Submitted by jake.carvey on Fri, 05/28/2010 - 10:34
Posted in
Any thoughts on a way to incorporate real world units?
In other words, could we use some sort of sensor (like a measuring wheel with known diamter) to measure real world distance traveled (or rotated) and feed it back into the system? Ideally, this would be independent of the drive mechanism so that components could be switched out during experimentation.
lasers? maybe an additional omega drive with an optical encoder? not my area of expertise, just throwing ideas out
jake
Lasers! Awesome - I want some
Lasers! Awesome - I want some *grin* But, seriously, an encoder can be used either at the shaft of the motor or at a final output stage. Putting it at the final output stage allows for measurement of true rotation of the drive component. This does result in lower resolution of measurement if you're gearing down the drive train (3:1 gear ratio means you'd lose 1/3rd of your total resolution) - but it allows for measurement of actual travel distance.
I'm not sure the benefits are really there for measuring at the output stage and the amount of resolution your are likely to lose. Why not measure at the motor and add a configurable parameter for gear ratio? Then, you get the best of both worlds - high resolution and real distance measuring.
!c
@shutterdrone: fair enough -
@shutterdrone:
fair enough - as long as there is a quick and simple calibration process that can be done in the field? I seem to remember this was always part of the goal?
I need to get my prototypes up and running - close
Jake - what I'm expecting for
Jake - what I'm expecting for the "simple gui" is you just put in:
For rotational axes:
Gear ratio
Steps/rev (if stepper)
RPM (if DC)
For linear axes:
Gear ratio
Steps/rev (if stepper)
RPM (if DC)
Diameter of final gear
Any one could be independently changed, so if you just change the dia. of the final gear, just change that.
!c
Is it too much to ask of an
Is it too much to ask of an arduino to calculate say an inch from so many steps. Have a calibrate function so you can move the axis say 1 inch/degrees and then have the arduino calculate a move of 16 inches internally. Is that clear? I am going to make this an OpenMoco weekend so I will understand the software better next week.
Cheers
RC, given the right software,
RC, given the right software, calibrating off of an input is no big deal. One just has to handle the feedback properly =)
!c