Re: Aaah! (Magically Moving Steppers)

At 01:43 29/12/2002 +0000, you wrote:

>Aaah! I spent some time trying to fix a problem with my parts coming
>out lobsided, and just believe I may have found it.
>
>My system is on, then I turn on the Spindle of the Sherline, turn it
>up to max. The Sherline spindle is not connected to the CNC system,
>it's the plain old power switch on the spindle of the mill.. I turn
>the speed control up to max.
>
>I then watch the rear-shafts of the motors, and though they were
>holding still before, they now roll back and forth every few seconds
>by a small amount. Aaah! I'm thinking this is due to induced mains
>hum throwing the dir/step pins. Any advice?
>
>My system is a little limited still, so it's basically:
>
>Xylotex driver board
>Switched Mode PC power supply (12V driving the steppers)
>BF Cap 47,000uf
>Some big NEMA23 steppers
>
>Could it be the switched mode power supply in some way? It driving
>the logic also, I understand that switched mode PSU's employ PWM to
>provide the correct power level (just like a chopper board), could
>this be interefering with the system?
>
>Any advice greatly appreciated.

Sounds like radiated and/or conducted interference from the motor is
affecting the Xylotex board amd/or the PC you are using to drive it.
Putting the Xylotex driver board in a metal box will help with the radiated
interference to some extent, but it still may be affected by pickup via the
cabling. Adding ferrite sleeves to all the cables (both signal and power)
going into/out of the Xylotex board will help with any pickup from the
cables, and will also help with any conducted interference travelling via
the mains supply. If you use big enough sleeves, taking an extra turn of
cable through the sleeve (i.e, push the cable through, loop it back & push
it through a second time) increases the effectiveness of the ferrite
considerably.

Using shielded cable for the signals to the board, and for the stepper
motor drive cables, can also be a good idea.

One of the potential problems with switch mode supplies is that they can be
transparent to some forms of mains-borne interference, and the BF capacitor
may not help (won't help with common-mode noise, for example). The
transformer/smoothing cap combination in a linear PSU can be better at
getting rid of conducted interference. If the problem is with the switching
supply (easily tested by substituting a linear supply), then it is likely
to be due to this kind of conducted interference, rather than the PWM, as
the cap should happily smooth any PWM pulses.


Regards,
Tony