Observed trials bike designing

General Discussion about large electric scooters and motorcycles and other things with no pedals.
User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Dec 31 2021 10:27am

Tech is always evolving. This all just further demonstrates how its a spectrum and you could draw your line arbitrarily at any point and say any technology past this point is cheating.

The other day I was reading about honda's CR series of mx bikes in the 90s. Apparently they implemented a primitive form of traction control using the ignition system back in the mid 90s. It had the unintended result of making the bikes sluggish and taking the edge of the burst of power commonly found in 2 strokes. I could see some guy who rides a 25 year old bike telling a guy on a new bike that he is "Cheating" and "back in my day we didn't have traction control"

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Dec 31 2021 10:52am

I think most of us welcome honest improvements in design. The mapping choices are a interesting option on a race bike. Nice to have the WFO option when you can take advantage of it. Nice saddle and hand grip setup also. Quick shift also nice for keeping the tranny from added abuse.

For a trials bike, the cockpit mapping options may be of some use. A multi disk flywheel that can be separated or connected by operator on the fly may be something that could help. I would like to experiment a bit with either -both. Also a shock "damper" bypass- release would allow much easier hopping, much like a pogo stick when you need it.

Not finding much info on the traction control setup - operation. Only words I find is that it works to smooth out - limit rpm spikes. Sort of an electronic flywheel. May be more of the type of mapping that may be at work at the higher rpm ranges in the power plot up the page a bit of the more current trials motor.

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Jan 13 2022 11:59am


User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Jan 14 2022 10:10am

Very cool bikes! wish I had the money to get one.

It looks like they have included some of the control options and strategies that we have been talking about. I love the tick-over/idle feature. I bet that was hard for the designers of the bike to tune for the right feel.

The regen lever is a cool idea. If I were designing a bike I would try to include a foot shift lever that has a different regen strengths much like the engine braking in a gas bike. Some habits are hard to break.

I'm pretty impressed with the execution here by EM. This bike includes something for everyone whether they are a new rider learning only in the electric environment or a seasoned pro gas rider. Other than a slight weight penalty for including a clutch there is really no reason not to have one. If you don't need it don't use it.

Is the swing arm and frame sections carbon or are those just covers? I kinda missed that in the vids. I'm a bit worried about repeatedly dropping carbon on rocks and logs.

Now I just need to cook up a plan to get a bunch of money.

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Jan 14 2022 10:38am

The carbon look is cool. Some sections (fork covers) appear to be carbon or carbon wrapped plastic. The frame sections are plastic tape with carbon or carbon look facing. Frame is near identical setup to gassers. Looks like they set the bar for folks to follow. Word is that the other major brands are all working on releasing their electric versions.

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Jan 14 2022 8:16pm

All the carbon is aftermarket add-on covers.

EM currently use the Silixcon controller which is very flexible from what I've seen, but I'm not sure it's really capable of the sort of control and logic that might be ideal for trials. Once the likes of KTM make a serious push into electric GasGas trials bikes I expect we'll start to see controllers with software designed with the task in mind from the outset. Or maybe an innovative company like Vertigo might take the leap into e-trial, although they seem to be so head down, tail up developing their ICE bikes that they perhaps don't see that they're developing semaphore while the world moves on to radio.

I'm not sure why EM wouldn't put the regen control on the existing brake levers? Either foot or hand or both. The first little bit of movement would be regen, then the hydraulic would come in. Exactly as electric cars currently do. I don't see why you'd muck about with a third brake lever. You might still keep the ability to set the underlying "engine braking" setting so the bike doesn't just freewheel when the throttle is cut, but regen braking is just braking - use the controls you've already got. Trying to juggle one finger clutch, then over to the regen lever, then back to the clutch - what a nuisance.

I wouldn't have thought the idle setting would have taken much tuning - just gradually increase idle RPM until it feels about right, basically similar RPM to an ICE bike with a little tweaking to cater to the different gearing.

The other downside to the EM clutch is $$$$. You get what you pay for.

I agree that I'd love to have the spare cash to get hold of one. Even though I think they're still a fair way off being as capable as the ICE trials bikes, I'm not as capable as the ICE bikes either!

User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Jan 14 2022 10:42pm

For this style of riding the braking needs to be too precise for it to be combined with the hydro levers. The chain slack would also be a problem compared to caliper braking at the wheel. Sometimes the rider needs to slip the clutch to keep the chain loaded while the wheel is stopped on an edge. There are too many situations where a rider is quickly transitioning between brake and throttle or even using them simultaneously. Imagine trying to rev your engine for a clutch dump while holding the brake to keep you in place. It wouldnt work if the lever is engaging regen.

As for the idle, I can see where it would be tricky to program. If you tell an electric motor controller to maintain 600rpm it will do it. If you start to apply load it will just apply more current to maintain your 600rpm. So then you have a trials bike that wont stall. Thats great but would feel very unnatural. If you tell it to apply a minimum current at all times for idle then its going to continue to rev higher just sitting there. Its a delicate balancing act that just kinda comes free with the analog nature of an ICE motor. Trying to explain it to a computer is a challenging prospect.

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Jan 15 2022 1:42am

You have a point about using throttle, clutch and brake simultaneously. That does make it a bit tricky. Maybe a pushbutton switch on the brake lever itself so you move your finger over a centimeter or so to engage regen or not, then regen varies according to the first part of the lever movement or force? Certainly needs a bit of consideration.

I don't think it would matter in the least if the controller was changing current to maintain 500 rpm against a slipping clutch - you're controlling the torque to the wheel via the slipping clutch anyway, so it matters not one jot if the motor is reacting to the load or indeed varying revs - you regularly vary revs against clutch without moving the bike at all on an ICE.
Once a friction device is fully slipping the rate of slip has far less consequence to the resultant friction than the force applied to the friction surfaces. Kinetic friction does not depend on the sliding velocity.
In fact I suspect it might be a good thing if it did maintain constant RPM, it would probably make it slightly easier to balance against the clutch if anything.
Holding pressure (balancing on the rear wheel) on an obstacle while building revs for a zap across a gap is an example of almost complete separation of engine rpm/power and control of torque to the wheel. It's no big deal, just requires skill development.
I don't think the EM's do stall either - there's an argument going on in some parts of the world saying that makes it unfair against ICE bikes. Now that's a pointless argument - electric is coming, get over it!

Gotta get myself a ride on a recent EM so I can get some real world knowledge!

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Jan 15 2022 10:55am

The regen lever looks like a afterthought imo. May make marketing sence, but agree, it should be incorporated into the existing controls. Looking for the first one to come in with a variable flywheel mass control. Separating that from idle speed, would allow tremendous amount of variety to impulse potential. Agree, may be time spend a bit of ride time on one just to see if any of the skills improve or better see what direction things should go. The electronic shock damper control has me quite interested at the moment. Somethirng like the pedal on a piano does to keep the felt off the strings, it could be activated momentarily to allow added pogo as needed.

User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Jan 15 2022 11:08am

I was at a trials competition recently and I asked the riders about the rules regarding electric bikes. They said in some sanctioning bodies an electric bike in the group gives all the riders an opportunity to restart their bikes without a any points.

As for the idle stuff. I'm sure its nothing for engineers and programmers to work out. I just always like to try and wrap my head around the concepts and what it would take to make them work as a a diy builder. In the videos you can hear the motor "bogging" as the rider works the clutch. So it does seem to simulate the behavior of the ICE motors riders are used to.

What I do think is really cool about all of this is that you can electrically simulate any type of gas bike. 2strokes and 4 strokes have different engine braking and power curve behaviors. Different flywheels weights have different behavior. All of that can be mimicked by electric motor control programming. So its conceivable that an e-trials manufacturer could custom tailor a perfect feeling bike for any rider.

I would also have a noise vs silent switch. If you are going to rev up and dump the clutch then the rpm motor noise would be pretty important. If you were someone who learned in a clutchless EV format then you might not want or need the noise.

Exciting times we live in.

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Jan 15 2022 9:57pm

Indeed.
That's funny about being able to restart if there's an e-trials in the class! Certainly doesn't happen that way here in Australia, nor in Germany from what I hear from a friend over there.

I'm still not quite convinced that, " All of that can be mimicked by electric motor control programming."
Perhaps it's just that it can't yet be mimicked because the controller hasn't been designed that can replicate a flywheel. There are huge issues to overcome to achieve that, so I'm not sure anyone is going to bother for a while yet. Not least is how big does your motor need to be in order to give the short high power impulse that is available from a smaller motor/flywheel combination? You'd need to be able to supply all the power a flywheel can store without a significant weight penalty.

I should soon be receiving my first shipment of a new flywheel/clutched, brushless e-trials bike, claimed peak 10kw, around 42kg. Really looking forward to getting them and having a play. I've got a Nucular 12F here that I'll put in one if the factory controller isn't flexible enough for me to play with.

DEHiCKA   100 mW

100 mW
Posts: 43
Joined: Mar 15 2015 4:16am
Location: Troy

Re: Observed trials bike designing

Post by DEHiCKA » Jan 17 2022 4:30am

bikerpete wrote:
Jan 15 2022 9:57pm
Indeed.
I should soon be receiving my first shipment of a new flywheel/clutched, brushless e-trials bike, claimed peak 10kw, around 42kg.
Interesting, is it a new company? do you have a link?
I would assume it does not have the standard trials 21"/18" wheels+tires, shock and fork, or 42kg would not be possible.
Trials 49KG middrive on modified stainless LMX frame
QS120/QS138 Geared + custom 8-62T #420/ 9-60T #428
Kelly KEB 315A/450A Peak/Fardriver 530A FOC/Nuc 24f
Michelin X-Lite 21"/18" Jitsie moto trials wheels

User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Jan 17 2022 12:28pm

bikerpete wrote:
Jan 15 2022 9:57pm
Indeed.
That's funny about being able to restart if there's an e-trials in the class! Certainly doesn't happen that way here in Australia, nor in Germany from what I hear from a friend over there.
Trials is such a rare riding style here in the US that I think they will take whomever they can get. I'm sure its different depending on the sanctioning body.

The competition I was at had probably 30 riders but some of them had come from 300-400 miles away so they are literally few and far between.

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Jan 19 2022 5:53pm

DEHiCKA wrote:
Jan 17 2022 4:30am
bikerpete wrote:
Jan 15 2022 9:57pm
Indeed.
I should soon be receiving my first shipment of a new flywheel/clutched, brushless e-trials bike, claimed peak 10kw, around 42kg.
Interesting, is it a new company? do you have a link?
I would assume it does not have the standard trials 21"/18" wheels+tires, shock and fork, or 42kg would not be possible.
www.dobbikes.com.au. The old model was 35kg, so the new one has put on a bit of weight but it should be significantly more capable.
The new model isn't shown yet as I'm waiting for some small changes to be made to make it compliant with western regulations. It's similar to the shown model, but all new frame and battery, triple clamp style forks and of course the new motor/flywheel/clutch setup. Toothed belt primary to the clutch.

DanGT86 wrote:
Jan 17 2022 12:28pm
Trials is such a rare riding style here in the US that I think they will take whomever they can get. I'm sure its different depending on the sanctioning body.

The competition I was at had probably 30 riders but some of them had come from 300-400 miles away so they are literally few and far between.
In Europe & UK it's a bit of a different scene, with many thousands of riders and of course all the contenders for Trial GP and the development classes for TGP. So they get a little more excited about the rules.
The very best riders from everywhere outside Europe are not even vaguely contenders for Trial GP - when you see what people like Pat Smage etc do, it puts it in perspective when he says he's maybe at 60% of the top TGP rider's level! They are very serious over there!

DEHiCKA   100 mW

100 mW
Posts: 43
Joined: Mar 15 2015 4:16am
Location: Troy

Re: Observed trials bike designing

Post by DEHiCKA » Apr 23 2022 8:00am

Picked up barely used Vertigo R2 300 in october. 68 kilos and 26Kw 52Nm @crank, what a petrol beast!
By my calculations this translates to 1800Nm wheel @1st gear minus 3-4% for the gearbox and the chain drive losses.
Clutch drops on the 2nd and 3rd feels brutal even at mid RPM.
Injector mapping goes up 12000RPM. Do not know the exact dimensions of the flywheel, but by my rough calcs (120 OD/80 ID/50 H in mm) it's kinetic energy at this rpm can be up to 15-20kJ. 12000 -> 6000 RPM drop will dump 12-15kJ.

My DIY trials bike is only ~390Nm. Currently in the process of upgrading it to the geared QS138 so I can get it close to 900Nm with the top speed around 50km/h.
Vertigo DHT.jpg
Vertigo DHT.jpg (373.83 KiB) Viewed 250 times
I have been digging into EM epure race specs and found the motor they used is some version of Ashwoods/DanaTM4 IPM200-50.
Datasheet says it has 74Nm peak torque. They use the same 2.777 primary gear pair as the majority of modern 2T trials bikes.
So with 2.777 * 57/11 = 14.38 total reduction for epure race it must produce close to 1000Nm wheel torque.
My only guess it is the 11Kw controller is why they down to 600Nm torque in specs.

Also EM center of gravity feels way higher than Vertigo. EM heavy battery goes all the way up and forward in the frame.
Vertigo has basically no significant wight higher than the motor cylinder head, only very light plastic airbox and air filter.
Trials 49KG middrive on modified stainless LMX frame
QS120/QS138 Geared + custom 8-62T #420/ 9-60T #428
Kelly KEB 315A/450A Peak/Fardriver 530A FOC/Nuc 24f
Michelin X-Lite 21"/18" Jitsie moto trials wheels

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Apr 25 2022 8:18pm

DEHiCKA wrote:
Apr 23 2022 8:00am
Picked up barely used Vertigo R2 300 in october. 68 kilos and 26Kw 52Nm @crank, what a petrol beast!
By my calculations this translates to 1800Nm wheel @1st gear minus 3-4% for the gearbox and the chain drive losses.
Clutch drops on the 2nd and 3rd feels brutal even at mid RPM.
Injector mapping goes up 12000RPM. Do not know the exact dimensions of the flywheel, but by my rough calcs (120 OD/80 ID/50 H in mm) it's kinetic energy at this rpm can be up to 15-20kJ. 12000 -> 6000 RPM drop will dump 12-15kJ.

My DIY trials bike is only ~390Nm. Currently in the process of upgrading it to the geared QS138 so I can get it close to 900Nm with the top speed around 50km/h.

I have been digging into EM epure race specs and found the motor they used is some version of Ashwoods/DanaTM4 IPM200-50.
Datasheet says it has 74Nm peak torque. They use the same 2.777 primary gear pair as the majority of modern 2T trials bikes.
So with 2.777 * 57/11 = 14.38 total reduction for epure race it must produce close to 1000Nm wheel torque.
My only guess it is the 11Kw controller is why they down to 600Nm torque in specs.

Also EM center of gravity feels way higher than Vertigo. EM heavy battery goes all the way up and forward in the frame.
Vertigo has basically no significant wight higher than the motor cylinder head, only very light plastic airbox and air filter.
Nice. It seems most people who have Vertigos really like them.
I suspect you'd be extraordinarily lucky (perhaps unlucky) to ever hit 12,000 rpm. 10 - maybe. I understand you can see the engine outputs from the ECU? Perhaps that's with an add on wifi unit? I'm not sure, just what I've heard people talk about.

I'll be interested to hear how you go with the geared QS.

Regarding the flywheel, you'll be well under the actual energy figure because the crankshaft itself has very significant mass - the crank ends go into solid steel discs that carry the eccentric. There's probably similar mass in that lot as there is in the flywheel, although not as well distributed for energy storage.

I have a theory that potentially one of the biggest factors contributing to the relatively poor performance of the EM compared to ICE bikes is simply that it is geared too high. A typical ICE bike in 3rd gear would be really struggling to do 65kmh (top speed of EM) I imagine (I should put a GPS on my bike and see), more likely you'd need 4th to get there, yet even 3rd gear is seldom used when really fast acceleration/rotation/lift is required. Certainly 4th is extremely rarely used except when you need to clear really big gaps, usually with a little bit of room to accelerate. At those higher gearings you need a really powerful/fast clutch to get good response - the clutch settings the elite riders use (who are the ones who really make use of 3rd & 4th gear) would be totally unrideable by us mortals - more like dog clutches in our hands! I know a very experienced pro rider who says the GasGas clutch at fastest settings is just insanely fast, he has to dial it down considerably.

So how on earth do they expect a bike with less torque, less flywheel inertia (I think) and a less powerful clutch to really perform? It's not going to happen!

If they geared it down close to an ICE bike second gear it would possibly transform it, but it would be very poorly received due to being so slow between sections. It would be an interesting experiment to put a really big rear sprocket on one.

I've heard from a couple top level riders that the EM Race feels like it's glued to the ground compared to ICE bikes, that it's really hard to generate significant lift. Also that the bike Gael rides must be significantly different to the production bikes.

EDIT: I just thought to add - the GasGas e-TXT was a standard ICE bike with an electric motor grafted into the crankcase & gearbox. Many people (me included at the time) kind of thought it was a bit of a half hearted foray into electric - taking the absolute easiest path to electrifying what they already had. Possibly it was.
But in hindsight I wonder if they were actually very well aware of what they were doing and recognised that an e-trials still needs a gearbox to get the best out of it across the full range of speeds.
I suspect that in future we'll see more bikes with a 2 or maybe 3 speed gearbox, Trials and Travel modes.

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Apr 26 2022 7:41am

Vertigo looks like a very nice bike. I have a Sherco 290 which does get away from me at times :oops: . Would like to try a modern 250 to compare. Most likely a 250 would be a better fit for my level of riding. I am not seeing any reasonably priced used bikes anywhere around.

Good points on the flywheel and travel speeds. 6 phase windings and dual controllers could be a suitable setup to gain on both issues. Bigger motors also.

Not sure how the sevcon is setup to run the 2012 Polaris Ranger EV, but the "low" setting on the stock three speed switch acts just like a low gear. Need to dig into this a bit deeper.

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Apr 27 2022 8:29pm

speedmd wrote:
Apr 26 2022 7:41am
6 phase windings and dual controllers could be a suitable setup to gain on both issues. Bigger motors also.

Not sure how the sevcon is setup to run the 2012 Polaris Ranger EV, but the "low" setting on the stock three speed switch acts just like a low gear. Need to dig into this a bit deeper.
I know I've got a bee in my bonnet about this, but I very strongly believe that trying to throw more electrical power at the problem is completely counterproductive. More power without the proper delivery characteristics is just going to make things worse and worse, not even considering the extra weight of motor, controller & batteries as you make it all more powerful.

Unless and until someone develops a controller specifically capable of providing the characteristics required for trials. Actually by extension for many other motorcycle activities - it's no coincidence virtually all the top hard enduro riders come from a trials background.
Personally I think we're a very long way away from doing without a mechanical clutch except for basic entry level riding.

Trials bikes are universally completely gutless compared to similar sized "dirt bikes", probably half the power or less. It's telling that most trials riders have no clue how much power their bikes make, and nor do they care. It's also telling that there are youngsters on 125's who are capable of challenging riders on 300's right up to the Expert class. The 125's are positively anaemic in their power output! Yet they can still be competitive simply because they can rev high and so store energy in the flywheel.
Power isn't a substitute for finesse in this sport.

This lack of power isn't totally by accident either. Yes, with 2-strokes it can be thought of as the engine can produce a total sum of power beneath the output graph (the area), by tuning the exhaust etc it's possible to shift power from one part of the graph to another. This is why the most powerful 2-strokes are "peaky", developing nearly all their power in a very small band. Having a peaky motor would be a disaster for trials! So 2-stroke trials bikes have to sacrifice peak power for smooth delivery right from idle.
However there's also the simple fact that too much power just makes it far too easy to spin the wheel. Having a heavy flywheel helps mitigate that, and this could similarly be done in the controller by limiting the rpm rate rise.

I'll deviate for a moment to explain my understanding of rpm vs torque in trials. If we look at the more advanced splatters and zaps it's easier to see one of the difficulties for a controller to overcome. In either of these moves, where the bike is sent more or less vertically airborne, the key mechanism is rotating the whole bike really rapidly around the rear axle so the centre of gravity accelerates upwards. Once that is headed skywards it's relatively easy to bring the rest of the bike along with it. The key here is that the back wheel has to be accelerated from usually close to 0 rpm to pretty much flat out (for that gear) in a fraction of a second. We're talking around 1/30th - 1/15th of a second from what I can see on video of the guns. If the acceleration is too slow the bike isn't going to rotate fast enough and will get going forward, not up. Given the light weight of the bikes, the rear position of the pegs and the fact the rider is trying to help the bike to rotate, there isn't actually a lot of torque required to get this rotation going. So RPM is often more critical than torque.
It's absolutely common that at the completion of a splat or zap there are still way more rpm available than needed, so you pull the clutch and control the drive with that, adding throttle as needed. This reinforces that RPM is often more critical than torque.
So now you need a controller/motor that can go from 0 to full rpm in say 1/30th second. Probably doable but perhaps needs an oversize controller to achieve it? Not my field of knowledge. There's also the impact on the bike of that sort of starting torque applied by the motor, it could be beneficial or it could be an added complication, I don't know.

If rpm rise is limited by controller then how do you get the massive instant torque & rpm when required, and also the ability to chop that torque totally when needed? The answer is by having a separate control regime operated by the clutch lever. So that the rpm ramp is 0-full over just a few mm of lever travel (as it is on current bikes - the engagement travel is really very short). That still doesn't go anywhere near to solving the problem of how that instant power is applied. Nor does it actually totally cut drive - there's still the inertia in the motor rotor etc. - that's when trials bikes get away from people, when they can't chop power with the clutch fast enough. Your clutch is your get out of jail free card!

A flywheel supplies power that decays as power is drawn down. So now the controller needs to supply big rpm that rapidly drops, but critical is that it drops in response to the load. There's no use punching up into a big splatter across a gap only to find that rpm has dropped to zero by the time the rear wheel hits the rock face! That will end badly indeed. So now our controller needs to see the load applied to the drive and adjust the decay rate accordingly.

Keep in mind that during all of this the throttle can only be used for really coarse control - I defy anyone to be able to finely control a throttle tube while they are simultaneously launching into a big jump, pulling the bars hard to their hips and then dropping their hips down to the rear tyre. :lol:

Consider this also - with a mechanical clutch in slipping state the amount of power delivered to the rear wheel effectively does not vary if the motor rpm are varied. This allows all sorts of stuff to happen, but one key one is that it's possible to dial up the power rather approximately with that coarse control throttle and apply it smoothly to the rear wheel via the clutch. This too can be done electronically I expect, but I don't know of any controller really setup to do it?

I could go on - I've spent a lot of time considering this! There's lots of other situations where these same factors are blended in different ways, but always it comes down to decaying power in response to load, ability to instantly cut all torque to the rear wheel and the ability to use throttle to dial the "available power" and use the clutch lever to finely manage the delivery of that available power.

These last two often seem to be totally disregarded in these discussions, but I think they're critical.
A throttle tube is slow, coarse and impossible to keep accurate as the bike and rider leap about. It's great for the basic power control but complete rubbish for the fine & rapid stuff.
A clutch lever is operated by one finger with the rest of the hand stabilised on the grip. The operating range is over just few mm. It's amazing how accurate and consistent riders become. I've reviewed videos where it's gone from fully slipping to fully engaged to partly slipping then out to full engaged in 1/15th of a second - 4 frames of a 60fps video while the rider was gapping to splatter! Absolutely not possible by throttle.

So I ask, why not just put in a flywheel and clutch?

Sorry for the totally excessively long rant. :D :oops:

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Apr 28 2022 2:11am

Having piqued my own interest I went out with a GPS on the bike a couple of days ago.

2nd gear not much over 15kmh
3rd gear high 20's
4th gear mid 30's
5th gear finally got over 50kmh
6th - didn't bother

I could possibly have eked a few more kmh out of each gear, but the bike was really not too far off peak rpm - it was fairly singing!

This is on a Beta 300. Beta are known for their low first gear - their second is typically just a bit taller than most bikes' first. But the big gap between 4th and 5th is typical of all bikes. So Iexpect this would be somewhat similar for all of them.

The DanaTM4 IPM200-50 lists as max 7000 rpm, which is less than any ICE bike.
So it seems that effectively the EM is running a ratio well up around 5th on an ICE bike. No wonder it tends to be underwhelming off the clutch! Dropping the clutch at 7000rpm in 5th gear on the Beta would be positively leisurely.

Interesting. I'd love to see how an EM rode with a huge rear sprocket.

User avatar
spinningmagnets   100 GW

100 GW
Posts: 12709
Joined: Dec 21 2007 10:27pm
Location: Ft Riley, NE Kansas

Re: Observed trials bike designing

Post by spinningmagnets » Apr 28 2022 7:08am

Absolutely not possible by throttle...So I ask, why not just put in a flywheel and clutch?
Why not both?

Competition improves the breed. If a certain configuration dominates a competition, so be it. It sounds to me that controllers and electric throttles need development and improvement.

If there was a competition for electric trials motorcycles, what would the givens be? If they all use the same voltage,, what should it be? 52V? 72V? 28S/102V?

How many amps are ideal?

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Apr 28 2022 8:38am

So it seems that effectively the EM is running a ratio well up around 5th on an ICE bike. No wonder it tends to be underwhelming off the clutch! Dropping the clutch at 7000rpm in 5th gear on the Beta would be positively leisurely.

Interesting. I'd love to see how an EM rode with a huge rear sprocket.
Gearing it lower without a manual clutch would be significantly more difficult to solely control via a twist throttle as you mention. Most of my recent "get-off's" on the 290 have been from too low a gear and my right wrist in less than perfect position, while landing a bit off position. My clutch skills still need a boat load of improvement. :oops:

Wondering if electronic actuator -variable pressure sensitive switch could improve - eliminate the single finger "reach" of the current clutch lever. Possibly combine a thumb side pressure switch for regen. Keeping all your fingers on or at least wrapped a bit more of the bars should help on tugs and landings. Still quite the clutch novice with my current setup and need to keep at it a bit more to see how I develop the skills.

User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Apr 28 2022 9:15am

I'm still hung up on the idea that if a controllers current ramp is too slow then dumping the clutch wont increase the instant power hit without a flywheel at least equal or more massive as its ICE counterpart. Ultimately the controller gets the final say regarding the time it takes the torque to ramp up to full. I don't know how fast this happens in an ICE bike. I would imagine it is within a few revolutions of the crank.

To make the bike well mannered it seems like the controller is going to have to be able to differentiate between clutch dumps and normal throttle application.

Side note, has anyone seen pricing for these DanaTM4 IPM200-50 and similar Dana motors? Is there a direct to consumer way to buy them?

User avatar
speedmd   1 GW

1 GW
Posts: 3177
Joined: Nov 14 2012 12:16pm
Location: new england

Re: Observed trials bike designing

Post by speedmd » Apr 28 2022 11:24am

I'm still hung up on the idea that if a controllers current ramp is too slow then dumping the clutch wont increase the instant power hit without a flywheel at least equal or more massive as its ICE counterpart. Ultimately the controller gets the final say regarding the time it takes the torque to ramp up to full. I don't know how fast this happens in an ICE bike. I would imagine it is within a few revolutions of the crank.
Played with throttle ramps a bit with the CA. Lots of variation available. Certain, there are lots better setups than what I was playing with in the basic trap. ebike controllers. Given the much higher torque of the electric motor at low rpms, they should (spin up the flywheel faster) be able to out perform the ICE setup with equal rotating mass. Will be a interesting year seeing how they do on the same tracks as the ICE bikes this year.

User avatar
DanGT86   100 kW

100 kW
Posts: 1024
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Apr 28 2022 8:40pm

I was in the same boat with the trapezoidal controllers and the CA3 tuning. I could never get the gain settings just right to keep the CA3 from oscillating trying to limit based on watts or amps. I eventually just gave up and set time based throttle ramp in volts/sec. Those old cheap controllers are just too violent! When I switched over to the Nucular controller the default settings were already way better than anything I came up with on the CA3.

All of that experimentation really drives home the point that advanced control is needed for any kind of precision riding like trials. There isn't just one control strategy that that suits all situations. Any ramp up that is slow enough to not oscillate at the current limiter of the CA would be inadequate for explosive trials moves. Any settings explosive enough for trials are horrible when trying to hit jumps on the trail. Tuning out the dead band in the throttle while rolling at med speed makes low speed control too twitchy.

It would be a monumental task to create a bike that felt natural in all offroad situations. It would definitely need throttle, freewheel, and regen settings that were tailored to all kinds of situations and trials maneuvers.

I haven't given up on the belief that a finely tuned controller with enough amps could obsolete a clutch but I am pretty convinced that a clutch is a way easier solution and way more familiar to most riders at this point. Id rather have one and not use it than realize at the end of a build that I want one.

bikerpete   100 W

100 W
Posts: 112
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Apr 28 2022 9:07pm

spinningmagnets wrote:
Apr 28 2022 7:08am
Absolutely not possible by throttle...So I ask, why not just put in a flywheel and clutch?
Why not both?

Competition improves the breed. If a certain configuration dominates a competition, so be it. It sounds to me that controllers and electric throttles need development and improvement.

If there was a competition for electric trials motorcycles, what would the givens be? If they all use the same voltage,, what should it be? 52V? 72V? 28S/102V?

How many amps are ideal?
WARNING - another very long read. Sorry, can't help it. This stuff occupies a lot of my thinking time.

Why not both?
  • Weight. The bikes used at Trial GP level are amazingly close to stock. This reflects that stock trials bikes are actually highly refined competition machines. Electric trials bikes are currently so overweight compared to ICE that they simply can't compete at the same level. This is just considering weight, not all the other factors that leaves them so far behind. Add bigger motors, bigger controllers & bigger batteries to supply them and it becomes completely out of the ballpark.
  • As I suggested, more power can be a disadvantage not an advantage. Traction and control are the name of the game, power can be their enemy. Once there's "enough" then more is worse.
I could probably add a bunch more, but these are really sufficient.

Electric throttles aren't really too much of an issue - yes they could certainly be improved, but when it's understood that a throttle can only be a coarse control then it's not such a big issue. Durable, consistent, linear - all good.

Controllers certainly need a huge amount of development to really solve the curious needs of trials. Having said that, I'd suggest that if controllers became capable of the needs of trials then I suspect those capabilities would find uses in other applications.

There are e-Trials competitions, both at club and Trial GP level.

Albert was at the pointy end of Trial GP for many years until retiring to e-trial in 2018.

This year for the first time e-Trial GP will ride courses the same as the main support class (Trial 2). Previously their courses were below even that level.
That's not to understate the difficulty of those courses though. The level of skill in any of the world round classes is absolutely mind boggling! The best riders in the USA, Australia and many other countries can barely complete many of the Trial 2 courses, let alone be competitive, and so very, very far from competitive in Trial GP. The best rider in the USA (by a fairly decent margin for a long time) once said he was at around 60% of the skill level of the top Trial GP riders!

At open level events (not closed classes) I believe the "givens" are pretty much restricted to similar standards as ICE bikes - tyres, minimum weight, safety features. I think there might be a maximum voltage but it's more around electrical safety than anything else - trials is pretty open about what you can ride. It's nothing like car "formulas".

How many amps are ideal? That's unanswerable.
See above about control, traction & power. If control was ideal then the amps might only need to be relatively low. If a heavy flywheel is incorporated amps might be low. If no flywheel then amps would need to be considerably higher.

I suspect it's very hard to comprehend the needs of trials without some significant experience in it.

One of the fundamental skills for trials (beyond the basic entry level riding) is being able to completely separate throttle and clutch, and to be able to use this to provide short, sharp, controlled accelerations.
One exercise is to ride walking pace along flat ground, with 3 markers 2 meters apart.
At first mark increase revs considerably but don't increase speed (slip the clutch).
2nd mark - squat down into the bike. Keep revs and speed steady.
3rd mark - release the clutch and extend up - "Go"
The bike will stand up into a small wheelie (depending on the rpm) but immediately drop back down - quick but controlled.

Easy hey?
Not!
Absolutely invariably people will either give the throttle an extra tweak as they Go which causes the wheelie to be far less controlled and encourages wheelspin, or they will try to pull on the bars or move their body backwards or forwards, etc.
This exercise builds the essential ability to drive the bike off the clutch. Notice that there is no mention of closing the throttle in there, but that's what will happen as you Go, it is far harder to keep the throttle on as you extend than to let the movement naturally roll the throttle off.
This simple exercise underpins all of the advanced moves that the Trial GP riders do, you can see it left right and centre - Rev, Squat, Go.
Sometimes it's really obvious, like when they stand there stationary in front of a big step, then pin the throttle wide open until the engine peaks before a massive explosion up as they drop the clutch - the bike rotates almost instantly, going well past vertical in a fraction of a second, the rider leaps up, often completely off the pegs, and pulls the bars up toward them trying to help lift 65kg up into the air. A fraction of a second later the rear wheel impacts and they control that impact with their legs to help maximise traction. Any excess rpm is managed with the clutch, by pulling it to the slip zone and controlling power with clutch slip.

I think there's a lot of relevance to e-trials design in understanding that basic exercise and how it transfers up the skill level.
It doesn't take a lot of imagination to see in it that the throttle can only be a coarse control - it's just not possible to keep a rotating handgrip accurately positioned while you're on a bike going from stationary to semi-inverted while trying to dead lift 65 kg into the air while jumping off your feet, then impacting a vertical wall and rotating forward again while trying to control the power at the wheel to the degree that the wheel might do just 1/4 of a revolution - all in well under a second! Mind boggling that some humans can do it at all really.
DanGT86 wrote:
Apr 28 2022 9:15am
I'm still hung up on the idea that if a controllers current ramp is too slow then dumping the clutch wont increase the instant power hit without a flywheel at least equal or more massive as its ICE counterpart. Ultimately the controller gets the final say regarding the time it takes the torque to ramp up to full. I don't know how fast this happens in an ICE bike. I would imagine it is within a few revolutions of the crank.

To make the bike well mannered it seems like the controller is going to have to be able to differentiate between clutch dumps and normal throttle application.
This is indeed at the nub of it. But it doesn't stop there.

I completely agree that the controller somehow needs to have a ramp for the throttle control. But I strongly believe that simple current ramping is vastly inadequate. It has to be a current/rpm ramp.
One of the commonly reported problems with e-trials bikes of all makes is that they have a horrible tendency to spin up wildly as soon as they lose traction for a moment. That's where there needs to be an rpm rise rate built in so it's possible to catch the spin and stop it before it goes too far. It's yet another function provided in ICE bikes by low power combined with big flywheel - the rpm always rises relatively slowly.
You could easily create an anti-spin routine, but that would be a severe hindrance in situations where you actually want a significant amount of wheelspin (they exist).

And as you say, when the clutch is dropped, in the absence of a serious flywheel, that motor had better provide instant amps or it's a complete fail. Again, an ICE bike provides that instant torque by virtue of the flywheel and also the nature of ICE.

An ICE produces power completely differently in one repsect to an electric motor.
In an ICE the throttle supplies a metered amount of fuel/air to the engine that effectively is irrespective of the power demand on that engine. I twist the throttle to 1/4 and it gives the same metered amount of fuel irrespective if I have the bike in neutral or am climbing a hill (more or less).
So if I have the motor spinning freely at half throttle, let's say the wheel is airborne, and the wheel impacts the ground then the engine is instantaneously going to provide all the power that comes with half throttle. The fuel/air mixture is already there, the spark is firing, it's all there waiting for a force to work against. The ramp rate is a maximum of 1 rev. in a 2 stroke.
That's not what happens with any controller/motor combo I'm aware of. The controller would have the motor freewheeling at low current and would have to react as the load gets applied. You'd need seriously fast current response for that situation!

So the controller needs to differentiate between clutch & throttle, but also be able to autonomously instantly switch current in repsonse to wildly varying load.
And those current ramp rates have to be tied to rpm ramp rates too.
And I'd argue that the response to load should be tied to a decaying RPM rate, certainly it has to be the slow rpm ramp at the very least.

I'm sure it's all doable in software, and maybe one day it will. It just seems like there isn't a controller in existence that can address all these needs yet.
speedmd wrote:
Apr 28 2022 11:24am
Given the much higher torque of the electric motor at low rpms, they should (spin up the flywheel faster) be able to out perform the ICE setup with equal rotating mass.
Therein lies the problem - you don't want it to spin up too fast.
It's one of the problems people have with the EM - it's really hard to use the clutch effectively because the motor/flywheel are too digital. They are either at idle or they are zipping up to full rpm at a rate of knots. Even if you only give it modest throttle, at zero load it just spins right on up.
This is where the throttle needs a relatively slow rpm ramp rate, independent of the current ramp rate.

In every other motorised activity it seems like power is the key goal. It just isn't in trials!
As I said peviously people on 125's can seriously compete almost all the way to the top of the sport agains people on 300's. Find me another motorsport where a vehicle with less than half the power, almost the same weight and identical design can seriously compete in the same race!

So I do keep coming back to the thought that a lot of this stuff is done so elegantly by a flyhweel, whereas it's just such a kludge to make it work electronically at present.

Post Reply