What determines wheel zippiness?

[Chriull]

30 minutes ago, mrelwood said:

 

Mike preferred a tilt in the opposite direction, forward. Aneta prefers a tilt backwards.

 

Have to reread that in detail! Maybe just @Mike SacristanSacristan stated rhe same with a different name?

My very personal unfounded opinion in regard to going up steep inclines and tilt calibration is, that it just feels more "natural".

With already the steep incline in front of the face and the toes almost touching the surface imo one subcounsuously is hindered by oneself to natural lean and push forward. With a firward backward tilt calibration one can apply the pressure more naturaly/without mental barrier?

But, as written that's just my very personal opinion - maybe i do the torque difference calculation for different tilts...

PS.:And as written some posts before another very personal idea of mine could be an "unnatural" forward tilt happening by firmware current/power limiting going slowly up inclines...

Edit: backward instead of forward tilt calibration...

Edited December 12, 2019 by Chriull

[Chriull]

4 minutes ago, Aneta said:

My understanding is that the angle of the pedals relative to "neutral" (calibrated) angle is the throttle input

No. EUCs are self balancing - so they have as primary "control loop goal" to keep the pedal even (at the calibrated position). This is not comparable to an car or e scooter throttle.

A higher angle deviation should just "increase the urge" for the wheel to reestablish balance. These different reactions of the firmware imo make (a part) of a wheel zippiness/hardness/softness.

[Aneta]

9 minutes ago, Chriull said:

With already the steep incline in front of the face and the toes almost touching the surface imo one subcounsuously is hindered by oneself to natural lean and push forward. With a firward tilt calibration one can apply the pressure more naturaly/without mental barrier?

Yes, with pedals level while climbing, vs. pedals dipped forward, it does feel safer and more natural, because the ground clearance is increased and you don't have to exert any effort whatsoever. It does feel like an invisible escalator with level steps under your feet.

Edited December 12, 2019 by Aneta

[Aneta]

5 minutes ago, Chriull said:

No. EUCs are self balancing - so they have as primary "control loop goal" to keep the pedal even (at the calibrated position). This is not comparable to an car or e scooter throttle.

A higher angle deviation should just "increase the urge" for the wheel to reestablish balance. These different reactions of the firmware imo make (a part) of a wheel zippiness/hardness/softness.

I see. I guess, I was wrong about angle = throttle; it's the temporal derivative of the angle (aka angular velocity) = throttle. Which makes sense, because gyros measure angular velocity (omega), not absolute angle per se (integrating angle from omega is prone to divergence due to gyro's bias).

I'll have to redo my experiments with clinometer mounted on the wheel to observe the angles at all times, not only during calibration. Stay tuned.

[Aneta]

57 minutes ago, mrelwood said:

I do have a language barrier to cope with, but to me what you have explained sounds clear. Did I understand wrong:

1: Flat calibration, not moving.

2: Tilt calibration, not moving.

3a: Tilt calibration during an incline.

3b: Tilt calibration during acceleration on flat ground.

If this is the case, I suggest you try the same with a hard riding mode. The result should be quite different.

Looking at your drawing, this is what I observed today: #1 is my normal calibration. I recalibrated the wheel as in #2 (if to the left is forward direction). #3B should look like #2 - on level surface, the pedals are still tilted up under my weight and it feels awkward. #3A is the "invisible elevator" I experienced on climbing, and what I called "it just worked". Effortless climbing.

 

[Chriull]

1 minute ago, Aneta said:

I see. I guess, I was wrong about angle = throttle; it's the temporal derivative of the angle (aka angular velocity) = throttle. Which makes sense, because gyros measure angular velocity (omega), not absolute angle per se (integrating angle from omega is prone to divergence due to gyro's bias).

I'll have to redo my experiments with clinometer mounted on the wheel to observe the angles at all times, not only during calibration. Stay tuned.

The input to the EUC is the pressure one applies on different parts of the pedal (== a torque on the pedal)

The EUC reacts by changing the motor current - which determines the countertorque by the motor. The torque (differences) creates an angular acceleration of the pedal.

The "setpoint" against which the control works is the difference to the calibrated angle of the angular position of the pedal.

So there are somehow "two steps of derivations" involved - from acceleration to position.

As a "sideeffect" forward/backward accelerations happen allowing us to have fun with our EUCs zipping around! 

Ps.: I'm not really sure if my above statement

23 minutes ago, Chriull said:

so they have as primary "control loop goal" to keep the pedal even

is not "oversimplified"/enough for a self balancing EUC? But could be...?

... but it's enough to setup and indentify the involved forces/torques. 

[mrelwood]

1 hour ago, Aneta said:

My understanding is that the angle of the pedals relative to "neutral" (calibrated) angle is the throttle input.

It almost is, but the applied force to tilt the pedals is a more precise description for the throttle. If you push the front of the pedals down when the wheel is facing a wall, the actual angle of the pedals may even be neutral but the wheel still tries to accelerate to counteract the force being applied.

Only if the riding mode is extremely soft (like softer than 16S ”soft” soft), does the actual change in pedal angle somewhat relate to the amount of current the wheel applies to the motor.

An infinitely hard riding mode works just like the equal and opposite force the ground applies on your feet when you stand (or jump up and down). It doesn’t require for the ground to move to apply the opposite force. You can’t make the ground to apply more force than you do, since it wouldn’t be a balanced system anymore. An EUC that applies more or less force than what is required to keep the wheel upright isn’t a self-balancing vehicle, and it doesn’t try to stay upright like EUCs do.

1 hour ago, Aneta said:

You recalibrate the lever so that at neutral position it sends the same signal to the controller as when you were pulling the lever. This is the same thing I'm doing

Not quite. No matter how you calibrate the neutral position on an EUC, the wheel only accelerates if you apply a force to tilt the pedals. 

1 hour ago, Aneta said:

- I recalibrated the throttle so that when going up, level pedals already provide the required throttle input

But to get the pedals to level you need to apply exactly the same force to tilt the wheel forward than you did when the pedals were calibrated flat! There’s no ”already”.

1 hour ago, Aneta said:

while I'm experiencing the same lack of any effort in my feet as if I was riding on a level surface at slow constant speed.

No. The effort is what tilts the wheel forward and accelerates. You can’t accelerate a self-balancing vehicle without leaning. It wouldn’t be a self-balancing vehicle. And it would be extremely hard to stay upright with one.

1 hour ago, Aneta said:

you don't have to exert any effort whatsoever.

Again, yes you do. Various aspects affect how easy the rider feels riding up a hill to be. But nothing can change the fact that all EUCs accelerate only as a response to user input.

1 hour ago, Aneta said:

#3B should look like #2

True.

[mike_bike_kite]

Oddly enough I was recently trying to come up with a report to show the quickest accelerating ECUs. To be honest I was guessing a lot with the formula to use but it's based on motor power, wheel size, voltage and battery power - I can't easily represent firmware etc.

Best accelerating EUCs

I have no idea if this represents reality and I have absolutely no idea how the Kiwano got on there. If anyone wants to have a go at a formula then I'll be happy to give it a shot.

Edited December 12, 2019 by mike_bike_kite
just changed title of report

[Mono]

6 hours ago, Aneta said:

I see. I guess, I was wrong about angle = throttle; it's the temporal derivative of the angle (aka angular velocity) = throttle.

It can't possibly be the derivative . No way one could build a calibrated controller from there. The angle is a much better model, like

5 hours ago, Chriull said:

The "setpoint" against which the control works is the difference to the calibrated angle of the angular position of the pedal.

and I am pretty sure that the torque is indeed monotonous in the angle (but not necessarily proportional to the angle) yet

4 hours ago, mrelwood said:

[...] the applied force to tilt the pedals is a more precise description for the throttle.

Regarding effort: as we apply weight force by just standing, it can be very effortless if we move to the right place on the pedal and still have sufficient footing. It all depends on foot position, pedal length and not being stressed out to lose balance.

Edited December 12, 2019 by Mono

[mrelwood]

37 minutes ago, mike_bike_kite said:

Oddly enough I was recently trying to come up with a report to show the quickest accelerating ECUs.

It would be useful to first describe what you mean by ”quickest accelerating”. Ones that start to accelerate with the least amount of effort, or ones that can actually be accelerated the fastest?

For the first it comes down to the smallest tire diameter, as long as the pedal is sufficiently long. The Kiwano doesn’t really work too well for this since it has a handlebar to limit one’s lean. Also, while the Onewheel should be excellent for this due to the unchallenged ”pedal” length, it didn’t feel very zippy when I tried one for a bit.

The firmware is quite important for both, but for the latter a lot can be calculated by just the sheer power the battery and controller can provide. Average sustained power rating of the motor doesn’t mean very much for this either.

Edited December 12, 2019 by mrelwood

[mike_bike_kite]

4 minutes ago, mrelwood said:

It would be useful to first describe what you mean by ”quickest accelerating”.

It was just a way of trying to show which wheel might accelerate of the line quicker than others. There is very little science involved in how I did it and just a lot of guess work in coming up with a formula.Are the wheels shown roughly the fastest accelerating wheels? what wheels shouldn't be on there (apart from the Kiwano) and what should be that aren't? Is there a formula that can be applied to a wheels data to say whether it will be quicker than another wheel? 

PS the report obviously isn't correct at the moment but leaving it showing allows me to get some feedback from you guys

[Mono]

7 hours ago, Mike Sacristan said:

I see now why you guys commented my FW 1.07 thread as if an EUC was entirely rigid and as if a rider was a stick without ankles.

That's one of the most powerful ideas in science. As Einstein put it very concisely:

          Everything should be made as simple as possible, but not simpler.

We make a model only as complex as necessary to explain what we like to explain. That's actually a pretty big deal. If we have the appropriate (simple) model, everything else is usually just some standard application of basic techniques.

 

7 hours ago, Mike Sacristan said:

When riding a EUC what we want to do is use force to rotate the wheel.
If a huge tractor tyre was lying down on the ground how would you rotate it 10 degrees clockwise?
Would you turn it by the middle? Or would you turn it by the edges?
If you could place a huge handle on it and twist the handle to twist the wheel.. where would you place the handle?
Closer to the center of the tyre or further from the center?

Of course, the further from the center the better the leverage. The second question is in which direction you want to push (or pull). Also an easy answer  

[Mono]

38 minutes ago, Chriull said:

The torque the rider weight "produces" is the vector product of the weight force times the vector from the axle to the "force attack point on the pedal"

I see, looks like the same what I had above. We have || v x w || = ||v|| ||w|| sin(angle between v and w) right from the definition of the cross product. I don't quite see that the direction of the resulting vector v x w here has a meaning beyond its length, does it?

1 hour ago, Chriull said:

So by tilting the pedal the angle between this two vectors and the vector from the axle to the attack point changes. If one keeps staying "on the pedal tip" while tilt back the by the riders weight induced torque increases (a bit).

Staying on the pedal tip when the pedals tilts back means a forward displacement of the CoG relative to the wheel axle, because the tip moves forward. That is exactly what would explain the amount of increased torque (static consideration). Dynamically, this displacement can only happen if the rider actively slows the wheel down to get ahead with their CoG.

1 hour ago, Chriull said:

Beside this "static" considerations the rider can de/increase the vertical force by crouching/standing up (fast enough to have a noticable effect).

That's why I call bending the knees the life saving reflex 

1 hour ago, Chriull said:

If while leaning forward while an acceleration one crouches/bends his knees so one does not stem ones inertia against the wheels acceleration while keepibg the weight force on the pedal (if such a move is in reality possible?!)

I think the whole point is to release the weight force (for long enough such that the wheel can move effortless to the desired place).

1 hour ago, Chriull said:

this could be (a part of) your proposal for trying to overcome an overlean. (Letting the wheel getting under one again by releasing sone force, but keep it acceleration)

Thinking about it, like this one could possibly put additional energy into the system (and possibly accelerate faster)? Maybe similar to surfers pumping? I still don't see exactly how the movement must look or how to compute the bilance...

[mrelwood]

1 hour ago, mike_bike_kite said:

It was just a way of trying to show which wheel might accelerate of the line quicker than others.

Well, the question remains, what exactly do you wish to calculate? Wheels don’t accelerate themselves. I think your goal leaves too much room for interpretation.

If I take a forward lean of 10 degrees on each, my 16S will accelerate faster than my MSX. If I wish to accelerate as fast as possible, the MSX will let me accelerate faster than the 16S.

1 hour ago, Mono said:

          Everything should be made as simple as possible, but not simpler.

I think this exactly is one example of simplifying too much.

[Chriull]

1 hour ago, Mono said:

I see, looks like the same what I had above. We have || v x w || = ||v|| ||w|| sin(angle between v and w) right from the definition of the cross product. I don't quite see that the direction of the resulting vector v x w here has a meaning beyond its length, does it?

You're right - the torque is a scalar. It just has a value and the sign, but no "direction".

1 hour ago, Mono said:

Dynamically, this displacement can only happen if the rider actively slows the wheel down to get ahead with their CoG.

Or the rider "accelerates" himself forward against the pedal. This would need some additional torque/acceleration from the wheel to keep the pedals tilted.

1 hour ago, Mono said:

I think the whole point is to release the weight force (for long enough such that the wheel can move effortless to the desired place).

If the whole force on the pedals is removed the wheel will not accelerate anymore...

[mike_bike_kite]

31 minutes ago, mrelwood said:

Well, the question remains, what exactly do you wish to calculate? Wheels don’t accelerate themselves. I think your goal leaves too much room for interpretation.

If I take a forward lean of 10 degrees on each, my 16S will accelerate faster than my MSX. If I wish to accelerate as fast as possible, the MSX will let me accelerate faster than the 16S.

I think this exactly is one example of simplifying too much.

In simple terms, time it takes to reach say 24 kmh from a standing start. I don't have those figures though it's possible @Seba does. I was just trying to estimate the faster accelerating wheels by looking at the specs for each wheel.

[Mike Sacristan]

1 hour ago, mike_bike_kite said:

In simple terms, time it takes to reach say 24 kmh from a standing start. I don't have those figures though it's possible @Seba does. I was just trying to estimate the faster accelerating wheels by looking at the specs for each wheel.

Some specs should be recalculated for true tyre size.
For instance the 16X and Nikola have a larger diameter than the Tesla.
The MSX has a larger diameter than the 18XL.
Then we have the weight of the wheels.

And then we have the topic of this thread. What determines wheel zippiness.
Size, power, weight. Then we have pedal size, pedal height, pedal modes (rider assistance) and also how the firmware is programmed.
The first three can at least be somewhat generalised and pedal height/size could also be included as that data is available.

Following this one can assume:
Mten3 > MCM5 > Tesla -> Nikola/16x -> 18XL -> MSX -> Monster

[mrelwood]

4 hours ago, mike_bike_kite said:

In simple terms, time it takes to reach say 24 kmh from a standing start.

That time depends mostly on how far the rider has guts to lean. 24km/h is low enough to encourage smaller diameter wheels where less lean is required making faster accelerations easier. But the actual maximum acceleration is a different measure. It depends mostly on the power of the battery and controller, but all the power in the world won’t help you accelerate fast unless you lean appropriately. A firmware behaviour that supports fast accelerations and inspires confidence is crucial, and not shown in specs in any way.

You do realize that to find out the fastest acceleration a wheel can make, one has to overlean and crash the wheel to find the limit? That’s why nobody really tries to clock the fastest accelerations, so no real data is available outside data logs of overleans.

I would probably do it faster on the MSX vs 16S, but a light rider might be faster on the 16S. Even with both of us having a lot of experience on both wheels.

Do you see now why it’s impossible to order the wheels by acceleration? Especially just by looking at specs?

Edited December 12, 2019 by mrelwood

[Mono]

13 minutes ago, mrelwood said:

But the actual maximum acceleration is a different measure. It depends mostly on the power of the battery and controller

I don't understand the point. The maximum acceleration we can only get at zero speed. That is, we talk about "time it takes to reach 1km/h" instead of reaching 24km/h?

[mike_bike_kite]

59 minutes ago, mrelwood said:

That time depends mostly on how far the rider has guts to lean. 24km/h is low enough to encourage smaller diameter wheels where less lean is required making faster accelerations easier. But the actual maximum acceleration is a different measure. It depends mostly on the power of the battery and controller, but all the power in the world won’t help you accelerate fast unless you lean appropriately. A firmware behaviour that supports fast accelerations and inspires confidence is crucial, and not shown in specs in any way.

You do realize that to find out the fastest acceleration a wheel can make, one has to overlean and crash the wheel to find the limit? That’s why nobody really tries to clock the fastest accelerations, so no real data is available outside data logs of overleans.

I would probably do it faster on the MSX vs 16S, but a light rider might be faster on the 16S. Even with both of us having a lot of experience on both wheels.

Do you see now why it’s impossible to order the wheels by acceleration? Especially just by looking at specs?

I'm afraid I still disagree. We can measure 1/4 mile times on motorbikes yet riders weigh different amounts, each rider also has different skills in controlling wheel spin and making fast gear changes plus each rider will take different amounts of risk. For range and top speed figures for wheels we use a 70kg rider so why not just use this same mythical rider? In my report I'm just estimating by looking at the available figures but, if you were particularly interested,  you could build a machine to test acceleration. Another method is just to look at acceleration times reported from wheellog etc - obviously riders will weigh different amounts, be riding in different conditions and some more cautious than others but given enough data you could make a reasonable stab at which wheel was faster accelerating. The extreme option might be to hand each wheel to kuji and tell him it would make a great youtube video if he tried to discover the fastest accelerating wheels. The out takes on the video would make compulsive viewing and he has the advantage that he looks like he weighs 70Kg.

[jonm42]

3 minutes ago, mike_bike_kite said:

The extreme option might be to hand each wheel to kuji and tell him it would make a great youtube video if he tried to discover the fastest accelerating wheels. The out takes on the video would make compulsive viewing and he has the advantage that he looks like he weighs 70Kg.

THIS!

[Mike Sacristan]

And then once the wheels are ordered by acceleration and it turns out the the MSX is the one with the fastest acceleration but they didn't read the foot note.
You just have to lean like this:

A lot of talk about simplifying things and then making them over complicated. 

Rider John Smith with a weight of X using a perceived effort of potato perceives the wheels as zippy in the following order.
Tomato, banana, apple and lastly watermelon.

[Cory Brown]

17 minutes ago, Mike Sacristan said:

And then once the wheels are ordered by acceleration and it turns out the the MSX is the one with the fastest acceleration but they didn't read the foot note.
You just have to lean like this:

A lot of talk about simplifying things and then making them over complicated. 

Rider John Smith with a weight of X using a perceived effort of potato perceives the wheels as zippy in the following order.
Tomato, banana, apple and lastly watermelon.

You're nuts if you think apple is zippier than watermelon!

[Mike Sacristan]

@Aneta i'm beginning to wonder if you really own a wheel or if you just are here to discuss physics. 
And @Mono has a very mesmerising avatar which I can't stop looking.

When climbing a slope the contact patch is further toward the front of the tyre. This puts the bias of our weight behind the center of the wheel.
This has been calculated and illustrated on this forum before.
To overcome it you need more forward lean. Tilting the pedals backwards will cause a dorsiflexion penalty making it very physically demanding so that only a few people will be able to keep their heels on the pedals. Olympic squatters come to mind.

I made figure 3 just for laughs to illustrate forward mass distribution with pedals tilted backwards.

Like I said in the other thread.. I understand the misunderstanding because when I tilted my pedals back many months ago I did it in the belief that it would emulate falling to neutral and hence I would get "effortless acceleration" because I had an X degree advantage (because gravity you know). As it has now been explained by others backward tilting simply changes the position of the pedals and the whole wheel and for our bodies a pedal decline worsens the lever making forward movement more demanding but braking easier. This is easy enough to try in real life by standing on your heels and then squatting vs standing on the balls of your feet and squatting. Or do it on a -7 + 7 decline/incline or a wooden board.

[Mono]

2 hours ago, Mike Sacristan said:

And @Mono has a very mesmerising avatar which I can't stop looking.

Thanks!