• Welcome to PiBoSo Official Forum. Please login or sign up.
 

Countersteering Controller Suggestion

Started by August1, October 12, 2016, 06:31:11 AM

Previous topic - Next topic

August1

October 12, 2016, 06:31:11 AM Last Edit: October 12, 2016, 06:45:35 AM by August1
I don't know really know which subforum to put this in but playing GPbikes and the recent Ride 2 with a crappy gamepad with an annoying deadzone has got me thinking about this again and it's this:

Isn't it possible and relatively easy and cheap to simulate countersteering (i.e. the ONLY way bikes turn above ~10 mph) by mounting a handlebar to a non-rotating torque sensor (transducer?)? There would obviously be no feedback through the bars whatsoever but 90% of the time when you're riding on the streets, there's little to no "feel" through the bars anyway right?

Anyways please excuse the complete lack of technical knowledge on my part but it's just something that I wanted to get off my chest.

HornetMaX

I don't think one can say there's little to no feel coming through the bars in real life.

Anyway, in principle it is possible to have a "countersteering controller": use direct steer angle (http://forum.piboso.com/index.php?topic=28.msg43#msg43), eventually on an input device with force feedback (e.g. a steering wheel). I think a couple one this forum have tried to build dedicated hardware to do that, as for bikes the steering range is very small compared to cars.

August1

When I say little to no feel, what I really mean is that you don't really get any sense of the bars turning when you steer unless the front is getting light or you're on the edge of grip limits. Most of the time the bars just feel pretty "solid" no matter how hard you push on them.

As for direct steer angle, it's a bit of a red herring for me. It's not implemented realistically in GPBikes and there are too many parameters to work with. What I'm talking about is simply converting torque forces from the stationary handlebar directly into standard left/right steering commands. Not exactly realistic I know, but seems like a much more realistic solution than rotating handlebars.

HornetMaX

Quote from: August1 on October 12, 2016, 07:12:53 AM
As for direct steer angle, it's a bit of a red herring for me. It's not implemented realistically in GPBikes and there are too many parameters to work with.
Direct steer angle = 1 parameter (max steering angle, pretty trivial).  Too many ?!

Quote from: August1 on October 12, 2016, 07:12:53 AM
What I'm talking about is simply converting torque forces from the stationary handlebar directly into standard left/right steering commands.
And how would you do that ? Convert a torque into what exactly ? A target lean angle (guess that's what you call standard steering command) ? That wouldn't really make sense to me.

Quote from: August1 on October 12, 2016, 07:12:53 AM
Not exactly realistic I know, but seems like a much more realistic solution than rotating handlebars.
Real handlebars do rotate (not much, agreed, but they do).

teeds

Also don't forget that although the bars don't rotate much, the bike beneath you does and I think the likes of DD's controller setup does imitate this well.

August1

Quote from: HornetMaX on October 12, 2016, 07:25:08 AM

And how would you do that ? Convert a torque into what exactly ? A target lean angle (guess that's what you call standard steering command) ? That wouldn't really make sense to me.


On a real bike it requires constant pressure on the bars to maintain lean so this makes sense to me.

doubledragoncc

I have tested this kind of system and it is never going to work due to the fact that it is far to sensitive and allows almost no control but just a flopping from side to side due to over sensitivity. The mechanical principle is piss easy but the fact that you are NOT riding along the road with the forces of nature means until those are simulated you are stuck with the basic concept we use today.

DD
GPBOC Live Streams: https://www.youtube.com/c/IASystemsComputerControls; i7 12700K 5.1GHz Z690 ASUS Strix Z690-A Mobo 32GB 3600MHz DDR4 RAM ASUS Strix RTX3080 OC 10GB DDR6X ASUS Ryujin 360 AOI Cooler ROG Thor 1200w PSU in ROG Helios Tower Case.

HornetMaX

Quote from: August1 on October 12, 2016, 07:56:14 AM
Quote from: HornetMaX on October 12, 2016, 07:25:08 AM

And how would you do that ? Convert a torque into what exactly ? A target lean angle (guess that's what you call standard steering command) ? That wouldn't really make sense to me.


On a real bike it requires constant pressure on the bars to maintain lean so this makes sense to me.
Except the relationship between input torque (what you call pressure) and lean angle is really not as simple.
Anyway, I think it will actually feel weird to push/pull a bar that doesn't move (on top of what you already noticed, that if the front shakes, you'll feel nothing).
But if you can try it, give it a go.

grimm

Hmmmmm, what if the pressure sensor, if it could work, was used to control the left-right rider lean? Rather than the bike steering itself.  ???

doubledragoncc

Pressure sensors move only about 1mm try controlling that. I know its pressure but it is too sensitive. And yes I have wasted money and tried lol.

I think rider movement needs to be via gyro and accelerometer electronics. remember that test I did when there grimm?

You all seem to forget one thing about counter steering, it is a small amount at the start of turning not a large movement and THAT is what is wrong with all these "I turn my wheel left to go right like in real life" bollocks everyone says. You dont turn the frickin bars to the left to turn right, you apply a bit of pressure to the right of the bars which moves it slightly to the left but it is also to stop the front wheel from turning into the turn ie right. How the fuck can holding your bars at 40 degrees to the left feel REAL when going right if on a REAL bike you would hardly even notice it!!!

In a sim you CANT have all the reality due to lack of technology, or putting it correctly, lack of lots of money.

DD
GPBOC Live Streams: https://www.youtube.com/c/IASystemsComputerControls; i7 12700K 5.1GHz Z690 ASUS Strix Z690-A Mobo 32GB 3600MHz DDR4 RAM ASUS Strix RTX3080 OC 10GB DDR6X ASUS Ryujin 360 AOI Cooler ROG Thor 1200w PSU in ROG Helios Tower Case.

TFC

Surely it comes under the realms of force feedback rather than applied pressure. You should have to apply pressure to counter the force feedback right?

August1

Quote from: doubledragoncc on October 14, 2016, 08:54:54 AM
Pressure sensors move only about 1mm try controlling that. I know its pressure but it is too sensitive. And yes I have wasted money and tried lol.


How much torque on the handlebars did it require to depress that pressure sensor the full 1 mm of travel? I'm guessing very little so it would have been very difficult to modulate the input. A torque sensor is completely static though. Let's say you use a torque sensor that's able to measure 0-20 NM of torque to an accuracy of 1 NM. I think your arms would be able to modulate a progressive torque input within that range.

doubledragoncc

To be totally honest you are all over complicating a simple thing because you want it real!!!!

Well heres the actual truth..............

Unless you got a lot of money and a great workshop or willing to pay someone to produce the parts it will NEVER be something everyone can afford. My systems are not cheap but they are a heck of a lot of work to make. Now add something like you are on about and oh boy you are looking at 3 times the cost plus!!!

Now the most important point. Because you are sitting still in a room, your body is working different from when on a bike in real life. You inputs are far greater than normal as you are not fighting actual forces. You have to have a force to fight in order to measure your movements. It is so hard to put into words for me, but I have tried every way to replicate this and it is just too expensive and to be honest I dont think even Piboso can write the program to do it correctly while only using a gamepad him/her self!!! In order to make a sim work the software must be tested with hardware to make it work, BUT there is no hardware so how the fuck can you say it works??? BTW I am not being funny I just use the word fuck as an expression to strengthen the point..

So.................First a system with force feed back and the actual movement of a real bike must be made to allow the programmer the chance to test the program with!!! Then you find out that the hardware concept is great for a real bike but shit while sitting still in a room!!! It is far more complex than just thinking about real world bike movement as your still sitting still in a fucking room!!! LOL

My system is totally unrealistic in its action of a real bike, BUT it is right for replicating you body movement like in the real world, so it is totally wrong but also right, its looking outside of the physics books. This is why all the thoughts about this sensor and that will not work as you think because you all forget the software we have now is written from physics not "Sitting Still In A Room" necessity!!!

Just my 10 cents on the subject

DD
GPBOC Live Streams: https://www.youtube.com/c/IASystemsComputerControls; i7 12700K 5.1GHz Z690 ASUS Strix Z690-A Mobo 32GB 3600MHz DDR4 RAM ASUS Strix RTX3080 OC 10GB DDR6X ASUS Ryujin 360 AOI Cooler ROG Thor 1200w PSU in ROG Helios Tower Case.

August1

October 14, 2016, 12:31:02 PM #13 Last Edit: October 14, 2016, 12:37:00 PM by August1
Mate, no need to get riled up  ;). I only proposed this control method because I feel like it may be a feasible alternative method of providing high-resolution control inputs besides the steering wheel or joystick solutions because ultimately isn't that why we use those solutions and not just stick to gamepads? Because it's easier to modulate control inputs and get rid of the dead zone issue?

doubledragoncc

LOL August not riled up bro that is why I wrote  "BUT there is no hardware so how the fuck can you say it works??? BTW I am not being funny I just use the word fuck as an expression to strengthen the point" I am an passionate person and talk so lol

DD
GPBOC Live Streams: https://www.youtube.com/c/IASystemsComputerControls; i7 12700K 5.1GHz Z690 ASUS Strix Z690-A Mobo 32GB 3600MHz DDR4 RAM ASUS Strix RTX3080 OC 10GB DDR6X ASUS Ryujin 360 AOI Cooler ROG Thor 1200w PSU in ROG Helios Tower Case.