On Tuesday 19 January 2016 03:35:32 John Dammeyer wrote:

> Hi Chris,
> Wiki defines Servo Motor as a motor that has closed the loop so it
> knows the position of the motor.
>
> Years ago there were discussions about closed loop with linear scales.
>  The trouble there is the amount the motor has to turn before the
> scales report data.  That can cause oscillations that are hard to
> damp.
>
> There is no reason a stepper cannot be the drive part of the motor
> with an encoder feedback.  Google TXM24C-1CG for example.  And there
> are others. The current name for that type of drive is step-servo.
>
> The UHU Servo driver http://www.uhu-servo.de/servo_en/ explains how to
> use a DC motor with either a 0 to +/- 10V or PWM and DIR.
>
> One of the first electronic lead screws also used a DC motor on the
> lathe lead screw.  Using counters connected to the spindle motor and
> the lead screw motor the difference between the counters is an error
> value that adjusts the DC voltage out to the motor.
>
> In each of these examples a varying DC voltage or PWM to a DC motor is
> the control.  A brushless DC motor still essentially works the same
> way too. Supply a voltage or PWM and the motor turns.
>
> Not so with stepper motors.  You essentially need to first create a
> driver that effectively turns the motor at a specific RPM based on a
> voltage input. Then you can close the loop with the desired number of
> steps in verses the encoder counts as an error that moves the motor in
> the correct direction for the required distance.
>
> John Dammeyer
> .
>
> > -----Original Message-----
> > From: Chris Albertson [mailto:[email protected]]
> > Sent: January-18-16 11:04 PM
> > To: Enhanced Machine Controller (EMC)
> > Subject: [Emc-users] Question about closed loop CNC
> >
> >
> > I've read that it is hard to use steppers for closed loop CNC.  If
> > this is still true could some one explain the theory.
> >
> > I have a small mill with DRO scales that are good to about .001 inch
> > and I'd like to put them to use with my planned CNC conversion.   I
> > understand software.  I've worked in the field for 30 years and have
> > experience with Linux (and other OS) device drivers, embedded real
> > time code and so on. If the problem is simply that no one has
> > bothered to write the code, I can fix that.   But if there is
> > something of a mechanical nature I'm not thinking of maybe some one
> > can point me at what I need to read up on.
> >
> > At least I think I might be able too automatically measure backlash
> > using the DRO scale data.  But my goal is to go closed loop if I can
> >
> > --
> >
> > Chris Albertson
> > Redondo Beach, California

Something that has not been mentioned here is that a real error because 
the stepper has lost a full step or more because it is being asked for 
step movement faster than it can maintain lock.

When that occurs, generally the motor stops/locks, and thestepper 
generator must be stopped and re-started from a low speed and the 
controlled acceleration re-applied.  But if its a multiaxis move, there 
will be a jog in whats supposed to be a straight line unless the whole 
machine is stopped while the errored axis is getting back into position.  
It would be a good idea to stop the errored axis at the caught up 
position, then restart the whole move so that it can proceed along the 
selected co-ordinate path, but because it errored at the selected speed, 
the new speed ought to be 10% slower until that move is completed.

Hopefully the part will still be usable, but you can only determine that 
after the finishing cuts.

Steppers can be persnickity. 2 things can help if the motor loses steps.
First is the signal path. Going thru 2 or 3 sets of opto-isolators can 
and will play hell with the step timings. So I, once I had identified 
that as a contributing factor, got rid of the isolators except for the 
ones actually in the input circuit to the driver module.

Second is the supply voltage to the motor. Example:
My toy mill, a small HF with bigger tables and ball screws on the XY, 
8mmx2.5 screws, is apt to act up at just above 10 ipm as it only has a 
28 volt supply for the motors.  A highly regulated 300 watt switcher 
with a 5 volt output to power the BoB.

When I did my toy lathe, I made a un-regulated supply of about 38 volts 
when loaded  but used the same motors and 2m542 drivers.  Its Z has a 
16x5 screw, with a 40 tooth gear on the motor turning an 80 tooth on the 
z screw. So as far as the motor is concerned, its turning a 2.5 pitch 
screw.

Where the toy mill can have trouble at 10 ipm,  the lathe, with its 
nominally 10 volt higher motor supply, can move the z at better than 60 
ipm!  I have a bigger supply made for the toy mill, some 50 volt 
supplies reset for 42 volts that will be installed when warmer weather 
comes back as that building isn't insulated, and I only keep it warm 
enough to stay above the dew point with electric heaters.

To my way of thinking, using steppers is 100% ok, and if the speed goal 
isn't met without stalls, put a damper on it if the stall is at a low 
speed, and raise the voltage until it Just Works(TM).  I tried one on my 
470 oz/in nema 23 motors, and a spare driver, hooked to that 42 volt 
supply, and the stepper ran quite well up to the limit of the 2m542's 
opto inputs.  Set at 8 microsteps, or 16, even 32 was tried, that speed 
limit was at the 350 kilohertz mark on the function generator dial. The 
motor could be made to turn something above 3100 rpms.  Slower at the 
higher microstepping because the opto speed limit had been reached.

2 optos in series because the BoB you bought is also an isolated version, 
_will_ hurt you. Particularly if using a PWM output for spindle speed 
control, you _WILL_ have to get rid of it if you want anything that even 
resembles linear speeds so the PID can control it right.
Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>

------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=267308311&iu=/4140
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to