"Beat Frequency" Horizontal Ridges caused by Z-Axiz

I know horizontal ridges are an artifact of FFF printing, and the ridges are generally smaller with thinner layers. I’m using Velleman Repetier Host “new” default settings, so my layers are 0.025mm.

However, I have larger, exaggerated ridges for every four to five layers. So if I print an object with 180 layers, I will have about 40 “exaggerated” ridges. It’s almost as though the vertical surface is a sine wave instead of a plane. These ridges are very regular, but there is only one “exaggerated” ridge for every 4.5 layers. It happens in objects such as cubes and the fan nozzle.

If I had a horizontally moveable Z-Axis I would look for wobble in Z. But the 3-drag Z-Axis only moves vertically. So…I speculate that perhaps there is a periodic “wobble” in the thickness of the layers caused by a “regularly irregular” vertical movement of the Z-axis. It seems certain to me that this must be a Z-Axis problem.

Did I state the problem in an understandable way? A picture is worth a thousand words, but I don’t know how to attach a picture. If any one understands the problem I’m having, are there any suggestions? I’ve seen it stated that the Z-Axiz is the weak link in this 3-Drag K8200 design. One speculation I have is that there is “backlash” in layers for which the filament was retracted back into the extruder. Has anyone figured an easy way to take the backlash out of the loose fitting Z-Axis nut that rides along the threaded rod?

Any suggestions at all?

Art

OK, I just realized that the frequency of the “ridges” on my vertical surfaces is the same as the pitch of the Z-Axis screw.

However, I am a loss to understand the mechanism for the imperfections. How can Z-Axis wobble translate into different layer thicknesses?

Unless I understand the mechanism, I cannot attempt a fix. The fix is obviously a hardware fix to the Z Axis. In fact, the entire metric screw/nut design is bogus. It should be a machined screw and a precision bearing. I cannot possibly be the only user experiencing this problem - especially considering the care with which I built and triple checked every step during assembly.
For instance, I not only triple checked all wiring, but I buzzed it out with an ohm meter before application of power. When I finally tested, there were zero wiring errors. Just a flakey mini USB port connector which required me to purchase a second board.

Sorry to read about the USB connector problem.
There was no need to purchase a second board though. You could have returned it for repair/exchange under warranty.
Please check the Z-coupler, is it 100% concentric? If not, please let us know, we’d be happy to replace it for you.
Soon, we will release an upgrade package with flexible coupler, machined trapezoid screw and nut, for those who wish to upgrade their printer.

No, my Z-Axis connector is not concentric. I noticed that during assembly.

I would purchase the Z axis upgrade immediately if it were available. Any idea how long before it finds its way into the distributor chain? I’m only guessing, but I suppose it might require a firmware change also.

Finding parts here in the USA is hit and miss. Velleman USA will sell only through distributors, and it is not always easy to locate parts from distributors even on the Internet. The parts I purchased through distributors were drop-shipped by Vellerman from Dallas Texas, do the distributors did not stock them anyway. It is frustrating to know there are parts in Texas that I cannot directly order.

I purchased a second board because I could locate only one board, and I didn’t want to risk being without a controller board. My power supply overheats, and I’m cooling it until I locate another for sale.

Returning things via overseas shipment is unattractive because of the long waits. I’m a new owner and don’t want to put my printer on the shelf for weeks while I await a part. I guess I have more money than brains. After spending nearly $1000 US so far, spending another $50 for a new power supply seems justified to protect my investment.

Once I’m sure everything is working, I will return all defective parts in a single shipment while using the spare parts I purchased to stay in operation. I plan to call the Vellerman USA phone number for return instructions, although I suspect all parts must go back to Vellerman Europe to verify they are really defective.

Regards,
Art Cominio
Melbourne, Florida, USA

We plan to have the upgrade axis available in the next couple of weeks.
The Velleman support office in TX is able to handle repairs and they can supply spare parts, there is no need to ship to Europe.
Please contact them for all repair/spare part issues, they are very helpful.

Thank You for the information!

Regards,
Art Cominio

Hi Art,
A search of this forum will reveal many posts on the subject of z axis ribbing or banding - I have been battling it ever since I got my K8200. I even suspected Slicer, but eventually got to understand that threaded rod lead screws are just not accurate enough for this application. I am glad to see that Velleman are bringing out a fix.

Placing a dial indicator on the bed and the probe touching the underside of the extruder beam, I then sent a manual command Z=+0,1mm and read the dial gauge. This was repeated about 40 times and the results logged in a spreadsheet and a chart drawn. It was very clear that there is a cyclical variation with the pitch of the thread (1,25mm), going from 0,087mm to 1,02mm! In addition, the average over all the steps was about 0,975mm. A number of threaded rods from different manufactures showed very similar results, even one I made by cutting a thread on some ground bar.

The explanation is that this causes a variation in layer height, which results in an inversely proportionate variation in the extrusion width (as the volume extruded is constant), which shows up as the ribbing or banding.

I ordered a 10mm dia x 2 pitch leadscrew and nut from RS Online (their part no 2637200) about 6 weeks ago, but they are not able to deliver for another 4 weeks ( I am in Johannesburg, South Africa), so I am going to look for another supplier.

Good luck,

Wally

Wally,

Thanks for the information. What you say makes perfectly good sense to me. I understand that variations in layer height affect the layer width. This makes a lot more sense than the idea that “wobble” is causing the problem. The defect is built into the machine - that is - it is limitation of the threaded rod used for the Z-Axis. A machined leadscrew will solve the problem.

I have one question. If the pitch of the leadscrew is different from the pitch of the threaded rod, then firmware changes are required, right?

Regards,
Art Cominio Melbourne Florida USA

Hi Art,

To change the steps/mm, open Repetier, click on the tab and then in the G-Code box. Enter and hit the button. This updates the firmware. 2560 steps/mm is the standard setting that the K8200 is delivered with; I changed mine to 2567 to allow for the pitch of the leadscrew being under spec.

There is a good write up on this at reprap.org/wiki/Triffid_Hunter’s … tion_Guide

I also noticed while trying to adjust the Extrusion multiplier in Slic3r that the lower layers showed under-extrusion, while higher layers (say 200 & up) showed over-extrusion. I put this down to leadscrew pitch error and now run Z at 2570 steps/mm.

Good luck,
Wally