Jump to content

Ninebot Z 'corner assist'? and 'posture control'?!


Planemo

Recommended Posts

Guys, having had a good chance to ride my Z6 recently, I noticed that the footplates dip forward when making tight turns. Is this normal?! If yes, is it supposed to help?

I never noticed it happening on my V5 at all. Not sure if I like it. It's not the end of the world, but I would be interested if it is possible to turn it off/adjust it.

Also saw the option in the app of calibrate posture something. Is this a left/right tilt calibration? Purpose? Is it for the sensor to shut down when fallen, or does it affect the above?

Link to comment
Share on other sites

Weird! You say it's unintentional but unless I am going mad the wheel seems to do it intentionally - I notice because when I pull out of the turn I am sure I can feel the footplates returning to level (which on mine is about 0.4 deg tilt back).

Is it all in my head then?!

I probably would never have thought about it if I hadn't ridden the V5 first, and given the choice I would like to try it without. If only to give a little more clearance on the plates. I wondered if the posture calibration would make any difference - I have already done the manual pitch (footplate level) calibration and it's bang on where I want it at the mo.

Link to comment
Share on other sites

The sensor still thinks it is is level when the dipping happens.

This is the same effect like with a badly calibrated wheel: pedals dip in turns because the sensor gets confused by the combination of sideways lean and bad calibration, making it think dipped forward is level.

I had the exact same thing on my ACM. Sudden dipping in turns, nice return to level coming out of a turn. One good calibration later, the pedals are always level, turn or not.

This isn't a feature, but they make the best of it by calling it corner assist. That's how I interpret this.

 

Link to comment
Share on other sites

We've had this with other brands a few years ago, implying (but not outright stating, we never quite knew what was going on) that the dipping in curves is some kind of assist thing, when it is just a weakness of the tilt sensor getting confused by bad calibrations. Now they just calibrate their wheels more carefully (or did some firmware tweaks or whatever) and there's no dipping.

The fact that the dipping is highly annoying and counterproductive (especially for riders just learning to ride!) and that dipping is no longer a thing now pretty much proves this was never a feature in the first place.

Inmotion and Ninebot seem to be behind the curve in this respect and repeat the same stuff.

(By unintentional I mean they wouldn't have this in the first place if it wasn't a problem with the hardware they couldn't 100% get rid of. It's a bug, not a feature.)

Link to comment
Share on other sites

Even weirder! Thanks for your help. So the wheel IS forcibly making the pedals dip....but not accelerating the wheel....how odd. Not sure I can get rid of it like you did with your ACM. As I say the pitch is calibrated very well, the only thing I am unsure of is the roll calibration. That said, the dip in turns seems pretty equal on both sides - it's not like it's bad one way and not the other.

Link to comment
Share on other sites

5 minutes ago, Planemo said:

So the wheel IS forcibly making the pedals dip....but not accelerating the wheel....how odd.

The wheel doesn't know the pedals are dipping. It thinks it is as level as it always is.

Nor sure what options there are on the Ninebot Z and what calibrations. But you can always try a nice calibration (wheel not tilted sideways and still while doing it) and see what happens.

Link to comment
Share on other sites

Meep - I saw in the other thread that pedal dip is actually a geometric consequence of turning a single wheel. Interesting. So if this is the case, those wheels that dont must actively apply tiltback to bring the plates back into a flat plane relative to the ground. I will look further into it.

Stephen: cheers, but I already ride at setting 1. I guess I could try setting zero but I am not sure it will make a fat lot of difference given the amount of tilt. I also run without brake assist.

Link to comment
Share on other sites

2 minutes ago, Planemo said:

Meep - I saw in the other thread that pedal dip is actually a geometric consequence of turning a single wheel. Interesting. So if this is the case, those wheels that dont must actively apply tiltback to bring the plates back into a flat plane relative to the ground. I will look further into it.

Stephen: cheers, but I already ride at setting 1. I guess I could try setting zero but I am not sure it will make a fat lot of difference given the amount of tilt. I also run without brake assist.

Back to the drawing board 😃 as they say! 😊

Link to comment
Share on other sites

12 minutes ago, Planemo said:

Meep - I saw in the other thread that pedal dip is actually a geometric consequence of turning a single wheel. Interesting. So if this is the case, those wheels that dont must actively apply tiltback to bring the plates back into a flat plane relative to the ground. I will look further into it.

I'm not 100% sure of this.

If we define dip as "any continuous nonzero pitch angle", then there shouldn't be a dip. That would be a technical ideal of how a EUC should be, that's how they work after all. Whether that is ideal for riding and you add some behavior tweaks is another question.

But probably it's much more complicated geometrically what really happens, you can't clearly separate pitch and roll, pedals are not on the center axis, nothing's perfect, etc. What was the question again?:whistling:

I can only give my experience from my wheel and what people posted here. That is: pedal dipping in turns is shitty (majority opinion unless we're talking about different things) and, since you can get completely rid of it by doing a nice calibration and nothing else, it's not intentional (a feature), so it must be unintentional (a bug) due to a technical limitation.

Link to comment
Share on other sites

Right, I had a think on this last night, I believe I got to the bottom of it and having just read the other thread I think @esaj is on the same thought process.

I put my (static) wheel on the floor, angled it to around the same roll as a tight turn and then turned the wheel slightly left, simulating a left turn. The contact patch of the tyre moves fowards. At the same time, the front of the pedal dips very slightly due to the compound angles involved. However, the dip wasnt as much as I am feeling. So whats happening? Then it became apparant that I was making my observations on a static wheel...

I think its this - the only way to maintain the same approach speed in a tight turn is to move the CoG fowards to match the new, slightly foward contact patch. Theres only two ways to do this - either the rider applies more weight to the front of the pedal of the wheel does it for you. This is where (and I think you are going to hate me for this meep) we are not giving the Ninebot engineers more credit then they deserve. What I believe they have done is write into the firmware a multiplier, taking a figure from the roll sensor and adding it to the pitch sensor. Thus, the more a wheel rolls, the more it pitches, foward or back depending on the wheels direction, which it also knows.

To try out this theory, I got back on my V5 and did a same radius turn as I did on the Z. Voila - I had to apply noticeably more weight to the front of the plate during the turn to maintain the same entrance velocity than the Z. In fact, from testing I found that I am not adding anything at all to the Z. Probably why, on my first ride on it after the V5, I felt like it was 'turning itself' during tight turns with very little effort from myself.

I hadnt noticed I was putting more weight onto the front of the V5 plates when turning during learning...us as humans just do it subconsciously to maintain balance (arent we great!) but it was only when going to a different wheel (with a roll multiplier) that it became apparant.

This would also explain why, when coming back out of the turn to zero roll quickly on the Z, I can still feel the plates leveling out even whilst vertical - the FW is 'catching up' albeit slowly.

Now you could argue whether this 'corner power assist' is desirable or not. I certainly think it should be adjustable or at least switchable. On the one hand I really like the slow speed turning of the Z, it was something I noticed right off the bat and couldnt at the time understand how such a heavy wheel could 'drive' itself through the turn with what seemed like zero effort from myself. On the other hand, because the wheel is adding artificial CoG changes itself and not me, the plates dip more, (because I am more upright) reducing plate clearance. Which I am not keen on at all.

This is all only my thoughts on it. I may be wrong. But if correct, I can see why Ninebot and others have done it when you compare it to a wheel without (say the V5). But I am not sure if their efforts were...well....worth the effort.

To answer as to why the wheel (ACM?) you had (meep) was ok after a calibration, I am thinking that maybe it was a wheel that did have the multiplier, but it either wasnt adding as much as the Ninebot FW does (hence the pedals felt 'flatter' after calibration but werent actually level) or because the roll calibration was incorrect left to right, meaning the wheel was adding more multiplier to one side hence you noticed a reduction to one side after calibration. In any event, the wheel must have had some sort of roll multiplier otherwise there wouldnt have been any change during cornering at all after calibration. Maybe the ACM just doesnt have as an agressive multiplier as Ninebot.

Thats my take on it, and it does appear that theres not much I can do about it on the Z. As I say, its not a massive problem, and it does have benefits, but it would be nice to try it without :)

Link to comment
Share on other sites

Nice description of the wheel geometry in turns, thanks. Now we know:)

And I certainly don't hate you, I simply don't trust the manufacturers.

Whether the pitch modification by roll is intentional in firmware or a consequence of a confused sensor, it's the exact same effect, so just from its existence we don't know.

The fact that there's no option to turn it off is (to me) another indication that they couldn't turn it off if they wanted to, if you know what I mean;)

But it's nice that you actually like the behavior, whether feature or bug. No problem then :)

(My ACM never had any asymmetric behavior as far as I noticed, just crazy dipping in turns and after a calibration everything was perfectly fine. But I guess a roll calibration is exactly what happens if you calibrate a wheel when it's tilted sideways, even though it really shouldn't.)

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...