Hello Guest it is March 28, 2024, 09:02:41 AM

Author Topic: Kernal Speed and Chatter  (Read 10746 times)

0 Members and 1 Guest are viewing this topic.

Re: Kernal Speed and Chatter
« Reply #10 on: September 13, 2009, 03:37:23 PM »
VMAX-

Respectfully, I have to disagree.

If you have a machine with a lot of mass and the accel rate is set too high then a rapid acel will cause your machine to jerk, then this will result in what I am interpreting as chatter.  Perhaps I am mistaken or using the incorrect terminology.  If so, I apologize in advance.

While what you say is partially true because the "chatter" it is due to the mass limitations of the machine, the software allows us to slow down our accels to a point where things become smooth.

CV takes this away from us to a certain extent because Mach now want to control velocity on it's own terms (well sort of) but we certainally do not have the same speed/accel control as in exact stop mode.  If 2 or more of your axis are dissimilar in terms of their tuning profile, you will see Mach slowing down one axis a lot more rapidily than the other in an attempt to keep things "Constant"  this struggle can result is Jerky motion (exactly what CV is supposed to prevent).  CV is a double edged sword under these circumstances.

Sid

"If you have a machine with a lot of mass and the accel rate is set too high..." - And how is that a software problem??  Is the software supposed compensate for you setting the acceleration too high?

I agree with Terry.  This is a machine problem, not a software problem. 

Regards,
Ray L.
Regards,
Ray L.
Re: Kernal Speed and Chatter
« Reply #11 on: September 13, 2009, 03:53:45 PM »
Hey Ray,

Thanks for chiming in.

Perhaps I could have been more clear.  In my previous post, I was using accel being set to high as an example of how chatter could occur.

Of course software can not compensate for incorrect settings.  However, it has been my experience when Mach is runing in CV, the speed blending which occurs to create the 'Constant' somehow overides the acel parameters in tuning in its attempt to run every axis at the same speed.  This causes undesirable effects.

Yes- I could adapt my machine so that the Axis were more similar in terms of the tuning profile but, it sure would be nice (read easier) to have some more control over CV (maybe like a CV axis scalling factor) to help smooth things out.

Sid
Re: Kernal Speed and Chatter
« Reply #12 on: September 13, 2009, 04:18:53 PM »
Hey Ray,

Thanks for chiming in.

Perhaps I could have been more clear.  In my previous post, I was using accel being set to high as an example of how chatter could occur.

Of course software can not compensate for incorrect settings.  However, it has been my experience when Mach is runing in CV, the speed blending which occurs to create the 'Constant' somehow overides the acel parameters in tuning in its attempt to run every axis at the same speed.  This causes undesirable effects.

Yes- I could adapt my machine so that the Axis were more similar in terms of the tuning profile but, it sure would be nice (read easier) to have some more control over CV (maybe like a CV axis scalling factor) to help smooth things out.

Sid


Sid,

I don't believe that's true, and it's easy enough to test - set the accel on one axis artificially low.  I think you'll see that ALL "blended" operations slow down accordingly.  If Mach ignored the accel on even one axis, we'd having servos faulting and steppers losing steps randomly, but that does not happen, even when there are wide disparities in the capabilities of different axes.  One of my machines runs an accel of 125 on two axes, with a max speed of 270 IPM.  The third axis has an accel of only 5, and a max speed of 30 IPM, and those settings are right on tha hairy edge of what it's capable of without losing steps.  On blended moves, the thing crawls, because of the third axis, but it does not lose steps.

Regards,
Ray L.

Regards,
Ray L.
Regards,
Ray L.
Re: Kernal Speed and Chatter
« Reply #13 on: September 13, 2009, 04:35:45 PM »
OK, I'm not workin today but I'll give it a go tomorrow.

vmax549

*
Re: Kernal Speed and Chatter
« Reply #14 on: September 13, 2009, 05:15:39 PM »
Actually the CV helps to maintain a constant feedrate AND that is a desired effect when milling. IT will make a smoother more consistant cut with proper CV.  AND the mach cv is normally BEST left in the default state with 200 lines look ahead(;-)

THe jerk is caused by improper axis tuning for the machine specifications OR a non ridgid machine that creates the jerk on axis reversals.

Mass has little to do with it and rigidity is not always a function of mass. Mass can help you OR hurt you depending on the design of the machine and WHAT you are trying to do with it.

SO back to the software chatter (;-) For the most part I don't think so.

Cutter chatter is normall a function of the machine process verses the rigidity/mass ratio of the machine AND the proper cutter selection and feed/speeds and DOC, tool length ratio. The list goes on and on. Get it right and smooth as glass get it wrong and CHITTY CHITTY bang bang.

R what you are calling chatter is not what I call chatter(;-) It could be Gouging is what you are seeing???

(;-) TP
« Last Edit: September 13, 2009, 05:20:58 PM by vmax549 »
Re: Kernal Speed and Chatter
« Reply #15 on: September 14, 2009, 09:39:12 AM »
Based on my reading of the CV Notes, I made the following changes to my Mach configuration:

CV Look Ahead 200 lines.
CV Dist Tolerance 100 units
Stop CV on angles > 90 degrees.

After restarting, I cut my Diamond at both 35 and 70 IPM and the cuts were smooth.  But when I cut at 100 IPM, I still had chatter.

I have since changed the CV Dist Tolerance to 50 Units.  I recut the Diamond at 100 IPM and I still had the same amount of chatter.

My X and Y motors are tuned to 450 Velocity and 30 Acceleration and Z is 150 / 10.

I would like to hear more from those that have some experience with these configurations changes.

For those that don't think software has an effect.  Yes, it does not cause chatter but it affects the way the machine operates.  For example if you have CV set and you cut a rectangle and a very high say 400 IPM, you will get rounded corners.  Changing to Stop will give you square corners.

vmax549

*
Re: Kernal Speed and Chatter
« Reply #16 on: September 14, 2009, 10:31:37 AM »
WE never said it would NOT effect how your machine ran. But it has little to no effect on chatter.

What is your spindle speed? The faster you try to cut the faster the spindle speed has to be to maintain the proper chip load.

(;-) TP
Re: Kernal Speed and Chatter
« Reply #17 on: September 14, 2009, 12:29:10 PM »
Sail, give us more information on the material you are cutting, the axial and radial depth of cuts, the type of cutting tool and tool holder you are using, how far the cutting tool is extended from the holder, what spindle RPM you are using and what type of machine you have. This will make it easier to figure out what is causing the chatter.

Paul T.
Re: Kernal Speed and Chatter
« Reply #18 on: September 14, 2009, 01:06:07 PM »
As stated by Vmax & Titchener above, I agree that you are probably over stretching the Feed rate due to low RPM. Also, with such small sizes I wonder why 1/2" EM ? & O flute ? There has been no mention as to the work holding or the remains of the material during the So-Called Chatter. I would suspect that what you are calling Chatter is not really what or where your describing. Deflection might be in play as well. Or even dull cutter induced surface character. I didn't see what material your cutting, could be surface work hardening, or grain tearing, or chip re cutting ! It could be any number of things involved here and without a visual or better description, then you could be easily leading all who try to help, astray ! And you end up without a viable solution.
CRUISER
  Don
Re: Kernal Speed and Chatter
« Reply #19 on: September 14, 2009, 01:07:41 PM »
Paul,

I am cutting .5" solid surface material.  I am using a .25" Solid Carbide UpCut Spiral O Flute Onsrud 63-775 bit.  I am cutting .125" with each pass at 90 degrees to the surface of material.  I am using a Milwaukee 5625 router in my machine and it has a standard .25" collet.  The bit has a CEL of .75" and only the cutters are showing out of the collet.  The router is set at 14,000.  The machine is a MechMate.

I think I understand and have answered all your questions.  

Thanks for you help.

Cruiser,

I am cutting with a bit that recommended by Onsrud for solid surface material.  I need to use .25" bit for details.

The little piece missing is from a tab that broke off.
« Last Edit: September 14, 2009, 01:25:13 PM by SailFl »