flywheel and clutch to maximize the benefits of responsiveness, while also making sure it doesn't become totally unrideable when you need fine control over power delivery at low speeds and in tight sections.

I take it you're meaning you're going to keep the crank and flywheel of the ICE engine and just run a chain to that? You're going to keep the flywheel I presume?
That'll be interesting. I think it could work out pretty well if the weight penalty isn't too big a concern, for hard enduro type riding it probably isn't too big a deal.

That's almost what GasGas did with their TXT-e, which was possibly the best of the e-trials bikes, if you could get one. They just went direct to your next step and replaced the ICE motor with an e motor.

I can't imagine how scary a Varg would be with all that power and no clutch! That sounds terrifying.

I think you're spot on the money that in dirt riding wheel slip/spin is not always a bad thing at all. And that it's an exercise in futility to handicap a rider with just a single control - we've got 4 appendages, might as well use them for more than just standing and hanging on!

Your comment about the fundamental difference between a flywheel decelerating and a motor accelerating is a good way to say exactly what I keep banging on about. I'll be using that.

I'll be interested to hear what you settle on for primary reduction ratio. Probably depends a fair bit on the donor bike and it's original purpose and RPM range.
Besides your straight speed requirements, if the bike was originally a high revving two-stroke then the flywheel might be pretty anemic for the much lower revving QS, it just wont store enough energy to really be useful unless you use a step-up ratio on the primary. I'd be trying to find a bike that was originally relatively low revving. Well, actually I'd just be using a trials bike, but there aren't any other bikes as far as I'm concerned. LOL.

 

[bikerpete]

I like your thinking.

Being more trials focused, my experience with the 2-speed DOB gearbox was that you really only need two or maybe 3 gears. For your use I guess 3 or 4 would be better. As you say though there's minimal weight penalty going from 3 to 6.

You could fix a disc of steel (or better a donut) in the crankpin gap (leaving the crankpin in) which would allow you to balance it all and also maybe allow you to slim down the flywheel to make mounting a sprocket easier and less bulky. ?

I'd say be prepared to discover that the acceleration of the motor even with a flywheel is still ridiculously fast. I've played with adding lots more weight to my flywheel on the 138 & although it makes a significant difference to the deceleration it makes a vastly smaller difference to the acceleration. It's not necessarily all bad, but it's certainly got trade offs & remains considerably different from ICE. It's nice to have that snap for some things, but it does make it harder to coordinate clutch release timing as you rev up the motor - the revs just climb so fast.
When I ran mine on a Nucular I could set the rpm rise rate to give the characteristic I preferred, that was really nice. If only the Nuc had an RPM fall rate too.

The other thing that was really good on the Nuc was having the Torque & Speed mode. That was so much nicer for managing lower speed/rpm stuff, it stops the motor wanting to instantly spin up when the load is off and behaves generally with much more subtlety.
Unfortunately the Nuc just doesn't handle idle. The best I could do was set the throttle minimum voltage below the physical minimum voltage so the motor idled. But then it became a bit flaky on turn-on whether it would decide there was a throttle fault. Doing that also means you have to have a Speed&Torque mode or at no load the motor just spins up faster & faster - unmanageable.
My Fardriver can set a fixed idle speed.

Maybe fire up a new thread?

Yes, your understanding of the design is correct. I may have to do some weighting of the rod pin to keep the crank roughly balanced, but given that it is captured between two bearings, I don't think it'll be a huge deal.

I'm hoping the weight penalty will be sub-30 pounds to start, as the head, piston, conrod, water pump, radiator, etc, will all be removed. I'll have a bottom end, clutch assembly, crank, and transmission / shift linkages. I'm looking at pre-counterbalanced bikes to minimize weight. As it's a two stroke with roller bearings, I don't need a high pressure oil pump, splash lubrication is fine.

The nice thing with doing direct drive to the crank, is it appears that the QS138 90 with flux weakening will easily spin to ~6k RPM, which combined with a slight reduction on the motor output shaft to crank gearing would allow me to spin the crank to ~7k or so, which is within the usable torque range of the stock bike. I might even be able to do a 1/1 motor to crank drive, and simply lengthen the final drive ratio. The other advantage of feeding the bike through a transmission is that you end up with a pretty significant reduction ratio out of the box (due to the crank to clutch gearing), as well as not having to worry about the packaging issues of sprocket location changing the suspension characteristics.

I have a feeling that the Varg would be amazing in their designed use case - MX, where your good riders are always trying to put the power down, and you've got a lot of grip, and the amateur riders aren't going to gripe about that. I think it'd also be great in supermoto, where you're spending almost all of your time hooked up in high traction situations. In the slower, more technical stuff, though, I think it'd be a handful, although if they deliver on the promised adjustability, you could tune it to your riding, but it's always going to constrain you to one map at time, and the value of the clutch/flywheel is that it allows you to modify how that map is applied to the ground, live.

On the power characteristics front, the 300 2ts are tractors, all power biased towards the bottom, so they match the power delivery characteristics of an electric about as much as an ICE can. The nice thing about KTM is they make wide (woods/enduro) and tight ratio (motocross) gearboxes, so a wide ratio gearbox will give me the ability to maximize the available power on tap for a given situation, while also giving me the ability to kick it up to 6th and do transitions / fire roads at up to 65mph. I'm also pretty curious if I can get more range by running the motor consistently in its highest efficiency areas. I'm sure a close ratio gearbox would still pay a lot of dividends, but I'd probably end up mostly using it as a 2 or 3 speed, whereas the wide ratio opens up a bit more flexibility. Honestly, a 3 speed would probably be fine (technical sections, faster hillclimbs, transition roads), but the modern stuff is more gears with a trivial weight penalty, so not a huge deal.

End goal would be to figure out what the ideal dimensions / packaging would be to have the motor mounted and driving the clutch basket directly, with adjustable flywheel weights on the motor output. My initial thoughts are the ideal is being able to mount the motor on the top of the transmission, moving a bit of weight backwards, and letting me drop the battery assembly as low in the frame as possible.

The last thing I'll note that folks often don't consider is how much the large amount of gyroscopic mass of that an ICE has stabilizes the chassis of a bike. 4 strokes have a lot of spinning mass, enough that it's very notable in hard enduro, which is why everyone favors the two strokes. I found the Alta and FreeRide deflected a lot off of bumps, especially on jumps - whereas a 4 stroke might slide very progressively, a 2 stroke might be a bit snappier, the electric bikes with very little rotating mass would break loose or deflect fast enough that it took some active management. The lack of inertia means electrics also come back in line pretty easily, but having the ability to tune the rotating assembly to suit your style can be a huge boost in confidence, especially for riders who aren't pro level / as comfortable riding around those characteristics. For me, stability in a chassis is a factor you want to be able to balance. The more stable, the more comfortable it is on fireroads and when you're doing big slides, but the more you have to work to get it to change direction in the tight stuff. The two strokes are, IMO, about the sweet spot of stability vs agility, so the goal would be to tune it to replicate those characteristics. Electrics have a huge advantage in throttle responsiveness over IC motors, so I'm hoping to have a very aggressive throttle tune that gets moderated through the flywheel and clutch to maximize the benefits of responsiveness, while also making sure it doesn't become totally unrideable when you need fine control over power delivery at low speeds and in tight sections.

I take it you're meaning you're going to keep the crank and flywheel of the ICE engine and just run a chain to that? You're going to keep the flywheel I presume?
That'll be interesting. I think it could work out pretty well if the weight penalty isn't too big a concern, for hard enduro type riding it probably isn't too big a deal.

That's almost what GasGas did with their TXT-e, which was possibly the best of the e-trials bikes, if you could get one. They just went direct to your next step and replaced the ICE motor with an e motor.

I can't imagine how scary a Varg would be with all that power and no clutch! That sounds terrifying.

I think you're spot on the money that in dirt riding wheel slip/spin is not always a bad thing at all. And that it's an exercise in futility to handicap a rider with just a single control - we've got 4 appendages, might as well use them for more than just standing and hanging on!

Your comment about the fundamental difference between a flywheel decelerating and a motor accelerating is a good way to say exactly what I keep banging on about. I'll be using that.

I'll be interested to hear what you settle on for primary reduction ratio. Probably depends a fair bit on the donor bike and it's original purpose and RPM range.
Besides your straight speed requirements, if the bike was originally a high revving two-stroke then the flywheel might be pretty anemic for the much lower revving QS, it just wont store enough energy to really be useful unless you use a step-up ratio on the primary. I'd be trying to find a bike that was originally relatively low revving. Well, actually I'd just be using a trials bike, but there aren't any other bikes as far as I'm concerned. LOL.

 

[DanGT86]

Bikerpete, have you asked the guys at Nuc about an rpm ramp down? Seems like a simple software move for them.

As for your idle couldnt you make a simple switch and resistor circuit to send a small throttle output for idle once things are all booted up?

Are these bikes you have been riding with clutches the 3kw bikes? I think as the power of the motor increases the less you would need any type of multi speed gearbox. If the weight of multiple gears were put into copper and magnets you would likely do better than gears. The heavier motor would double as your heavier flywheel.

 

[bikerpete]

I tried to put a feature request in to Nuc, but the system seems broken. I haven't pestered them with an email - both requests are probably pretty niche so not likely to be high priority for them.
I will send them an email, but I won't be holding my breath for them to work on either of them.
Hands up all those who want these two features? Ahh, 3 of us. :lol:

The problem with just sending a throttle input is twofold:
1. In torque mode the rpm varies wildly with small changes in load so you're constrained to speed + torque mode.
2. In an ideal world the controller would adjust current to maintain constant RPM as load varies.

I'll just have to keep disagreeing with you on the "throw more power at it" strategy. Without vastly different controller software that will only make things worse.

For starters increasing rotor mass is an awful way of increasing flywheel effect. You're increasing the mass of a small diameter rotor, but what you want to do is maximise the peripheral mass of the largest diameter you can. So an external flywheel as big as say the motor housing will provide vastly more energy storage per unit mass increase. Even if you had a large diameter, narrow rotor you'd still be better off putting mass into a larger external flywheel.

Next, if you're going to run a high gearing then you're going to need some serious clutch friction in order to be able to go from slipping to locked, fast. That means more plates, bigger diameter, stronger springs & heavier lever pull. A typical 300cc trials clutch is almost instant if you want it to be in low gears, but click it up to 4th or 5th (where you can get to maybe 60kmh) and it becomes too sluggish to be effective. So you're now looking for a bigger, stronger clutch than even a 300cc has.

Finally you come to that fundamental issue that much of the time you want to accelerate very fast, but only to a relatively low top speed. That is really, really hard to do in a high gear! I find the only way to do it is to give it everything for a moment, then be ready to pull the clutch if it gets going to quick, which then cuts all drive. If you're not spot on with going to a slipping clutch, I can tell you it can get scary, fast! Remember this all happens commonly in 1/4 to 1/2 a second.
It's far nicer to have the bike in a gear that self limits the terminal velocity.

My ”3kW" bike is pulling almost 9kW through the controller, in a 50kg bike geared for a top speed of 57kmh. It is staggeringly responsive already. My 300cc trials bike feels like a Land Cruiser compared to an MX5 in comparison. I've already flipped it dramatically twice by giving it too much throttle with insufficient preparation on flat trails. Putting a 4kw motor in it with the same high gearing would be horrible!

"Better" is very situational. Most people seem to default to bigger, faster, more powerful = better. Fine, nuanced, control is the definition of "Better" here, and more power can just be the enemy of control.

Here's something to think about.
If we stick with the idea that a decelerating flywheel is vastly different to an accelerating motor then we are constrained by the physics of flywheels.
The energy they return is proportional to the ratio between the starting rotational velocity and the ending velocity.
It's pretty common in trials to want to accelerate really quickly to a relatively low speed, and then to rely on the flywheel to carry you up and over an obstacle, cresting it with very little momentum left.
If we're in a high gear we can't spin the motor up too fast for fear the bike is going to pick up too much speed. So now we've got a very narrow rev range to recover energy from the flywheel.
So how do we get a good strong acceleration but still have enough inertia in the flywheel to carry us up & over?
The only solutions are to have a really big flywheel or to revert to controlling the power with clutch &/or throttle.
The whole point of relying on the flywheel is because it does it so nicely compared to us twiddling controls. So we've just made an elegant solution completely inelegant!

I can assure you from experience that revving an over geared, under flywheeled bike to the moon and popping the clutch is so different from the desired outcome as to be pointless. The response is completely arse about - the bike leaps forward at a rate of knots, then just stops. It's frustrating in the extreme. You're left having to try to nuance the throttle and clutch to get the right outcome, and it is astronomically more difficult than just using a well balanced flywheel and gear.

As I've said before, one day there'll be controllers that can emulate all this in software, but they sure aren't available on the open market right now!

Bikerpete, have you asked the guys at Nuc about an rpm ramp down? Seems like a simple software move for them.

As for your idle couldnt you make a simple switch and resistor circuit to send a small throttle output for idle once things are all booted up?

Are these bikes you have been riding with clutches the 3kw bikes? I think as the power of the motor increases the less you would need any type of multi speed gearbox. If the weight of multiple gears were put into copper and magnets you would likely do better than gears. The heavier motor would double as your heavier flywheel.

 

[bikerpete]

I got the Nucular Feedback login sorted today so there are now two feature requests in there.

One for an Idle setting
One for a "De-Acceleration" limit.

"Deceleration" is already used for the braking rate limit, a slight misapplication of the terminology I reckon, but no matter.

 

[ecstatic_subjectivity]

For what it's worth, having put a good amount of time in on a Brammo Empulse, a clutch drop on that bike is completely pointless. There's zero rotating mass behind it, the clutch is only there so you have a way to smooth out shifts, if you need it (it shifts fine without it, due to the lack of rotating maass). Also, because of the relatively even output of the motor, you don't really see significantly more power to the ground at any specific RPM, so there's no "getting it into the power" by using the clutch.

If folks are curious about the techniques behind the slow / trials sort of riding, this video might be interesting:
https://www.youtube.com/watch?v=7oDyewNtTF0
Note how often he talks about the clutch as both a mechanism to dump power to the rear wheel, but also as a safety lever. Of course, you can ride a motorcycle without ever doing a wheelie, but for these types of riding, the varying types of wheelies (clutch up, power, bounce, etc) are all critical to navigating the terrain in front of you.

There's a lot of truth to the common refrain in electric drivetrains that you can just put a bigger motor in there, but we're unfortunately limited by what we have available in terms of motor, as well as packaging and weight concerns - those powerful motors get big and heavy fast. It's also worth noting that peak HP for a modern woods bike is about ~55hp/40kw at the crank, and they've got 6 gears to play with, to run you from about 25mph in first to 80+mph in 6th. In terms of "absolute" power, I'm absolutely taking a huge step down, but I'm pretty sure it'll work just fine given my experiences with the ePure Race, thanks to the advantages of the electric powertrain.

I'll start a thread when my parts arrive!