Vibration of threaded rods (z-axis)

Hello dear Mendel90-User,

I started 3D-printing with the Mendel90 some weeks ago and I'm happy with Nopheads design and the delivered kit.

My current activity is to find the best setup to be consistent with my wishes.

During the different tests with higher speed / acceleration settings I had a lot of vibration of the two threaded rods (they have a low resonance frequency depending on the boundary conditions - rod dimension, bearing, prints near the heated bed, etc.).

In a first step I reduced this vibration by two additional mounts with ball bearings (D22/8 x 7) and assembled the mounts on the free ends of the thread rods. Of course, it would be fine to combine the different mounts (axis / thread rod) in one part, but for now it is o.k.

If someone has the same problem, please feel free to download the files on thingiverse.
upper z-axes support

Regards
Martin

PS: In some cases it can be help to reduce the noncircular rotation of the rods (if depending on the clamps of the microstepper / provided that the thread rods are straight).

Seems like a very bad idea to me and a classic mistake people keep making with RepRap Z axes. They will over constrain the lead screws and likely cause Z wobble on the prints.

---

[www.hydraraptor.blogspot.com]

Hi Nophead,

if you print the parts in the right way, there is a possibility for movements in axis direction, so it is not over constrain. When you fixture the rod extension to the ball bearing you are right. The goal is to reduce the very high amplitude when it runs nearly eigenfrequency. Certainly I'm open for improvements of this parts.

Regards
Martin

Edited 1 time(s). Last edit at 06/03/2013 06:48PM by Bazooka.

If the free end vibrates why would it matter?

If the free end is constrained and the rods are not perfectly straight they exert a force on the smooth rod which will displace it slightly.

---

[www.hydraraptor.blogspot.com]

I'm sorry Martin but nop head is 100% correct on this issue. Trying to constrain the free end of the threaded rod is always a bad idea.
What did sound like a better idea to me was this one - [forums.reprap.org] - although even then, with the right sort of flexible tubing in the connector, it probably won't make a huge difference.

Hi Nophead,

I understand you are not happy with this modification and the printer is certainly running without the additional bearings on the upper side. I agree with you that a fixation in all direction (over constrain) isn’t helpful.

About my motivation:
Every type of vibration results in additional forces within the structure. The forces take effect on the microstepper and stepper bearing, on the brass nut and thereby also on the guide bars. The forces wich are depended on resonance are very difficult to calculate, so my motivation is to find an improvement to minimize these effects by increasing the frequency.

First option is to minimize the acceleration (reduces the excitation). The second possibility is to find a way which increase critical frequencies by using a suitable system. The calculated resonance frequency of the thread rod is at Z on bed position appr. 45Hz, at Z=200 mm appr. 240 Hz, with the additional support between 200Hz (Z=0) and 400Hz (Z=200).

By wobbling you generate also forces within the structure and the upper support has only an influence on the force path. The best solution would be a perfect geometry of the thread rod and couplings, but I know how difficult it is.

Would it be an option to use the upper support without the screws, that means the upper support will be more flexible (fixation only with the small clips)? This will also result in higher frequencies.

Regards
Martind

PS: During my previous trials i could find only a positive behaviour, but I need more experience about the upper support.

The easiest way to make the resonant frequency higher would be to use 8mm threaded rods, like the sturdy version as they are much stiffer.

I don't see any problem at the speeds I run at. What speed and acceleration are you seeing resonance at and does it affect the print in any way?

---

[www.hydraraptor.blogspot.com]

This is exactly the situation I was wondering about...
My right Z-axis wobbles at the top - see video.
However, I dont think this affects print quality, so I ignore it.
Comments?

z-axis threaded bar movement on mendel90

I have learned that it make sense to reduce frequency loads on any machine.

Currently I didn't see an impact on print quality (and I did not write it). At the first time I saw the high deflection of the rods during the infill print of a screw-adapter (sliced with Cura, no change in firmware settings). After I had reduced the acceleration settings in the firmware (2000 and 1000 mm/s^2) it results in less deflection.

About M8 or M10 rod: first eigenfrequency (Z=0) appr. 60Hz (M8) and appr. 80Hz (M10).

@weirdgod: wobbling / resonance frequency of the thread rod are different things. At the moment I don't see an impact of wobbling on the print quality (no movement of z-axes during printing a layer).

Conclusion: No one needs to use the upper support but everyone can test

Did you try reading the RepRap magazine vol.1 There's an article on Z axis issues... there's even a formula yo can use to help eliminate banding.

Interesting article on the problems of the z-axis. Thanks.
Unfortunately I can not find any statements to a cantilever flexural resonator as my z-axis.

I'm probably the only one where the threaded rods vibrate strongly. Very strange....

Mine vibrate maybe 1mm with acceleration set to 2000 and rapid infill at 50mm/s. However, a plastic gnome standing on the X motor bracket for most of the day hasn't fallen off yet.

---

[www.hydraraptor.blogspot.com]

Quote

However, a plastic gnome standing on the X motor bracket for most of the day hasn't fallen off yet.

Excellent idea, this gnome! Demonstrates wobbling issues and step loss causes for those who get dizzy when reading "acceleration calibration".

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I had the same problem with my Z axis and what I did was to unscrew the rods at the base and place a 10 mm heatshrink tube on the rods, shrinked the tube and placed the rods back. This helped a lot and now I have much less vibration.

Just to concurr with nophead and others, adding constraint at the top of the threaded rod does not help anything.

The position of your X/Y axes MUST be solely determined by your smooth rods/linear bearings. The threaded rods serve only to push the X-Axis upwards, nothing else.

I also see no reason to eliminate resonance during Z moves, as far as I am aware no slicers make any X/Y movments during the motion of the Z axis, so there is no possible way in which motion of the Z-rods in the X/Y plane can induce motion of the nozzle in the same.

I have had a machine with M5 rod, badly bent during a car move rotating with over 10mm of runout during rotation and still producing beautiful prints at 0.15mm for a customer demo I did back in early 2012.