Ninebot z10 creeping forward

[Chriull]

On 11/17/2018 at 8:05 AM, Alex_from_NZ said:

This is balancing it level, letting go it does the same thing so I know it's not how I'm holding it or anything. 

It's not the ground level either as it will do it in any direction on the same bit of floor. 

The pedals are level according to my spirit level, no angle there. 

 

The EUC works on the principles of an inverted pendulum (https://www.youtube.com/watch?v=MWJHcI7UcuE). Main difference imho is, that for an inverted pendulum like in the video the aim is to keep the pendulum upwards (angle of the pole), while with an EUC the aim is to keep the pedals at a specific tilt angle (gyroscope on the motherboard). By this the rider is the "final control instance" to accelerate or decelerate (== control the EUC).

So to keep the EUC standing still without a rider on top the center of gravity (cog) of the parts of the EUC ("after the axle, not touching the ground") would have to be exactly in vertical alignment with the axle for the choosen pedal tilt angle (angle reported for this by the gyroscope on the motherboard).

This is technically just impossible - every dust particle would change this again. The gyroscope itself provides a digital output (quantisation error), has limited accuracy and additionally some drift. And noone of us consumers could anytime calibrate an EUC precise enough, anyhow...

Just by the difference in the control loop (primary aim "dampened acceleration control" instead of holding the pendulum upwards) most EUCs have the "chance" to more or less stand still. 

Also just tried if my KS16S has this "creeping atitude": i'd say minimal but yes. Just by the smaller tire i have to stabilize it and by this i hold it normaly unconsciously in place - so without concentrating especially on this one would never notice anything.

My conclusion would be that this is a normal behaviour (maybe additionally amplified by the asymetric design) and no fault in any way. Just if this "asymmetric" behaviour hinders/interferes trollying or other normal modes of operation one should contact the reseller for repair/replacement.

 

[meepmeepmayer]

Looking at the Z, the real question is: why does not every Z drift?

Image for clarification. Note the "slight" asymmetry:

I would file this under "That's the price for a sexy design". But who knows, maybe it does indicate something if one Z drifts and the next doesn't. One can never be to careful with EUCs.

[Chriull]

12 minutes ago, meepmeepmayer said:

Looking at the Z, the real question is: why does not every Z drift?

Image for clarification. Note the "slight" asymmetry:

This asymmetry is compensated by the weight distribution.

By this imo if one tries to calibrate the pedal to some forward or backward tilt the "creeping behaviour" should be worsened.

Quote

I would file this under "That's the price for a sexy design".

including the wide tire - imo without this it could never have the side to side stability to make the drift so easy to be experienced.

Quote

But who knows, maybe it does indicate something if one Z drifts and the next doesn't. One can never be to careful with EUCs.

... this topic just reminded me of @Hirsute getting dragged around the stage by his EUC again : (right at the beginning of the video)

 

[Mono]

1 hour ago, Chriull said:

Just by the difference in the control loop (primary aim "dampened acceleration control" instead of holding the pendulum upwards) most EUCs have the "chance" to more or less stand still. 

I don't understand this sentence. If we consider that the EUC or inverted pendulum is supported by a single point, it will just fall if the controller does not act and move the supporting point. The only way I see how the EUC motor does not need to move is that the contact patch of the tire is long enough such that it is under the CoG of the entire wheel in its current position.

It seems possible though that the controller has a stand-still mode (or it could also be a keep-current-speed mode), which is active as long as the pedal angle isn't changed from the outside. I can't quite imagine that this is the state of current standard technology though.

[Chriull]

 

54 minutes ago, Mono said:

2 hours ago, Chriull said:

Just by the difference in the control loop (primary aim "dampened acceleration control" instead of holding the pendulum upwards) most EUCs have the "chance" to more or less stand still. 

I don't understand this sentence. If we consider that the EUC or inverted pendulum is supported by a single point, it will just fall if the controller does not act and move the supporting point.

Imo any EUC (except (1)) will move if it could be frictionless stabilized to the sides. But in normal vertical state so little that it is (almost) not noticable when held. And if it's leaned (more or less carefully) against something the friction stops any movement.

54 minutes ago, Mono said:

The only way I see how the EUC motor does not need to move is that the contact patch of the tire is long enough such that it is under the CoG of the entire wheel in its current position.

(1) Or since the contact patches of the tires and of the axle are not punctual  and additionally by the friction at the contact patches a state can be found where the EUC can just stand still. Not really a simple task but possible.

54 minutes ago, Mono said:

It seems possible though that the controller has a stand-still mode (or it could also be a keep-current-speed mode), which is active as long as the pedal angle isn't changed from the outside. I can't quite imagine that this is the state of current standard technology though.

Me neither - as i could not imagine how to distinct between an "outside" pedal change or pedal change by gravity on the wheel itself. Would need some pressure sensors which would overly and senselessly complicate an EUC?

This nice stand still behaviour is imo just a consequence of this above rough idea of this "different" control loop. I'd bet that if one fixes a pole with a weight on the top onto the EUC and prevent that the EUC can fall sidewards (?frictionless?) it will not be able to balance the pole. (... but not too much  )

My guess is that if the control loop would be so "reactive" like for the inverted pendulum the balancing efforts of the rider would easily lead to overshots (oscilation issues) dumping the rider.

[Mono]

2 hours ago, Chriull said:

This nice stand still behaviour is imo just a consequence of this above rough idea of this "different" control loop. I'd bet that if one fixes a pole with a weight on the top onto the EUC and prevent that the EUC can fall sidewards (?frictionless?) it will not be able to balance the pole. (... but not too much  )

You mean a fixed pole? If the pole can fall on it's own of course the EUC can't know how to balance it, as it can't know the current angle of the pole. If the pole is fixed I take your bet.

2 hours ago, Chriull said:

My guess is that if the control loop would be so "reactive" like for the inverted pendulum the balancing efforts of the rider would easily lead to overshots (oscilation issues) dumping the rider.

IMHO rather unlikely, if I understand the setup correctly. I'd rather expected a "run-away" with some stationary velocity, like without the pole. Kind-of easy to test.

[Chriull]

47 minutes ago, Mono said:

You mean a fixed pole? If the pole can fall on it's own of course the EUC can't know how to balance it, as it can't know the current angle of the pole. If the pole is fixed I take your bet.

fixed. The free part is the shell/pedals on the motor axle. Maybe someone with beginner wheels (if they are "stable" enough) gives this a try 

47 minutes ago, Mono said:

3 hours ago, Chriull said:

My guess is that if the control loop would be so "reactive" like for the inverted pendulum the balancing efforts of the rider would easily lead to overshots (oscilation issues) dumping the rider.

IMHO rather unlikely, if I understand the setup correctly. I'd rather expected a "run-away" with some stationary velocity, like without the pole. Kind-of easy to test.

I meant that "counter balancing" actions of a panicked rider could lead to this "overshots". Maybe also instinctive balancing actions of unexperienced riders could lead to unwanted results.

Did you ever see beginners on a segway getting the "shake" - looks frightening. They are more or less in standstill and get the segway driving a small distance back and forth with insane accelerations/movements. (Would be an argument for the "same" control loop?... *hmmmm* )

 

[Chriull]

3 hours ago, Mono said:

I'd rather expected a "run-away" with some stationary velocity,

This seems to be a chance to get back to the topic again. With this stationary velocity beeing the general solution to keep the pole (rider/wheel) balanced without external influences and standstill as one specific solution.

In a real world with imbalance, friction, not flat surfaces and wind resistance this specific solution will always result instead of standstill in some specific veloctity to balance the forces.

[Mono]

59 minutes ago, Chriull said:

In a real world with imbalance, friction, not flat surfaces and wind resistance this specific solution will always result instead of standstill in some specific veloctity to balance the forces.

Yet this assumes that the support is a single point which it isn't, in particular if the tire has thread grooves traversing the middle. Also small roughness of the ground has a similar effect and can stop the movement at a mini-incline (and not force a velocity as you suggest).

In any case, empirical evidence already suggests that your model can not be correct: many wheels seem not to creep. My V8 keeps (almost?) standing upright even with the motor off and blocked tire. With the motor on it sometimes creeps and sometimes doesn't.

 

If it disagrees with experiment, it's wrong. And that simple statement is the key of science. It doesn't make a difference how beautiful your guess [i.e. theory/model] is. It doesn't make a difference how smart you are, who made the guess, or what his name is; if it disagrees with experiment: wrong. That's all there is to it.        -- Richard P. Feynman
 

[Alex_from_NZ]

Hi Everyone.  

So i installed the mudguard and trolly handle to test and as expected it still creeps forward (accelerating as it does so).

I have not done anything else about it yet as i am still unable to walk.  Looks like ive done in a few ligaments argg.

I tried to get back on it yesterday to do some more practise but couldnt so its sitting in my garage at the moment until im able to do anything other than sit here lol. 

Im still undecided what to do..  I dont think in riding scenarios it will cause me any problems, its more the opportunity for it to runaway at speed if i fall, or for it to not be able to be leaned against something while turned on.

[Mono]

4 minutes ago, Alex_from_NZ said:

or for it to not be able to be leaned against something while turned on.

That's a good point, it's something I do once in a while. Even if it doesn't move in this scenario it would probably drain the battery much faster. You should be able to check in the app.

[meepmeepmayer]

7 minutes ago, Alex_from_NZ said:

I have not done anything else about it yet as i am still unable to walk.  Looks like ive done in a few ligaments argg.

Woot, what did you do? It's not just from your riding training as seen in your video?

Pls go to a doctor

7 minutes ago, Alex_from_NZ said:

Im still undecided what to do..  I dont think in riding scenarios it will cause me any problems, its more the opportunity for it to runaway at speed if i fall, or for it to not be able to be leaned against something while turned on.

Did the seller say anything?

I don't have many ideas other than opening it up (make sure that does not affect warranty) and hoping you find a secret gold treasure or just drugs someone has hidden in there, which make for an unusually uneven mass distribution.

[Alex_from_NZ]

3 minutes ago, Mono said:

That's a good point, it's something I do once in a while. Even if it doesn't move in this scenario it would probably drain the battery much faster. You should be able to check in the app.

It does, but only one of the batteries.... which is one of the reasons i first thought this was fault related instead..

I posted a screenshot of the battery pages from the app on the page back, batt one is discharging but batt 2 is not. :/ 

[Alex_from_NZ]

2 minutes ago, meepmeepmayer said:

Woot, what did you do? It's not just from your riding training as seen in your video?

Yep, from the accident in that video.  I landed on my left leg with leg locked straight and it twisted the wrong way and to the side.  

I should have stopped there, but i thought of Side Street Reny learning to ride and thought, hes older than me, i can keep pushing myself too!

I can not i found out. lol.

2 minutes ago, meepmeepmayer said:

Pls go to a doctor

Been, on crutches and drugs.. Not the good ones though grr. Had an Xray and got an MRI scan booked for tomorrow. Wish me luck!

2 minutes ago, meepmeepmayer said:

Did the seller say anything?

Have not spoken to them yet, getting it there with my current condition will be near impossible.  Driving is... difficult right now plus my cars ignition coil needs to be replaced but i cant bend over enough to get at it atm lol.  When it rains it pours right?  

2 minutes ago, meepmeepmayer said:

I don't have many ideas other than opening it up (make sure that does not affect warranty) and hoping you find a secret gold treasure or just drugs someone has hidden in there, which make for an unusually uneven mass distribution.

I think i will leave it unopened, purely as its less than 2 weeks old and only been riden once at this stage i would rather let them deal with it if i decide to do anything.  I prolly should take it back, but maaaaaan i just want to ride it sooo much!

[meepmeepmayer]

Talk about bad luck... Well, good fortunes at your doctor. It's probably nothing serious, just something lengthy...

I guess when you know more you can decide whether returning it makes sense (if you can't ride anyways for longer and it has the creep) or you just need to wait for your knee to work again and that would be it.

[novazeus]

On 11/18/2018 at 3:33 AM, Alex_from_NZ said:

Im off work to go get a scan on my knee tomorrow, so will likely try install the mudguard and handle just to see if there is any difference like some have said.  I hope this is all it is but i have my doubts..

Would be great if someone who owned the wheel could chime in.. Has anyone got their wheel without the mudguard and handle installed noticed this?  It is rather noticeable.. ie, lean it against something and it will try creep forward or end up rocking back and forth..

I hope im not breaking etiquette here, but im going to tag some people i know who own this wheel into this to see if they can lend their experienced viewpoints.

@Marty Backe @houseofjob @Lukasz @fbhb (i know we discussed it previously) @RoberAce @novazeus 

pretty sure there’s no creep on mine.

how did u hurt yourself?

[novazeus]

no, mine does creep but i figured out how to set the pedals level and the creep is almost gone.

[Chriull]

6 hours ago, novazeus said:

no, mine does creep but i figured out how to set the pedals level and the creep is almost gone.

This is imho the solution to get rid of the creep - change pedal angle (by calibration i assume?) so the asymetric casing is "balanced". But this could in the worst case lead to "uncomfortable/strange" pedal angles if the casing is badly balanced?