There have been a couple of articles describing this reverse motion effect.
A motor with a disturbed motion, either from a scrambled pulse stream or
from an external disturbance, can spin with significant torque, in the
opposite direction to the pulse stream at a speed of three times that which
would be expected from the pulse rate. The combination of a motion of 3 full
steps in the wrong direction by the rotor with an advance of the magnetic
pull of the stator in the wanted direction causes the rotor and stator to
again be in sync and to produce torque. 

A single reversed direction pulse when the motor is running can easily cause
a two step position error (one from the error the step was about to correct
to keep producing torque, the second one from the wrong pulse). Particularly
at lower speeds, the rotor will oscillate after each step. If the wrong
direction step happens to occur as this ringing is rebounding from an
overshoot, the motor will continue in this reverse direction locking in at
the 3x rate as described above.

Look for the single reversed pulse problem - it may well be the whole
problem!

Don Labriola

-----Original Message-----
From: Ed Nisley [mailto:[EMAIL PROTECTED] 
Sent: Thursday, November 30, 2006 10:58 AM
To: [email protected]
Subject: Re: [Emc-users] Mysterious direction-signal changes
-SNIP-
One possibility is that my controller / motors are unduly 
sensitive to starting out in the wrong direction. You can 
convince a stepper to spin the wrong way if you torque it 
enough; it doesn't sound happy, but it'll run the wrong way 
at least for a while after you spin it up.

What may be happening is that the motor gets up enough steam 
in the wrong direction and loses lock when the direction 
suddenly flips over. At that point it ramps up in the wrong 
direction; the fractional stepping torque isn't enough to 
overcome being in backwards sync with the whole-step poles.

Also, I have seen on the real 'scope (but not recorded, so 
apply a salt shaker here) invalid direction signals lasting 
for tens to hundreds of milliseconds. That'd be enough to 
ramp up to speed the wrong way, then go -thunk- when the 
direction signal flips back the right way.

Throw in another blast from my footgun and it might all be 
the same problem under the covers.

Onward...

-- 
Ed




-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to