New Bioloid Humanoid
New Bioloid Humanoid
by Pev
Mon Aug 07, 2006 2:52 pm
Well here he is the 18 DOF Bioloid Humanoid as designed for the new Comprehensive Kit.
So he has lost a single degree of movement, which is his waist servo. No big deal really as it was a bit redundant IMHO. There are also some changes in the construction of the frames for the legs but these are minor. Robotis are also supplying a much more comprehensive demo program which includes walking routines. Have only just started to play with them but he seems pretty good and his walk is quite quick compared to the RN-1. He also turns well as he has the rotating hip servos. All in all agreat improvement over the old 19 DOF primarily due to the new routines and motions included by Robotis.
Pev
So he has lost a single degree of movement, which is his waist servo. No big deal really as it was a bit redundant IMHO. There are also some changes in the construction of the frames for the legs but these are minor. Robotis are also supplying a much more comprehensive demo program which includes walking routines. Have only just started to play with them but he seems pretty good and his walk is quite quick compared to the RN-1. He also turns well as he has the rotating hip servos. All in all agreat improvement over the old 19 DOF primarily due to the new routines and motions included by Robotis.
Pev
by pepperm
Mon Aug 07, 2006 5:31 pm
Hi
I see from your picture that the servos seem to be the original AX-12 servos and not the AX-12+ servos that Limor mentioned. Is this the case or is it that I can't quite see the markings.
In the event that they are the AX-12+ servos, what are they like compared to the AX-12, if you are able to make a comparison.
Well done in makeing the humanoid so quickly too Pev. How many hours did it take?
Looks like they have all sold from Robosavvy too because the Buy link "Product Not found", maybe a glitch though.
My Bioloid should be with me in a few days. I can't wait
Mark
I see from your picture that the servos seem to be the original AX-12 servos and not the AX-12+ servos that Limor mentioned. Is this the case or is it that I can't quite see the markings.
In the event that they are the AX-12+ servos, what are they like compared to the AX-12, if you are able to make a comparison.
Well done in makeing the humanoid so quickly too Pev. How many hours did it take?
Looks like they have all sold from Robosavvy too because the Buy link "Product Not found", maybe a glitch though.
My Bioloid should be with me in a few days. I can't wait
Mark
by Pev
Mon Aug 07, 2006 7:56 pm
Hi,
You're quite right they are just AX-12s not the + version. I had one of the first comprehensive kits shipped to europe so have 19 AX-12s and 1 AX-S1. The kits are very good and I've built a number of different 'bots.
I did make the mistake of building a 19 DOF humanoid from the original instructions last night then realised I need to build the 18 DOF for the new software to work. It took about 3 to 4 hours to build and test this morning and then a bit of tinkering with the offset setting in the motion editor to get it to walk without falling backwards.
The offset is important but the PDF QuickStart Manual covers it well and it seems to make all the difference with the supplied walking routine. I'll try to get a video done and posted.
The demo program is good but I think I'll tinker with that too to get the IR sensor a bit more sensitive as mine walks into walls sometimes lol.
All in all a lot better than the 19 DOF humanoid as I said, but mainly due to the demo program and the new walking routines. Its a quick walk too and the turning is cool with the rotational hips.
Am hoping the expansion pack will be available soon as I'd like to get the Bluetooth module and expand the autonomy by linking it to a laptop for more detailed processing and the camera looks cool too.
Am trying the crack the IR remote capabilities at the moment so if you know of any good IR analyzers for a PC or PPC I'd be really interested.
Be interested in how you get on with your kit when it arrives in the next few days, am sure you'll have a blast.
Pev
You're quite right they are just AX-12s not the + version. I had one of the first comprehensive kits shipped to europe so have 19 AX-12s and 1 AX-S1. The kits are very good and I've built a number of different 'bots.
I did make the mistake of building a 19 DOF humanoid from the original instructions last night then realised I need to build the 18 DOF for the new software to work. It took about 3 to 4 hours to build and test this morning and then a bit of tinkering with the offset setting in the motion editor to get it to walk without falling backwards.
The offset is important but the PDF QuickStart Manual covers it well and it seems to make all the difference with the supplied walking routine. I'll try to get a video done and posted.
The demo program is good but I think I'll tinker with that too to get the IR sensor a bit more sensitive as mine walks into walls sometimes lol.
All in all a lot better than the 19 DOF humanoid as I said, but mainly due to the demo program and the new walking routines. Its a quick walk too and the turning is cool with the rotational hips.
Am hoping the expansion pack will be available soon as I'd like to get the Bluetooth module and expand the autonomy by linking it to a laptop for more detailed processing and the camera looks cool too.
Am trying the crack the IR remote capabilities at the moment so if you know of any good IR analyzers for a PC or PPC I'd be really interested.
Be interested in how you get on with your kit when it arrives in the next few days, am sure you'll have a blast.
Pev
by pepperm
Mon Aug 07, 2006 10:02 pm
Thanks for that, interesting that you mention IR though. I was reading through the pdfs last night and noticed the IR input option. It looks interesting. There is loads of info on the web about IR protocols but I need to get my Bioloid to play with to investigate further.
I also plan to use some wireless modules I got from http://www.active-robots.com/products/r ... adio.shtml for the serial comms. It just removes the need for a cable.
The bot accepts ZigBee wireless boards but to be honest I am not sure how to use them with a PC. The manual seems to talk about bot to bot comms only.
I hope to be posting a lot when I get my Bioloid working though. I am particularly interested in hacking into the Mega128 I/O connections that must be on the CM5 board, perhaps producing an I/O connector that can be used with standard gyros and servos for additions that could be used from "C" code.
Mark
I also plan to use some wireless modules I got from http://www.active-robots.com/products/r ... adio.shtml for the serial comms. It just removes the need for a cable.
The bot accepts ZigBee wireless boards but to be honest I am not sure how to use them with a PC. The manual seems to talk about bot to bot comms only.
I hope to be posting a lot when I get my Bioloid working though. I am particularly interested in hacking into the Mega128 I/O connections that must be on the CM5 board, perhaps producing an I/O connector that can be used with standard gyros and servos for additions that could be used from "C" code.
Mark
by inaki
Tue Aug 08, 2006 12:21 pm
With regards to IR, I have had problems reading communications commands because IR leds keep all the time ON. Probably because they work as obstacle sensors too.
So in order to analyze the communication signal you must separate the IR signal that is coming from the distance sensor from the one that comes from the IR communications. When sending bytes through communications port it sends additional signals. I used a Sharp GP sensor to detect IR signals along a video camera to actually see the signals. However I was unable to read anything useful on the sensor, I think because of the overlapped IR signals that are coming out of the AXS1 or perhaps because the signals are not 40Khz. Most appliances use 40Khz IR carriers but I don't know if this is the case. The bad about the Sharp sensor is that it has a filter around 40Khz and is blind to other frequencies.
I checked also if a universal IR learning remote could capture the signals but it could not. I guess the problem was the overlapped signals again or perhaps the carrier.
Perhaps the first step would be to check actual frequency of IR signals on the AXS1 side using a scope. Then capture the commands for that frequency. I am not positive about signals being overlapped or in sequence. To me it would be more logical to be in sequence. I mean, the distance sensor should be OFF when communications data is ON.
As a side note, the signals used for communication seem to be brighter that the ones used for obstacle detection which would mean that the frequency is much higher.
So in order to analyze the communication signal you must separate the IR signal that is coming from the distance sensor from the one that comes from the IR communications. When sending bytes through communications port it sends additional signals. I used a Sharp GP sensor to detect IR signals along a video camera to actually see the signals. However I was unable to read anything useful on the sensor, I think because of the overlapped IR signals that are coming out of the AXS1 or perhaps because the signals are not 40Khz. Most appliances use 40Khz IR carriers but I don't know if this is the case. The bad about the Sharp sensor is that it has a filter around 40Khz and is blind to other frequencies.
I checked also if a universal IR learning remote could capture the signals but it could not. I guess the problem was the overlapped signals again or perhaps the carrier.
Perhaps the first step would be to check actual frequency of IR signals on the AXS1 side using a scope. Then capture the commands for that frequency. I am not positive about signals being overlapped or in sequence. To me it would be more logical to be in sequence. I mean, the distance sensor should be OFF when communications data is ON.
As a side note, the signals used for communication seem to be brighter that the ones used for obstacle detection which would mean that the frequency is much higher.
by Pev
Tue Aug 08, 2006 2:20 pm
inaki wrote:I checked also if a universal IR learning remote could capture the signals but it could not. I guess the problem was the overlapped signals again or perhaps the carrier.
Hi Inaki,
I tried the Universal learning remote too with no luck. THink I may try an IR receiver hooked up to a BasicX processor next
Pev
by inaki
Tue Aug 08, 2006 9:16 pm
I have checked it again. The carrier is 40Khz. Each command is 16 bits long, plus a header and several control bits. I have not figured out yet what these control bits mean but the format looks pretty straightforward.
The distance sensor has a different pattern, it emits a pulse at about 40Hz with a small width of 83us, and it is OFF when sending a command, so it should not interfere with commands. So the light is pretty dim when used only as obstacle detector.
So, conclusion, it should be pretty easy to send commands from other device like a microcontroller or even a PC, or, why not, to read commands from the robot.
The distance sensor has a different pattern, it emits a pulse at about 40Hz with a small width of 83us, and it is OFF when sending a command, so it should not interfere with commands. So the light is pretty dim when used only as obstacle detector.
So, conclusion, it should be pretty easy to send commands from other device like a microcontroller or even a PC, or, why not, to read commands from the robot.
by pepperm
Tue Aug 08, 2006 10:27 pm
All very intersting but surely a good thing to do would be to get the AX-S1 receive and the code in the CM-5 to understand code being sent to the AX-S1 by standard TV style remote controls. That way you could write code that would respond to commands from a remote.
I think I read somewhere that there is an IR remote control for the Bioloid, that and the comments in the manual suggest to me that receiving TV style IR signals is possible. We just need to figure out how.
Remote controls use a standard code sequence defined in a standard called RC-5 I believe.
Still not got my Loid yet though so I can't experiment yet. Any day now....
Mark
I think I read somewhere that there is an IR remote control for the Bioloid, that and the comments in the manual suggest to me that receiving TV style IR signals is possible. We just need to figure out how.
Remote controls use a standard code sequence defined in a standard called RC-5 I believe.
Still not got my Loid yet though so I can't experiment yet. Any day now....
Mark
by Pev
Tue Aug 08, 2006 11:45 pm
pepperm wrote:All very intersting but surely a good thing to do would be to get the AX-S1 receive and the code in the CM-5 to understand code being sent to the AX-S1 by standard TV style remote controls. That way you could write code that would respond to commands from a remote.
I think I read somewhere that there is an IR remote control for the Bioloid, that and the comments in the manual suggest to me that receiving TV style IR signals is possible. We just need to figure out how.
Remote controls use a standard code sequence defined in a standard called RC-5 I believe.
Still not got my Loid yet though so I can't experiment yet. Any day now....
Mark
That is a great idea but I have tried and and failed. Maybe you'll have better luck when you get yours. I have requested protocol info from Robotis.
Pev