Re: [CAD_CAM_EDM_DRO] Re: Tooth belts & gear suppliers

> In MY opinion, the added "slop" would introduce error in EITHER direction=20
> between the motor and the lead-screw position, equal to two or three steps.=
> =20=20=20
> Of course, this would be "non-cumulative", I guess, but "sloppy" for=20
> machining!=20=20

Which probably gets masked by backlash in most of the machines featuring
such belts.

These kinds of errors are not ones that can be seen by eye. I
routinely put a dial indicator against the toolpost and make test
moves with the cnc control in MDI. Sounds like maybe some tests with
various amounts of weight hanging over a pulley to simulate cutting
forces are in order. This will be very hard to measure though as an
acme screw is not backdriveable. What I probably will measure is the
small amount of slop known to be in the worn nut mounting. But if the
belt slop is present, it should be measureable by commanding a move
under static load.

For a servo machine, I could see some wisdom in putting the encoder on
the screw... heck, I'm tempted to use a long V belt to the motor on a
certain temporary ("computerized power feed" class) project.

Chris

-------------------------------------------------
Christopher C. Stratton
Engineer, Instrument Maker, and Horn Player
ccs@... 617 628 1062
http://web.mit.edu/~stratton/www/brassbuild.html