Z-Axis making noise

[2Bits]

When I jog thru the area where the majority of the time the z-axis operates during a job it makes a howling noise. Like a loose fit and dry so it can vibrate and make the noise as it travels thru this area. Speed seems to have no impact when jogging this area. This would be the lower end of my z travel. Not to be confused with a resonance issue. If I give it some extra lube, it will quiet down a little and then back at it. Jogging above that area is quiet and as expected.

While I didn't build this unit, I believe it to be a acme 1/2-10 lead screw with a round flange mounted AB delrin nut. Length of lead screw is about 12".
Z-axis plate & gantry is HPDE and holds a Bosch Colt router

So what's at fault? The nut or the screw or both?

Mark

** Update **
I should have looked closer, not a newbie, but should have looked closer at the hardware. The nut is a plain non-antibacklash delrin nut

[2Bits]

Really? No guesses, theories, thoughts?

Did I post this in the wrong place?

[ger21]

Mine makes a similar noise every now and then, mostly when it's cold. My guess is misalignment.

[mike_Kilroy]

Really? No guesses, theories, thoughts?

Did I post this in the wrong place?

I assume it is run with a servo motor not stepper motor..... if so, sounds like plain old classic backlash caused velocity loop instability. if a servo, I would bet on this.

a servo system is tuned for best performance. velocity loop tuning is a function of load inerta: more inertia, more velocity gain to goose the motor along so it responds fast. less inertia = less vel gain to do the same response. make sense? ok, now a 'rule of thumb:' velocity loop will be stable over about a 5:1 inertia change.

use example to see what this means: your motor has a relative inertia of "1" and say your load has an inertia of 10x the motor; hence total inertia is 11. but if you have an area with excessive backlash, the load can dissapear as it disconnects from the motor; hence the total inertia is now down from 11 to 1. so if your vel loop is tuned for the 11 inertia, it will now be unstable at the inertia of 1. result is oscillation - on ac servos this is usually at a 50-300hz rate so it makes the growl sound you hear.

How to fix?

either remove the backlash (buy a new $ 50.00 new delrin nut since it wears and the metal screw doesnt) or reduce your velocity gain to be be stable with no load - and so at load it will be spongy and not be so stiff anymore.

[ger21]

I'm pretty sure he's using steppers.

[mike_Kilroy]

I am not sure if this same thing occurs with steppers or not.

[2Bits]

Well the answer is:

I am using steppers

The Z-Axis was sort of mis-aligned. There was no end bearing or thrust bearing. Just out there free to the world (I didn't build this).

The nut which I was told was anti-backlash on all axis was just plain nuts and this one was worn. My guess to it being open ended.

So the cure was installing a bearing on the bottom of the Z axis and new anti-backlash nuts from DumpsterCNC.com. No more noise, nice and quiet and calibration was .001 to .002 on all axis when I finished. The X and Y were also misaligned and corrected.

For my first ever CNC, I guess this has been ok, as it's become such a huge learning tool to fix everything that was wrong with it. When I first got it, I was lucky to get within .015, I've now got it to .002 - .005


Mark