RC bike counter-steering demo

[janssnet]

 

In an attempt to create an RC LEGO motorbike, I had to figure out a steering method. Moving a weight from left to right was (successfully) done before, but I couldn't find LEGO bikes with counter-steering.

Inspired by the videos of many real RC bike lovers, I came up with this LEGO-ish implementation. It's not for LEGO purists, it contains modified parts. But it demonstrates pretty nice how counter-steering works and how it can be implemented on a LEGO bike (using a servo). I therefore thought it might be of interest to some of you.

If modifying LEGO parts makes you sick, please skip this video.

If you enjoy creating new parts (out of other LEGO parts), great! Let me hear your thoughts.

On this matter, I personally start to enjoy the use of custom springs more and more (will do a separate video on this subject) and ...

I'd love to make a case for an axle with one ball socket. Together with a (custom) spring, can be used in almost any vehicle for suspension or anything else. Have a look at the video and you'll see what I mean. Was an essential element to create this steering.

Enjoy watching, looking forward to hear your comments.

https://youtu.be/AZQkJCd0VKg

 

 

Edited January 7, 2021 by janssnet

[Gray Gear]

Very cool, I like it! 

[BrickBuildingFox]

Very impressive!

Edited January 4, 2021 by BrickBuildingFox

[MajklSpajkl]

Very interesting. I wasn't even aware of this concept. Regarding parts altering - to each their own. No judgment ;-) Have you tried with LEGO original parts only?

 

[janssnet]

3 hours ago, MajklSpajkl said:

Very interesting. I wasn't even aware of this concept. Regarding parts altering - to each their own. No judgment ;-) Have you tried with LEGO original parts only?

 

Yes, ran into 2 problems. The strength of the springs need to be very accurate to enable the front wheel to move freely, the original LEGO springs are too stiff for this application. Secondly, I couldn't find a way to fit the standard spring in the (very limited) space available. And to be honest, I enjoy these custom springs very much. Lots of useful and useless possibilities. The most useless is a suspended seat on the Bobber-bike, since it's hardtail. Had to consider the convenience of the driver :))

[kieran]

Works really well, look challenging to control, 

[Bartybum]

Awesome work, thoughts on adding a gyro sensor and making a PID controller handle the countersteering/balancing? I did something similar in Garry’s Mod a couple years back and it works remarkably well

[janssnet]

9 hours ago, Bartybum said:

Awesome work, thoughts on adding a gyro sensor and making a PID controller handle the countersteering/balancing? I did something similar in Garry’s Mod a couple years back and it works remarkably well

Interesting thought! Have a gyro available from an old car project.

With the gyro installed the connection to the steering servo would not require springs, can be done with fixed axles. Will be interesting to see if this reacts as quick as a self-stabilising bike to keep it straight. Have never seen it on the real RC motor bikes though .... 

[Bartybum]

51 minutes ago, janssnet said:

With the gyro installed the connection to the steering servo would not require springs, can be done with fixed axles.

You could use fixed axles, but you’d wanna add some form of suspension to help with vibrations in the system. A similar problem I found was that the kP and kD constants had to be reduced the faster you went, or else you’d suffer from speed wobbles and overcorrection.

52 minutes ago, janssnet said:

Will be interesting to see if this reacts as quick as a self-stabilising bike to keep it straight.

If done right it should react quicker. I had my PD controller scheme target a roll value based on user input.

53 minutes ago, janssnet said:

Have never seen it on the real RC motor bikes though ....

Minus the robot, this is a pretty cool example:

You don’t see it in RC bikes because they always use flywheels to stabilise themselves. Much mechanically simpler

Edited January 5, 2021 by Bartybum

[janssnet]

11 hours ago, Bartybum said:

You could use fixed axles, but you’d wanna add some form of suspension to help with vibrations in the system. A similar problem I found was that the kP and kD constants had to be reduced the faster you went, or else you’d suffer from speed wobbles and overcorrection.

If done right it should react quicker. I had my PD controller scheme target a roll value based on user input.

 

Did a quick setup. Took a car gyro (only one axis) mounted it vertically on the bike. Put it on reverse, so when the bike falls over to the left, the steering turns also left, to compensate (counter-steering) and to put the bike straight again.

Made a quick video, please find it here.

Conclusions:

- It works! Although less stable when going straight than without gyro (which might be a matter of adjusting the sensitivity of the gyro), the steering becomes far less unstable. It almost works like 'going back to center' steering. 

- Springs continue to be necessary for the steering axles. Tried without, but it makes the bike very very sensitive for any change in steering direction by the servo. Almost impossible to steer.

Seriously considering to put it on the Bobber-bike, for which every bump in the road currently seems to be too much. What do you think, will the gyro stabilise for bumps in the road?

Thanks again for your suggestions. Great stuff!!

 

[Brickend]

This is great work - I was recently playing with motorbike geometry myself, and used part 45590 in the headtube to control the steerer movement and to provide some return to center; would this add any benefits to this setup, or could possibly be used to add dampening in place of springs?

[Bartybum]

@janssnet any chance at a YouTube video? It won’t let me access the Google Drive

4 hours ago, janssnet said:

What do you think, will the gyro stabilise for bumps in the road?

Bumps in the road are probably better dealt with using softer suspension rather than a software solution I feel.

[janssnet]

8 hours ago, Brickend said:

This is great work - I was recently playing with motorbike geometry myself, and used part 45590 in the headtube to control the steerer movement and to provide some return to center; would this add any benefits to this setup, or could possibly be used to add dampening in place of springs?

Not exactly sure how you did it. Can you share a picture?

For the front suspension of the upcoming Bobber-bike, I came up with this idea. Not 100% happy yet, may have to loosen the springs a bit (as Bartybum suggested). Have a look here

 

Edited January 6, 2021 by janssnet

[Bartybum]

Hehe this is coming along really nicely. Looks quite good. Can you document what the effects of changing the damping constant do? Better yet, can you adjust their values on the fly with a potentiometer?

[Brickend]

4 hours ago, janssnet said:

Not exactly sure how you did it. Can you share a picture?

For the front suspension of the upcoming Bobber-bike, I came up with this idea. Not 100% happy yet, may have to loosen the springs a bit (as Bartybum suggested). Have a look here

 

I just quickly mocked this up to hopefully explain the concept.

The yellow connectors represent the handlebar, which is attached to a 5L grey axle (fork steerer) that runs through 45590 in order to provide resistance and a degree of return to centre, whilst damping out some movement. 

Of course there are multiple ways to build this and can be tuned with multiple 45590s add more resistance. The other way I've tried is to put a 2l liftarm on the steerer and have it turn against 45590 pieces to each side of the steerer if the attached configuration is not suitable.

Without looking at the system as a whole it's hard to know if this is of any use to you, but thought I'd add it to the discussion anyway!

 

[janssnet]

8 hours ago, Brickend said:

I just quickly mocked this up to hopefully explain the concept.

 

 

 

Thanks for sharing. I'm afraid this is not going to work. The bike will not be able to drive in a straight line.

The 45590 is to stiff to allow the front wheel to move freely to keep the bike straight. It's hard to explain, requires serious physics. Bottom-line, if the front wheel has a castor-angle (approx. 30 degrees) and is able to move freely, it will automatically keep the bike straight (at a certain minimum speed).

Hope this helps.

[Bartybum]

5 hours ago, janssnet said:

it will automatically keep the bike straight (at a certain minimum speed). 

A more accurate term would be that it stops the bike from falling over. It won't go straight (because free caster steering alone (with no counterweights) only provides passive stability), but it'll rest in a turning/spiral state and won't go unstable. To go straight you need a feedback controller

Edited January 7, 2021 by Bartybum

[amorti]

@janssnet Is this project still alive?

I've been working on this one which initiates counter steering in a different way. It works well, but it's clumsy, and the moving steering neck means you couldn't put much bodywork on there.

I didn't share it here on EB since you can't build it without CaDA alternating beams (Lego still has only 11&15 lengths, and for this you need at least 2x 7's.

Your bike has the front wheel mounted one stud ahead of the steering pivot (I think?).

In all tests I've done, the wheel needs to be mounted 1 to 2 studs behind the pivot to have any chance at being stable.

It's not realistic, however you will see same in most/all RC bikes.

Maybe give it a try, see if you gain that missing stability?

[aFrInaTi0n]

Onboard footage of @amorti's Café Racer:
 

 

The wobble of the front fork is a consequence of having the Insta360 Go2 camera in the back of the back, resulting in the cog also being moved + having less weight at the front fork (specially in acceleration situations). Without the camera-equipment mounted, it rarely wobbles.

Top speed meassured with SkyRC GPS probe is 28 kph / 17.4 mph

Edited June 5, 2022 by aFrInaTi0n