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

Uploading program to CM5

Bioloid robot kit from Korean company Robotis; CM5 controller block, AX12 servos..
6 postsPage 1 of 1
6 postsPage 1 of 1

Uploading program to CM5

Post by clusher » Thu Jul 17, 2008 12:33 pm

Post by clusher
Thu Jul 17, 2008 12:33 pm

Hi all.

I'm programming the CM5 using C and WinAVR.
To upload a program to CM5 when it has the original Robotis firmware I have to go to Robot Terminal and press shift+3 for a while, then type load and transmit file.

But how can I change to a new program after uploading my program? The same way doesn't work, and I can't simply transmit the file... What am I missing? So far, I've been uploading back to Robotis firmware and transmit my new program... but it has to be simpler than that! Do I need to program a loader or something?
Hi all.

I'm programming the CM5 using C and WinAVR.
To upload a program to CM5 when it has the original Robotis firmware I have to go to Robot Terminal and press shift+3 for a while, then type load and transmit file.

But how can I change to a new program after uploading my program? The same way doesn't work, and I can't simply transmit the file... What am I missing? So far, I've been uploading back to Robotis firmware and transmit my new program... but it has to be simpler than that! Do I need to program a loader or something?
clusher
Savvy Roboteer
Savvy Roboteer
Posts: 57
Joined: Thu Jul 17, 2008 12:27 pm

Post by limor » Thu Jul 17, 2008 4:10 pm

Post by limor
Thu Jul 17, 2008 4:10 pm

it may be because you press 3 times only.
whenever you press the reset red button on the CM5 while you press continuously the '#' key in robot terminal (it will display many ###########), you will see the bootloader prompt regardless of what firmware is in the flash memory.
it may be because you press 3 times only.
whenever you press the reset red button on the CM5 while you press continuously the '#' key in robot terminal (it will display many ###########), you will see the bootloader prompt regardless of what firmware is in the flash memory.
limor
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1845
Joined: Mon Oct 11, 2004 1:00 am
Location: London, UK

Post by clusher » Fri Aug 01, 2008 3:51 pm

Post by clusher
Fri Aug 01, 2008 3:51 pm

Just noticed i forgot to thank you for the answer. You were right, of course.

Now a radical different question. I'm trying to copy cat a small experience I saw on youtube: http://www.youtube.com/watch?v=HpFGROm-LQM

I've decided to use a position control based on the torque applied on the ankles. As the slope increases, so does the torque and I try to adjust the legs the way the video shows.

The first thing is the torque is really noisy. Just by being on a standup position I get torques of 0-120 on either direction, while I expected to see a zero torque. The second is that as I increase the slope the torque doesnt change until it is too late (robot falling over or sliding...)

I started out by making a velocity control loop, but the problems with the servos were the same...

Any ideas how to improve the feedback? I tried lowering the center of mass (bending the knees), but didn't too much...
Just noticed i forgot to thank you for the answer. You were right, of course.

Now a radical different question. I'm trying to copy cat a small experience I saw on youtube: http://www.youtube.com/watch?v=HpFGROm-LQM

I've decided to use a position control based on the torque applied on the ankles. As the slope increases, so does the torque and I try to adjust the legs the way the video shows.

The first thing is the torque is really noisy. Just by being on a standup position I get torques of 0-120 on either direction, while I expected to see a zero torque. The second is that as I increase the slope the torque doesnt change until it is too late (robot falling over or sliding...)

I started out by making a velocity control loop, but the problems with the servos were the same...

Any ideas how to improve the feedback? I tried lowering the center of mass (bending the knees), but didn't too much...
clusher
Savvy Roboteer
Savvy Roboteer
Posts: 57
Joined: Thu Jul 17, 2008 12:27 pm

Post by StuartL » Fri Aug 08, 2008 5:51 pm

Post by StuartL
Fri Aug 08, 2008 5:51 pm

The torque feedback algorithm in the AX-12 standard firmware is shocking. We're working on an improved algorithm but in the short term there's an immediate work-around:

Change the stance so that some of the servos are under high load, e.g. the hips, thighs or knees. When you lose the load from these you know that the leg is unloaded.

Or, alternatively, fit an IMU :)
The torque feedback algorithm in the AX-12 standard firmware is shocking. We're working on an improved algorithm but in the short term there's an immediate work-around:

Change the stance so that some of the servos are under high load, e.g. the hips, thighs or knees. When you lose the load from these you know that the leg is unloaded.

Or, alternatively, fit an IMU :)
StuartL
Savvy Roboteer
Savvy Roboteer
Posts: 350
Joined: Mon Jun 04, 2007 3:46 pm
Location: Thatcham, Berkshire, UK

Post by clusher » Tue Sep 02, 2008 3:25 pm

Post by clusher
Tue Sep 02, 2008 3:25 pm

Back from holidays ;)

Yeah, I'm waiting for a few extras for bioloid, including the IMU.

The work around seems good, but it gets too slow to read 3 or 4 torque values on a closed loop... Did you guys finish remaking the AX-12 firmware? I read somewhere it would have more reliable readings and faster response...
Back from holidays ;)

Yeah, I'm waiting for a few extras for bioloid, including the IMU.

The work around seems good, but it gets too slow to read 3 or 4 torque values on a closed loop... Did you guys finish remaking the AX-12 firmware? I read somewhere it would have more reliable readings and faster response...
clusher
Savvy Roboteer
Savvy Roboteer
Posts: 57
Joined: Thu Jul 17, 2008 12:27 pm

Post by StuartL » Tue Sep 02, 2008 9:52 pm

Post by StuartL
Tue Sep 02, 2008 9:52 pm

"Finished" is the wrong word. We haven't yet done the torque feedback or PID control, the SYNC_READ isn't written yet either. However the other stuff (including on the fly reflashing) is all done.

I'm going to see Matt on Thursday, I'll see if we can get more of the servo firmware done in the next few weeks :)
"Finished" is the wrong word. We haven't yet done the torque feedback or PID control, the SYNC_READ isn't written yet either. However the other stuff (including on the fly reflashing) is all done.

I'm going to see Matt on Thursday, I'll see if we can get more of the servo firmware done in the next few weeks :)
StuartL
Savvy Roboteer
Savvy Roboteer
Posts: 350
Joined: Mon Jun 04, 2007 3:46 pm
Location: Thatcham, Berkshire, UK


6 postsPage 1 of 1
6 postsPage 1 of 1