Re: Lost steps

Hi Mariss, Thanks for the noise info.
In my case the signal wires are not close to the motor wires except
maybe within the case, and there they are not bundles together. I'm
also pretty sure the problem is lost steps and not added steps. The
external motor wires are shielded but I don't think the shield is
grounded. Maybe I should do that on general good practise grounds.


I need more time to test out all the info I've been given here.
I suspect I'll find its the acceleration problem



--- In CAD_CAM_EDM_DRO@egroups.com, "Mariss Freimanis"
<geckohall@h...> wrote:

> Hi,
>
> Something else you might want to check for is noise. If your motor
> drive is a switching type (chopper, PWM, etc.)then you may be

getting

> noise pickup from the motor phase wires into the STEP and DIRECTION
> inputs. This can happen if these wires are bundled in a common

cable

> with the motor leads even for a few inches.
>
> Noise is always additive. The motor will move the required number

of

> steps plus whatever are contributed by noise. Obviously with missed
> steps you will be short of your position. This makes it easy to

test

> for.
>
> Home the axis in question. Check that position accurately. Make a
> large move away from home position. Measure that position

accurately.

> Repeat many times. If you are consistently "long" by a variable
> amount each time, it is almost certainly noise.
>
> How to correct. Check your cable routiong. Keep the outputs at

least

> an inch away from inputs. Check the step input with a scope if you
> have one. Verify a 0 to 5VDC swing on the input. See that the rise
> and fall times on the edges are clean and fast. Route the Step,
> direction and common inputs thru a shielded cable. Check the power
> supply voltage for spikes. Bypass the supply with a capacitor at

the

> drive if spikes are noted. Hope that helps.
>
> Mariss
>
> --- In CAD_CAM_EDM_DRO@egroups.com, Ozzie@h... wrote:
> > After a long slow process I now have CNC mostly setup on my
> Shoptask
> > machine. The setup includes EMC running on Linux Redhat 5.2,
> > Camtronics 5A chopper driver in half step mode, and 150 in-oz
> motors.
> >
> > In repeated attempts to determine the backlash on the axis, I

find

> no
> > consistency in dial indicater readings. On the X axis I lose
> > .002-.003" cumalatively in one direction. On the Y axis the
> readings
> > are erratic with variations in the.002" area.
> >
> > Before beginning I partially disassembled the carriges and cleaned
> > and
> > lubricated the ways, and adjusted the gibs. All seems to work
> > smoothly.
> >
> > I understand these motors are not power-houses, but they should
> > certainly operate the machine with no load on it.
> >
> > I'm using moves of a half inch each direction at 6 ipm for the
> tests.
> >
> > The motors came with the machine when I bought it used and were
> > powered by a box, one axis at a time, with "The Motion Group,

Inc."

> > on
> > it.
> >
> > The Camtronics driver tests well and puts out the correct

voltages.

> >
> > Any ideas, and how should I go about finding the problem????
> >
> > I'm a total newbie-dummy.