Hello Guest it is August 17, 2019, 11:39:16 AM

Author Topic: Need help with resonance problem  (Read 6164 times)

0 Members and 1 Guest are viewing this topic.

Need help with resonance problem
« on: November 04, 2009, 05:42:17 PM »
Hello all,

I've converted a Grizzly wood lathe to CNC by adding some scrounged actuators with step motors.  I'm having a problem with the Z-axis.  It gets a few steps out of sync with Mach3 and starts cutting where it shouldn't.  In a run of 15 minutes it gets off maybe as much as a dozen steps.  Computer is fine.  Electronics is fine.  These are my background and I wish the problem was here because I could deal with it.  It's not.  It's mechanical.

I slowed the Z-axis *waaay* down to speed of 25 and acceleration of 2 and the problem went away.  So, now I have to figure out how to identify the resonance and how to deal with it.  This is an area I have no experience in.  Can someone point me to a recipe or tutorial on how to identify, measure and compensate for resonance?  I'm hoping I don't have to do this by trial and error because that could take forever.

Now, there's a Lovejoy-style coupler between the step motor and leadscrew.  I'm suspicious of the rubber spider between the two halves.  Seems to me that this could exacerbate resonance problems.   The leadscrew is almost 30" long and at least 1/2" dia., pretty massive.  The stepper's 1/4" shaft looks puny next to the leadscrew, and it is probably twisting under the torque and contributing also.  Would a larger stepper with a more robust shaft help?

Any help will be appreciated.

Dave
Re: Need help with resonance problem
« Reply #1 on: November 04, 2009, 06:20:10 PM »
You've verified that it behaves if BOTH velocity and acceleration are cranked way down, but doesn't sound like to verified WHICH one is the real problem.  In a stepper, resonance is more likely to cause a stall than a few lost steps.  Sounds to me like you're just seeing lost steps due to over-driving the the axis.  Leave acceleration very low, and see how high you can crank the speed and still get reliable results.  Once you find the limit, back off by 30-50%.  Now do the same with acceleration.

What stepper drivers are you using?  Are you microstepping?  If so, by what factor?  Any good stepper driver will have mid-band resonance correction,  but it sounds like your problem is occurring at higher speeds.  Microstepping will help mitigate high-speed resonance problems.  Mechanical dampers should rarely, if ever, be required in a properly setup system.

Regards,
Ray L.
Regards,
Ray L.

Offline RICH

*
  • *
  •  7,351 7,351
    • View Profile
Re: Need help with resonance problem
« Reply #2 on: November 04, 2009, 07:18:22 PM »
Just for info, as the stepper motor speed is decreased the motor torque goes up. Now as you increase the
speed of the motor the torque decreases.  So when you decreased the velocity setting, there was more torque
/ power available to be delivered to the axis to overcome cutting, axis binding, whatever.......
 
Everthing has a natural  frequency, even you, so the screw / shaft has one. Based on the support conditions of the screw , the end mounting conditions, the deflections of the shaft, and rpm, the combination can create a repeatable vibration ( represented many times as a mass on a spring osciilating up and down ).  Should the vibration / "frequency" also be the same as the natural frequency or a harmonic of the shaft, a situation occurs such that the shaft can become resonant. Continued vibration / stimulation of the shaft while running increases the resonance and  the forces generated can become extremely large. Those forces will be transmitted into the supports.
So that may be summarized as critical speed.

You would be shocked / frightened to see the force due to resonance. So i don't think your scew is "resonating" .
The rubber type coupler would actualy tend to greatly reduce the forces into the motor.

RICH
 



Re: Need help with resonance problem
« Reply #3 on: November 04, 2009, 09:38:25 PM »
This motor is German, Nanotec - Munich, 4H5618C0408, 3.4V, 2.85A, 1.8deg.  NEMA 23. 

 I'm not microstepping it.  The controller can do half and quarter steps if that will help.

The error is always the same direction.  I.e., when I stop it to check position, the carriage is always farther away from home than it should be.

In playing around with jog mode, I was able to run the stepper at 900 and acceleration of 120 without obvious stalling.  But in actual use I was still seeing errors down at 250 and 40. So I cut it to 25 and 2 and the errors stopped.

I'll generate some code to run it back and forth short distances, lots of reversals, for use in determining the velocity and acceleration at which the errors begin.  Given the mass of the leadscrew relative to the size of the stepper, I'm expecting that acceleration is going to be the problem.

Should I be half-stepping to improve performance?

Dave
Re: Need help with resonance problem
« Reply #4 on: November 04, 2009, 10:08:45 PM »
When I converted my Sieg X2 to cnc, I started with an inexpensive drive from Stepperworld, which sounds similar to what you have, Shortly after that, I bought a Xylotex controller, and set microsteping to 1/8, and was much happier.  three years ago, I bought some gecko drives, and a toroid transformer to drive the motors at 60 volts,,,,,,, I finally put it all together a month ago,,, and could kick myself in the..... for not doing it sooner.  The board you have sounds like a unipolar drive,,, it energizes the coils only one way,,, Xylotex and Gecko uses unipolar drive which flips the polarity of the coils back and forth,,, the xyltex board is limited to 30 volts, and the geckoes 80 volts. 
Re: Need help with resonance problem
« Reply #5 on: November 04, 2009, 10:20:17 PM »
This motor is German, Nanotec - Munich, 4H5618C0408, 3.4V, 2.85A, 1.8deg.  NEMA 23. 

 I'm not microstepping it.  The controller can do half and quarter steps if that will help.

The error is always the same direction.  I.e., when I stop it to check position, the carriage is always farther away from home than it should be.

In playing around with jog mode, I was able to run the stepper at 900 and acceleration of 120 without obvious stalling.  But in actual use I was still seeing errors down at 250 and 40. So I cut it to 25 and 2 and the errors stopped.

I'll generate some code to run it back and forth short distances, lots of reversals, for use in determining the velocity and acceleration at which the errors begin.  Given the mass of the leadscrew relative to the size of the stepper, I'm expecting that acceleration is going to be the problem.

Should I be half-stepping to improve performance?

Dave

You should definitely go to quarter stepping.

Regards,
Ray L.
Regards,
Ray L.
Re: Need help with resonance problem
« Reply #6 on: November 05, 2009, 02:06:19 AM »
I found out my controller supports 1/8 steps, so I used it and indeed, it solved the problem.

And, now that I see it in operation, it makes perfect sense.  While I was still at full steps I watched the motor as it ramped up and down.  There was a discrete speed, fairly slow, where I could actually see the motor stumble and reverse briefly.  Made more noise too.  Thinking about it, it makes perfect sense.  Those full steps were huge slugs of torque that were exciting the resonance of the system.  Using smaller steps avoids that excitation.  Everything is smoother and quieter.

Thanks for the advice.

Dave