[outsider787]

On my newly CNC converted BF20, on the Y-axis alone, over a range of several inches, there's some vibration of the table when the table is moved over a certain area of its movement.

When I look from the back (where I can see the screw) I can see the screw vibrate quite a bit over that portion of movement.

I tried greasing up the ways and greasing up the screw. No change.
If I drop the ipm speed down a lot, the vibration stops. As it is, it's not that fast and still vibrates.

I don't know what else to do.
Any advice?

Here's a short video outlining the problem. (the volume is not the loudest)

CNC vibration - YouTube

[Winnfield]

What is that thing behind the coupler that is wobbling like crazy? Did you use angular contact bearings in your ballscrew end mount?

[Winnfield]

It sounds like the motor is missing steps.

[outsider787]

What is that thing behind the coupler that is wobbling like crazy? Did you use angular contact bearings in your ballscrew end mount?

I am using the stock screw (for now) that the machine came with.
That thing that (appears to) wobble is a spacer between the nut an the the bearing (which replaces the hand crank). It's made out of square stock, so 'it appears' wobble. It's not.

[Winnfield]

Did you measure the actual distance traveled and compare it to your command? That will tell you if it's steps missing.

[LUCKY13]

Which drivers are you using fo rthose steppers? That sounds like midband resonance.



Jess

[outsider787]

Jess,
My specs are: (taken from Hoss's website as recommended)
From Keling:
3 x KL-6050 (drivers): $41 ea
3 x KL23H2100-50-4B (steppers): $55 ea
1 x KL-350-48 (power supply): $50

Ever since I bought the BF20, the y-axis seemed a little tighter then the x-axis.
It almost felt as if the gibbs were slightly engaged when moving the table in the y direction (the gibbs were not engaged, but felt that way).

I don't know if that translates into this.

adrian.

[DRock]

My G0704 will vibrate during a certain section of the Y-axis when I am hand-cranking it at a certain speed.

I am not sure what the cause of this problem is but try either going up or down slightly with your feed rate and see if anything changes, you might just be matching the resonant frequency causing a little vibration.

[LUCKY13]

That does sound like there is something inherant to the mill base.


These mills seem to outwork other mills in the same size/weight. I have always expected that the way the column mounts has something to do with there unexpected ridgidity.


But this makes me wander now if there may be a down side to the column being mounted on the rear of casting. Possibly causing the base to bow?

Maybe its just the size of the casting is just right to amplifly the resonance freq of the screw ( acting like a speaker).



I would check that Y axis to see if its straight/flat. But you could try adding a flywheel(balancer) to the stepper motor. There is a thread in the Gecko section about this and a symple balancer added to the end shaft of the stepper would cure vibrations like this. Now they would be caused by drivers that didnt have mid band res correction, but they still cured it and it might work for yours.


The drag you meantioned does make me wander though about the base being warped/bowed. The screw being mounted crocked could also act that way. As far as the vibration I would also try mounting the mill down to the bench if its not. That is a unusaual vibration if its not being cause by the stepper driver ( not having mid band res correction).


Jess

[outsider787]

How would I find out if it's flat? What am I looking at? The dovetail straightnes along the axis?

I was looking a little more carefully at the loose end of the shaft (behind the table) through the portion of vibration, and the loose end of the shaft doesn't just vibrate, it really wobbles. This is all happening when the table is moving away from me, which is putting the shaft in compression.
I'm thinking that either the threaded shaft may be bent (any amount would cause vibration) or that I don't have the nut tightened enough on the shaft against the thrust bearing.

I can do about 20-25ipm on the y-axis. Anything higher is a no go.
And yes, the stepper does skip steps at the higher feed rate.

I can live with the 20-25ipm for now, at lease until I take things apart to have another look at the y-axis bearing attachment.

[RTTIM]

A couple rules of thumb that might help you:

1. If the issue is resonance the vibration will peak at the speed where the natural frequency of the component is being excited and will fall off both below and above that speed.

2. Conversely, if the vibration level continues to increase as the speed is increased you're looking at some type of mechanical issue that is affecting the load on the screw. It could be one or more of a variety of issues, including imbalance, misalignment, a bent screw shaft, mechanical binding of the slide or a bearing, etc.

Do you know anyone with a vibration data collector or vibrometer? I know it might be overkill, but ideally, a FFT of data collected at the several speeds would give you the most information. If you order normalize, then overlay the plots from several speeds it will give you a real good idea of the fault frequencies and trend.

Tim

[LUCKY13]

A couple rules of thumb that might help you:

1. If the issue is resonance the vibration will peak at the speed where the natural frequency of the component is being excited and will fall off both below and above that speed.

2. Conversely, if the vibration level continues to increase as the speed is increased you're looking at some type of mechanical issue that is affecting the load on the screw. It could be one or more of a variety of issues, including imbalance, misalignment, a bent screw shaft, mechanical binding of the slide or a bearing, etc.

Do you know anyone with a vibration data collector or vibrometer? I know it might be overkill, but ideally, a FFT of data collected at the several speeds would give you the most information. If you order normalize, then overlay the plots from several speeds it will give you a real good idea of the fault frequencies and trend.

Tim

RTTIM, your post makes good sense. This vibrometer you speak of I expect would be something hard for most people to get access to. Though he may be able to find such.

Something I havnt tried, but it may be possible, is taking the knock sensor from a automotive aplication and hooking it to a volt meter. It would be a live reading though unless something like a storage scope was used, but still it could give feedback of whats going on.


It could be a lot easier than all this . The flywheel/balancer I meantioned would be a very easy quick test to see if its mid band ress causing the problem. Something as simple as a round disk of aluminum on the end of the shaft of the stepper. Then drill and add bolts to the outer edge of the disk. On these balts add some washer that can vibrate when there is vibration. This will kill mid band ress problems very easy.

Here is a thread that explains and shows some guys killing vibrations in system with these simple tricks. Its a pretty long read but it clearly shows these balancer/dampeners fixing the problems.

http://www.cnczone.com/forums/steppe...er_damper.html


Now it may be possible that there is a mid band ress problem, plus parts of the machanical drive, and or the mills design making this mid band problem even worse. Either way if a simple balancer helps it shows atleast part of the problem. I do expect any slack or missaligned parts could add to a problem like this, or even bring it on when there would be no problem. WIth others saying they have expeiranced the same thing in this mills Y axis you have to consider it being part of the mill structure itself.



Outsider, yes the dovetail flatness would be something your looking for. The axis should not be stiff and if the base has twisted or bowed the dovetails would be effected. Ofcourse slack in the bearings & the screw not being centered could be a problem also. Its very possible that there is a build up of differnt things adding up to this vibration. I would by all means read the thread I listed and try the simple fix for mid band driver ress. The of course fixing any machnical problem you have will just make it work and perform better to. These machines are a collection of different systems so it all has to work together. Just like with a ballscrew drive you cant get zero backlash if the end bearing supports are moving around with clearance. ANd you cant have a smooth axis with a screw that doesnt run centered ( or a accurate one). Ans if a dovetail is not flat/striaght it will have tight & loose spots in it. ANy time you have any loose spot in any machnical part vibration can happen.


Jess

[RTTIM]

Jess,

I agree that the place to start is with the basics - make sure the machine is level and mounted correctly, soft foot compensated for, etc. so it's not deforming the base.

Outsider's comment that the y axis has always "seemed a little tighter" is worth noting as you pointed out. As you suggested, I'd look at the flatness/straightness of the dovetails. It could also be misalignment of the screw bearings and the nut in the table. I might even pull the screw and check it for run-out to make sure it isn't bent.

A problem with the stepper motor can be diagnosed by physically swapping the X and Y axis motors - if the problem disappears it was the stepper, if not, it's probably in the drive screw mechanism somewhere.

The important thing is to be methodical, take it one step at a time and keep good notes.

Tim

[outsider787]

Thanks for all the advice guys,
I am pretty sure it's not the stepper, since this only occurs over a 3-4" section of movement, and it's only in one direction.

I will take apart the bearing block and inspect that, as it's easier to get to then the nut, and see if there's any problems there.

adrian

[DRock]

I solved mine by taking it apart and cleaning/lubricating all surfaces.

If your screw wobbles at the end this is NOT GOOD as it should be rotating perfectly along its center axis.

[outsider787]

I took a deep breath earlier tonight, and decided to get to the bottom of this.

Took apart the y-axis bearing block, removed the screw, then slid the table back and forth. There certainly is one section where it's a little tougher to slide the table. It's not a uniform smoothness. I adjusted the dovetail adjustment screws, and came to a happy medium. The difference in tightness is not night and way, but certainly noticeable when sliding the table around by hand.
Cleaned all the surfaces, re-greased everything, and re-installed the bearing block and tightened the jam nut against the bearing until it the bearing was the right tightness.

While I was at I also adjusted the tightness of the screw nut. This one was really not very tight at all.
Once everything was back together, I'm happy to report smooth operation on the y-axis!

I think the initial problem was that the shaft was that the jam nut (where the coupler attaches to the shaft) was not tight against the bearing and that looseness is what I was seeing as the shaft wobbled, and crated the vibration.

I can now do nearly 2500 mm/min (100ipm) to match what I could do on the x-axis all along.

Thank you everyone for all your help.