Timing Belt Machine Improvement and Pictures

A fellow builder showed me a machine he is building that was inspired by my work and suggestions. He used a high quality linear slide and a linear bearing at the forward end of the mail shaft. I was a little concerned about stability of the new timing belt design and also about excessive wear to the vee rail over time. Using his example I made a simple guide bushing for the front of my machine. I think it will be a significant improvement. Here is a picture of the new bushing. Then a picture of the new slate base added for stability, and finally all the detail pictures for the timing belt machine. Most of the pictures were taken before the front bushing was added.

support bushing

 

 

 

 

 

This entry was posted in Uncategorized. Bookmark the permalink.

4 Responses to Timing Belt Machine Improvement and Pictures

  1. Boris says:

    It’s interesting you’ve moved to a pulley / belt design, which seems more complicated and ordinary. I really liked the elegance and mechanical simplicity of the earlier, apparently unique three wheeled design. Do you see any compelling advantages in the pulley design over the earlier design?

  2. shagmatic says:

    the three wheel design is fine as long as there is a responsible adult monitoring the toy and providing the corrections necessary for proper control 🙂

    the problem with the three wheel design is that it does slip. it would probably be possible to stop it from slipping by knurling the shaft (i had tried this in the past) or coating it with diamond grit. slipping is no problem if someone is making corrections as it slips but running from a fixed program such as following audio there is no opportunity to correct in an open loop system.

    the timing belt machine does not slip unless heavily loaded or with very fast acceleration. i have added code to control the stroke length in response to the audio and this can result in very fast acceleration. i will be adding code ot match the acceleration to the stroke length so this is minimized. of course it will also limit the speed of response.

  3. Baz says:

    At the risk of suggesting the obvious, what about a position sensor to measure shaft location and slip? A series of randomly spaced black/white lines and a suitable mouse-like sensor should be able to detect position and velocity…. Add a control loop to damp the slip… The timing belt just looks so tricky to build.

    • shagmatic says:

      there are many ways to get and utilize position information. the questions are whether it is really necessary and whether a simpler method can achieve adequate accuracy. an open look system is not corrected. errors due to mechanical slipping or missed motor steps are not corrected and if not random will accumulate. the timing belt design is not that difficult to build and eliminates the possibility of the motor and the shaft getting out of sync. still the possibility of the motor missing steps exists. this can be overcome by using a motor of adequate torque and also limiting acceleration and speed to within the capabilities of that motor. i have cnc machines that work just fine with open loop control, however you do need to respect the system’s limits.

      it is also possible to add an encoder to the motor shaft which is much easier than a position sensor on the ram shaft. position information can be used by the control program to correct for motor misses. however this is not as simple as a first glance at the problem will indicate. you need to set limits to position accuracy and decide when and ow to make the necessary corrections. it is much easier to leave this to the controller by using a servo motor. if the motor is adequate all you need to do is set the tolerance for position error and let the controller do the tough part of the job. you can monitor the fault output of the driver and stop the machine if errors are not corrected. my cnc router has servo motors on the X and Y axes because these are high speed and may encounter considerable resistance to motion when making heavy cuts. i am still using a stepper on the Z axis that does not move as fast and is not as heavily loaded.

      my next machine will use a cable drive that will be a little easier and probably cheaper to build but will require a custom drive wheel. i will be testing it with both stepper and servo motors.

      adding a position sensor to the ram shaft of the drive wheel system will not be easy. the position sensor needs to accurately follow the shaft or it will be useless. it can not be friction coupled to the shaft. the environment is too messy to use an optical sensor. a magnetic or capacitive could work but i do not know of anything that would be easy to implement. the drive wheel system work fine if adjusted properly and as long as there is user feedback to make corrections. it is not good as an absolute position system, for that oyou need positive drive and either a truly adequate (and acceleration limited) mptpr or feedback from an encoder on the motor shaft.

Leave a Reply

Your email address will not be published. Required fields are marked *