Closed PeterBowman closed 4 years ago
Related:
* Not prioritary, just to point out. For starters, SerialServoBoard looks PositionControl-oriented, whereas we both agreed on exposing the more native PWM interface.
Returning to the scope of the issue, yes, I totally agree we should recover the full range at firmware level.
Serial/USB
setup we once had (mentioned at https://github.com/roboticslab-uc3m/yarp-devices/issues/234#issuecomment-533788381).Canbus
too.Updated Lacquey firmware: https://os.mbed.com/users/raulsr988/code/LacqueyFetch/rev/f31b6ddb9cd3/
Note: Currently the previus firmware is installed on the Mbed.
Waiting to install the latest changes (38176f164de5bd0873013406efda53323c1a415c) in Teo to be tested completely. Right now the tests have been done under Pcanview
, sending messages directly to the Mbed. All works correctly. I'll close this issue
Just one more thing to finally mark this as closed: merge this change into firmware/LacqueyFetch/ (also update README).
Commit https://github.com/roboticslab-uc3m/yarp-devices/commit/cfd8d88e8555b8fd399e4ab80db95df9082e138f#diff-68eeea6c71ceb478044ee0f4dc44b6a0 (https://github.com/roboticslab-uc3m/yarp-devices/issues/211) adds a new PWM control layer around the LacqueyFetch device, resembling more closely its internal way of actuating the motor. However, our firmware is severely limited in that only full-positive (+100%), full-negative (-100%) and stop (0%) commands are forwarded, thus ignoring the whole
[-1,1]
value range (source):https://github.com/roboticslab-uc3m/yarp-devices/blob/8726fa368f8b3becf26a2f263f4d451e1836eb1b/firmware/LacqueyFetch/main.cpp#L76-L92
Here,
data
is achar
. Make it a 32-bitfloat
so that it matches https://github.com/roboticslab-uc3m/yarp-devices/commit/38176f164de5bd0873013406efda53323c1a415c.