Re: [CAD_CAM_EDM_DRO] Re: what gear ratio ??? - more
Posted by
Doug Fortune
on 2003-07-23 22:42:51 UTC
cnchomeman wrote:
between the position of the leadscrew and the position of the
motor, thus probably leading to nasty oscillations.
True, mounting the encoder on the (geared down) motor shaft, will
increase resolution - this is good (generally).
quadrature per revolution)....
which is easily achievable by Mach2 (http://www.artofcnc.ca )
and achievable by TurboCNC (http://www.dakeng.com/turbo.html ) and
probably also EMC/Linux.
motor is woefully insufficient for heavy milling on a Bridgeport.
Luckily, 400 oz*in peak (times 3) is OK and hopefully 40*3 = 120 oz*in
will do for continuous light cuts. The real answer is either:
- a bigger motor
or
- higher gearing (say *10 ) in conjunction with a lower resolution encoder
(like 100 CPR).
But considering the difficulties (and extra backlash, setup & expense) caused by
compound levels of gearing, perhaps 3:1 will be sufficiently powerful to do
some useful cnc machining.
------------------
ONE OTHER THING:
Note that the above implies a 0.000 055555 inch/step accuracy. That might
be true if the motor driver kept the tracking (of the motor) exactly in
step with the desired position, but that doesn't happen. Some people use
a rule of thumb that says to expect around a +/- 10 count deviation
(perhaps a LOT more during highspeed moves, but thats usually while
rapiding and not cutting, so exact accuracy doesn't matter in that case).
Thus the expected accuracy is around 0.000 0555 * 10 or about 1/2 thou.
Again, this is a great number, considering the conflicting requirements.
regards
Doug Fortune
http://www.cncKITS.com
.
> Doug wrote:Because (if there is any gearing) there will be a discrepancy
> > Let me further say, a 10:1 ratio might be best with those smallish
> > motors, to get decent machining power at typical machining speeds....
> > however the rapids might suffer. Also, to get more than 2.5:1 or 3:1
> > you'll need more than 1 stage of belt drives, which complicates things.
> >
> > So all in all, try the 2.5 to 3:1 first and see if that works for
> > you. Put the encoder on the motor, not the leadscrew.
> >
>
> Why does the encoder need to be on the motor? For resolution?
between the position of the leadscrew and the position of the
motor, thus probably leading to nasty oscillations.
True, mounting the encoder on the (geared down) motor shaft, will
increase resolution - this is good (generally).
> If he has 300 PPR encoders thats 1200 lines and 5 pitch screwsI call that 300 CPR (cycles per revolution) and 1200 ciq (counts in
quadrature per revolution)....
> he is at .000 1666666" per step when its mounted on the screw.True.
> it's .000 0555555" per step when mounted on the motor.Also true (with a 3:1 reduction).
> He will need a lot of pulses to go 100" per minute.Also true. Say 120"/min == 2"/sec -> 36,000 pulse/sec
which is easily achievable by Mach2 (http://www.artofcnc.ca )
and achievable by TurboCNC (http://www.dakeng.com/turbo.html ) and
probably also EMC/Linux.
> Or is there some other reason?The reason for the gearing down, is that a 40 oz*in continuous
>
> Eric
motor is woefully insufficient for heavy milling on a Bridgeport.
Luckily, 400 oz*in peak (times 3) is OK and hopefully 40*3 = 120 oz*in
will do for continuous light cuts. The real answer is either:
- a bigger motor
or
- higher gearing (say *10 ) in conjunction with a lower resolution encoder
(like 100 CPR).
But considering the difficulties (and extra backlash, setup & expense) caused by
compound levels of gearing, perhaps 3:1 will be sufficiently powerful to do
some useful cnc machining.
------------------
ONE OTHER THING:
Note that the above implies a 0.000 055555 inch/step accuracy. That might
be true if the motor driver kept the tracking (of the motor) exactly in
step with the desired position, but that doesn't happen. Some people use
a rule of thumb that says to expect around a +/- 10 count deviation
(perhaps a LOT more during highspeed moves, but thats usually while
rapiding and not cutting, so exact accuracy doesn't matter in that case).
Thus the expected accuracy is around 0.000 0555 * 10 or about 1/2 thou.
Again, this is a great number, considering the conflicting requirements.
regards
Doug Fortune
http://www.cncKITS.com
.
Discussion Thread
Doug Fortune
2003-07-22 23:13:22 UTC
Re: [CAD_CAM_EDM_DRO] what gear ratio ??? - more
cnchomeman
2003-07-23 22:00:06 UTC
Re: what gear ratio ??? - more
Doug Fortune
2003-07-23 22:42:51 UTC
Re: [CAD_CAM_EDM_DRO] Re: what gear ratio ??? - more
cnchomeman
2003-07-23 23:22:00 UTC
Re: what gear ratio ??? - more
Jon Elson
2003-07-24 10:39:09 UTC
Re: [CAD_CAM_EDM_DRO] Re: what gear ratio ??? - more