by garycmartin » Tue Dec 28, 2010 1:00 am
by garycmartin
Tue Dec 28, 2010 1:00 am
Many thanks, I just built a Humanoid A and don't yet have a windows VM on my Mac to use any of the Robotis provided software for debugging/testing. After about 5-6hrs building and checking, I switched it on only to hear the repeating Do Re Me "some of the cables are not properly in place" tones. Arrrgh! After more visual checking that all was wired correctly in order, I installed the driver 2.2.14 from FTDI; plugged in the Dynamixel2USB; used one of the spare Dynamixel cables to connect into the 510 controller; powered it up and set it to MANAGE mode; ran your Dynamixel Monitor application and edited the preferences to give it the right /dev/tty.usbserialXXXXX name. It found a different set of IDs each time I disconnected/connected, and any viewed data values for a servo jumped all over the place – ouch... I then started testing smaller chains of servos one by one and narrowed it down to the left leg, testing each servo in that leg I noticed one cable where a wire had partially pushed back out of its plug, probably when it was inserted into the servo. After re-seateing that one wire using long nosed pliers, all is now well, and my Humanoid A is happily walking about
One minor bug I did notice with Dynamixel Monitor application that initially had me confused: If you use the disconnect button after testing some set of servos and then change the servos plugged in (say from one arm chain to another), when you use the connect button to re-connect, it shows the correct list of IDs but if you click any, the ID in the main detail window will be stuck on the first servo ID you click on, and you can only modify that one servo. The trick to avoid this glitch is to quit the Dynamixel Monitor application each time you change the servos physically connected to the bus and then run it again – each servo then can be manipulated and examined individually.
Many thanks again!
Many thanks, I just built a Humanoid A and don't yet have a windows VM on my Mac to use any of the Robotis provided software for debugging/testing. After about 5-6hrs building and checking, I switched it on only to hear the repeating Do Re Me "some of the cables are not properly in place" tones. Arrrgh! After more visual checking that all was wired correctly in order, I installed the driver 2.2.14 from FTDI; plugged in the Dynamixel2USB; used one of the spare Dynamixel cables to connect into the 510 controller; powered it up and set it to MANAGE mode; ran your Dynamixel Monitor application and edited the preferences to give it the right /dev/tty.usbserialXXXXX name. It found a different set of IDs each time I disconnected/connected, and any viewed data values for a servo jumped all over the place – ouch... I then started testing smaller chains of servos one by one and narrowed it down to the left leg, testing each servo in that leg I noticed one cable where a wire had partially pushed back out of its plug, probably when it was inserted into the servo. After re-seateing that one wire using long nosed pliers, all is now well, and my Humanoid A is happily walking about
One minor bug I did notice with Dynamixel Monitor application that initially had me confused: If you use the disconnect button after testing some set of servos and then change the servos plugged in (say from one arm chain to another), when you use the connect button to re-connect, it shows the correct list of IDs but if you click any, the ID in the main detail window will be stuck on the first servo ID you click on, and you can only modify that one servo. The trick to avoid this glitch is to quit the Dynamixel Monitor application each time you change the servos physically connected to the bus and then run it again – each servo then can be manipulated and examined individually.
Many thanks again!