Legacy Forum: Preserving Nearly 20 Years of Community History - A Time Capsule of Discussions, Memories, and Shared Experiences.

Premium stability test

Bioloid robot kit from Korean company Robotis; CM5 controller block, AX12 servos..
33 postsPage 2 of 31, 2, 3
33 postsPage 2 of 31, 2, 3

f

Post by UncleBob » Tue Mar 16, 2010 3:19 am

Post by UncleBob
Tue Mar 16, 2010 3:19 am

I think your foot sensor feedback is doing as well as my balancer.

BTW what difference do you find turning on / off the gyro during your balance test? Have your tried turning it off during walking?
I think your foot sensor feedback is doing as well as my balancer.

BTW what difference do you find turning on / off the gyro during your balance test? Have your tried turning it off during walking?
UncleBob
Savvy Roboteer
Savvy Roboteer
Posts: 398
Joined: Sun Dec 27, 2009 5:25 am

Post by FabinovX » Tue Mar 16, 2010 10:18 am

Post by FabinovX
Tue Mar 16, 2010 10:18 am

the difference is that when you push strongly the bioloid with gyro on, it doesn't balance to find its equilibrium point as will do any solid object.
as soon as the whole surface of the feet touch the ground, the robot stabilize instantly without balancing anymore.
when shaking the robot, the foot keeps the contact with the ground.
when you turn the gyro off, the robot acts as a brick and fall more easily.
when walking, the gyro is really helpfull to stabilize the robot.

but the real advantage of the AX-S20 is that it gives more informations really interesting.
to make a program as mine it's really simple with this balancer. just use the pitch and roll values instead of calculating lots of things using noisy values.
the hardware do the job for you, you just have to get the data.
it saves a lot of variables (which quantity is quite limited in cm510).
To correct path when walking, you can use the Z azymut and so on.
that's really an interesting piece of technology and i long to see it on sale.
the difference is that when you push strongly the bioloid with gyro on, it doesn't balance to find its equilibrium point as will do any solid object.
as soon as the whole surface of the feet touch the ground, the robot stabilize instantly without balancing anymore.
when shaking the robot, the foot keeps the contact with the ground.
when you turn the gyro off, the robot acts as a brick and fall more easily.
when walking, the gyro is really helpfull to stabilize the robot.

but the real advantage of the AX-S20 is that it gives more informations really interesting.
to make a program as mine it's really simple with this balancer. just use the pitch and roll values instead of calculating lots of things using noisy values.
the hardware do the job for you, you just have to get the data.
it saves a lot of variables (which quantity is quite limited in cm510).
To correct path when walking, you can use the Z azymut and so on.
that's really an interesting piece of technology and i long to see it on sale.
FabinovX
Savvy Roboteer
Savvy Roboteer
Posts: 74
Joined: Mon Feb 01, 2010 4:12 pm

Z

Post by UncleBob » Tue Mar 16, 2010 11:20 am

Post by UncleBob
Tue Mar 16, 2010 11:20 am

Ya I have an idea about the Z as well.

Since now it knows the direction, you can program it to go somewhere specific in your house.
Ya I have an idea about the Z as well.

Since now it knows the direction, you can program it to go somewhere specific in your house.
UncleBob
Savvy Roboteer
Savvy Roboteer
Posts: 398
Joined: Sun Dec 27, 2009 5:25 am

y

Post by UncleBob » Tue Mar 16, 2010 2:13 pm

Post by UncleBob
Tue Mar 16, 2010 2:13 pm

I finally get something working. The problem is due to the motion file in the robot is corrupted. I reupload it and it seems better. But the tsk file get the side mixed up. As soon as robot starts, it shift to one side. I try change 17 & 18 goal - position, it is better but not perfect. I think still need to change the sign of the formulae somewhere.
I finally get something working. The problem is due to the motion file in the robot is corrupted. I reupload it and it seems better. But the tsk file get the side mixed up. As soon as robot starts, it shift to one side. I try change 17 & 18 goal - position, it is better but not perfect. I think still need to change the sign of the formulae somewhere.
UncleBob
Savvy Roboteer
Savvy Roboteer
Posts: 398
Joined: Sun Dec 27, 2009 5:25 am

Post by FabinovX » Tue Mar 16, 2010 3:39 pm

Post by FabinovX
Tue Mar 16, 2010 3:39 pm

did you try the code i did especially for you yesterday.
If not here it is :

FabinovX wrote:Ok UncleBoB, try this with your type C, it should work fine :

http://robosavvy.com/Builders/FabinovX/ ... Type_C.rar

please, tell me if it really does or not.

File updated (i forgot to invert gyro offset with 9 and 10)


if you made modification to adapt the code with axs20, just paste the two line about ID 9 and 10 join offset in the callback and the whole function mesure_delta of this file to yours.

cheers
FabinovX
did you try the code i did especially for you yesterday.
If not here it is :

FabinovX wrote:Ok UncleBoB, try this with your type C, it should work fine :

http://robosavvy.com/Builders/FabinovX/ ... Type_C.rar

please, tell me if it really does or not.

File updated (i forgot to invert gyro offset with 9 and 10)


if you made modification to adapt the code with axs20, just paste the two line about ID 9 and 10 join offset in the callback and the whole function mesure_delta of this file to yours.

cheers
FabinovX
FabinovX
Savvy Roboteer
Savvy Roboteer
Posts: 74
Joined: Mon Feb 01, 2010 4:12 pm

Post by FabinovX » Tue Mar 16, 2010 8:34 pm

Post by FabinovX
Tue Mar 16, 2010 8:34 pm

Here is my new video :

phpBB [media]


and here is the code :

http://robosavvy.com/Builders/FabinovX/ ... o+arms.rar

Dont forget it will only work fine without modifications with a standard type A premium.

have fun :)

cheers
FabinovX
Here is my new video :

phpBB [media]


and here is the code :

http://robosavvy.com/Builders/FabinovX/ ... o+arms.rar

Dont forget it will only work fine without modifications with a standard type A premium.

have fun :)

cheers
FabinovX
FabinovX
Savvy Roboteer
Savvy Roboteer
Posts: 74
Joined: Mon Feb 01, 2010 4:12 pm

gyro

Post by UncleBob » Sat Mar 20, 2010 3:14 am

Post by UncleBob
Sat Mar 20, 2010 3:14 am

I just compare my robot (without a gyro) walking with one of my video on youtube. I cannot see any big difference.
I just compare my robot (without a gyro) walking with one of my video on youtube. I cannot see any big difference.
UncleBob
Savvy Roboteer
Savvy Roboteer
Posts: 398
Joined: Sun Dec 27, 2009 5:25 am

Post by FabinovX » Sat Mar 20, 2010 11:28 am

Post by FabinovX
Sat Mar 20, 2010 11:28 am

Hi,
well as far as i understand, the gyro is there to help stabilization.
so gyro response are calibrated not to disrupt motion page data send to the actuators if the robot does'nt need stabilization.
that is to say if the robot is walking on a flat surface for example.
but if it walks on a litlle irregular surface and the chest is starting to balance to much, then it will help to stabilize.

cheers
FabinovX.
Hi,
well as far as i understand, the gyro is there to help stabilization.
so gyro response are calibrated not to disrupt motion page data send to the actuators if the robot does'nt need stabilization.
that is to say if the robot is walking on a flat surface for example.
but if it walks on a litlle irregular surface and the chest is starting to balance to much, then it will help to stabilize.

cheers
FabinovX.
FabinovX
Savvy Roboteer
Savvy Roboteer
Posts: 74
Joined: Mon Feb 01, 2010 4:12 pm

new tricks

Post by UncleBob » Thu Apr 01, 2010 10:23 am

Post by UncleBob
Thu Apr 01, 2010 10:23 am

I got a few new tricks in mind for my bioloid. With easter here I hope to do some stuff. Also I just got a SSD drive for my macbook. It is superfast now for me to make vid as well.
I got a few new tricks in mind for my bioloid. With easter here I hope to do some stuff. Also I just got a SSD drive for my macbook. It is superfast now for me to make vid as well.
UncleBob
Savvy Roboteer
Savvy Roboteer
Posts: 398
Joined: Sun Dec 27, 2009 5:25 am

Post by pikaj00 » Sun Apr 18, 2010 4:54 am

Post by pikaj00
Sun Apr 18, 2010 4:54 am

tried your code works great thanks alot! mine is a little jumpy when standing still but not to bad.. love it thanks again
tried your code works great thanks alot! mine is a little jumpy when standing still but not to bad.. love it thanks again
pikaj00
Robot Builder
Robot Builder
Posts: 23
Joined: Tue Mar 09, 2010 4:31 am

Post by pikaj00 » Mon Apr 19, 2010 3:27 am

Post by pikaj00
Mon Apr 19, 2010 3:27 am

just a question with this code and motion installed can you not use the remote controller to walk anymore? mine stopped working aswell as the demo mode no more works.. thanks talk to you later.
just a question with this code and motion installed can you not use the remote controller to walk anymore? mine stopped working aswell as the demo mode no more works.. thanks talk to you later.
pikaj00
Robot Builder
Robot Builder
Posts: 23
Joined: Tue Mar 09, 2010 4:31 am

Post by FabinovX » Thu Apr 22, 2010 9:59 pm

Post by FabinovX
Thu Apr 22, 2010 9:59 pm

Hi,
the code i gave is a stand alone code.
but if you want, you can copy and paste it to the original code given with with the premium.
to make your robot less jumpy, you can try to modify the dead zone values in the code.
but the bigger are the ranges between negative and positive value, the less reactive is the robot.

see you ;)

Cheers
FabinovX
Hi,
the code i gave is a stand alone code.
but if you want, you can copy and paste it to the original code given with with the premium.
to make your robot less jumpy, you can try to modify the dead zone values in the code.
but the bigger are the ranges between negative and positive value, the less reactive is the robot.

see you ;)

Cheers
FabinovX
FabinovX
Savvy Roboteer
Savvy Roboteer
Posts: 74
Joined: Mon Feb 01, 2010 4:12 pm

Post by mattman » Sun Apr 25, 2010 8:42 pm

Post by mattman
Sun Apr 25, 2010 8:42 pm

thanks for the code! pretty good stuff
thanks for the code! pretty good stuff
mattman
Newbie
Newbie
Posts: 3
Joined: Sat Apr 24, 2010 7:17 am

Is not working

Post by cajacava » Wed May 05, 2010 3:58 am

Post by cajacava
Wed May 05, 2010 3:58 am

Hi, I'm really new in all this. I download the motion and task files from your post in my premium kit but when I run the program the robot send the two legs to the same side or to the front at the same time and I can't controle it. Can you help me please? I'm using the gyro of the kit
Hi, I'm really new in all this. I download the motion and task files from your post in my premium kit but when I run the program the robot send the two legs to the same side or to the front at the same time and I can't controle it. Can you help me please? I'm using the gyro of the kit
cajacava
Newbie
Newbie
Posts: 2
Joined: Wed May 05, 2010 3:52 am

Post by FabinovX » Fri May 07, 2010 8:32 pm

Post by FabinovX
Fri May 07, 2010 8:32 pm

Maybe, that's due to the noise of AX12 position feedback.
you can try to enlarge dead zone in the soft.
but first you may try this :
Just put the robot on the ground before starting the program.
if it still goes crazy, then you should calibrate your robot.
In effect, if your ax12 can't reach their goal position when standing there will be a difference between goal and present position, so that, the robot will analyse this and move for nothing.

EDIT : the gyro is just use to stabilize when there is quick movements,
it doesn't help when you do slow movements.
it's only AX12 position feedback that helps in this case.
for information, my very first try was based on AX12 torque feedback but it was more noisy than position feedback and the soft was really more complicated and longer because torque value changes according to the rotation direction too.

Hope it will help.

Cheers
FabinovX
Maybe, that's due to the noise of AX12 position feedback.
you can try to enlarge dead zone in the soft.
but first you may try this :
Just put the robot on the ground before starting the program.
if it still goes crazy, then you should calibrate your robot.
In effect, if your ax12 can't reach their goal position when standing there will be a difference between goal and present position, so that, the robot will analyse this and move for nothing.

EDIT : the gyro is just use to stabilize when there is quick movements,
it doesn't help when you do slow movements.
it's only AX12 position feedback that helps in this case.
for information, my very first try was based on AX12 torque feedback but it was more noisy than position feedback and the soft was really more complicated and longer because torque value changes according to the rotation direction too.

Hope it will help.

Cheers
FabinovX
FabinovX
Savvy Roboteer
Savvy Roboteer
Posts: 74
Joined: Mon Feb 01, 2010 4:12 pm

PreviousNext
33 postsPage 2 of 31, 2, 3
33 postsPage 2 of 31, 2, 3