I finished my C-Beam kit from Ooznest some days ago and I thought you might be interested in some of the „issues“ during the build
First of all, it has been a big pleasure to build this machine. When you´re designing things yourself most of the time there is a point where you just value a well thought out kit that just fits, as expected from a aluminum profile kit. 😛
It took roughly two afternoons to get everything together, followed up by some additonal hours of choosing the right software and adjusting the settings. Hendrik found a nice little programm called cnc.js. I had a look at the standard solutions openbuilds suggested and chllipeppr but they didn´t seemed quite right. It worked straight away and despite I haven´t tested it yet with a real cutting job I can really recommend it.
As I´m using the Kress from my MPCNC for testing purposes on the C-Beam I had to do some minor modifications to the tool clamp. I had to get it lower than the intended mounting point and all I had to do was grinding the screws a little down on the edges.
That way It was possible to screw the clamp from the backside of the plate securely. Not as good as the corners that are intended for this but it should work. At least two of them are securing the clamp from above.
What I didn´t liked so much and I suppose it´s not a common thing with the kit is the quality of the supplied cables. They were crimped on the isolation and that meant they didn´t had any reliable contact to the PSU. After three seconds the PSU switched off. My first thought was that the PSU was faulty and the second was that I might had connected something wrong. As I repeatedly check connections before I power something up I usually suspect the error elsewhere. 🙂
The bare metal shoudln´t stick out at the front and crimping the isolation leads to missing pressure on the cable itself causing a loose connection. I corrected this to get it stable and safe and everthing works fine now.
Those were the only two points I had to modify. Everthing else went smoothly. Pictures from the final machine will follow soon 🙂
I had the demand for some more serious aluminum milling. I did that with the MPCNC before and it was kind of okay but I wouldn´t like to mill higher part counts with it. The spindle gets hot after some time and leads to more flex in the toolhead. It´s not made for aluminum so I looked around for a more suitable platform.
I found the openbuilds C-beam machine to be nicely designed and versatile enough to be worth adding to my machinepark. I looked around for a europe based shop and found openbuilds poland , v slot europe and ooznest .
I decided to go with the ooznest version as they had the best full kit. I didn´t wanted to source all the parts from here and there. Sometimes it´s nice to have a complete package 🙂
The sipment was fast after the payment went trough and so I had a first look at the package today. All was nicely packed and labeled according to the build section like X, Y and Z Axis. The printed parts for the power supply and the controller board looked pretty nice and felt solid.
Can´t say much yet but a real manual would be more comfortable than a build video. It´s way faster to look at a drawing than to forward, play, stop, rewind, play, the buildvideo 😀
Surprise! I had a common chip clearance problem. I considered this as a cause from the beginning and also got some tips from the vicious1..com forum. I milled two complete sets of 3D-printer frames without problems so I couldn´t believe that this was really the cause.
I drew a test pattern and managed to get it right with an additional ,5mm helical drilling movement. That seems enough for the chips to clear the hole.
I don´t have that much time to get into those issues but the parts didn´t came out perfectly rectangular yet. I tried to manually adjust the stepper position before powering it up but thats still not enough to get the precision I need. I´m using all of the available travel so I´m getting quite a big error in the end.
I need to readjust the frame and the feet as well.
I currently got some problems with drilling holes again. It seems there is something odd in the Z- axis. I didn´t assembled it carefully enough. I suppose that the spindle isn´t lowered in a straight line, instead it is tilted on the way down pushing the endmill into the material sideways and ripping a hole into my plywood.
I´m using this as a reason to reprint the tool- and steppermount.
When I originally build the MPCNC I had no drillpress and the terrible idea to tap a M4 thread into the stainless. This is a bad idea as it clearly makes the alignment of the screws in the tool mount to the conduit almost impossible.
I redrilled the holes and I am now using the original path with the nut traps inside the conduit. Along with the adjustment of the leadscrew this hopefully improves the precision and rigidity of my z-axis.
Finished toolmount replacement. The one on the left was printed with an pre R17 BCN3D Sigma and the right one printed with my DIY printer.
Now everything is assembled again. The tubes are now „perfectly“ straight and parallel. I used a glass plate to verify the alignment.
I did a short test run and I had the same result as before. A nice circular vibration and a not very pretty drill hole. I´m using a diamond cut endmill at the moment and went successfully through two sets of printer frames. I´ve been told in ryans formum (viscious1.com) that diamond cut bits are everything but ideal to dive 12mm into beech plywood. They are very robust so they were the only ones that survived when it came to an error. They are compressing the shavings within the hole and that leads to more and more pressure hence the heat rises and the endmill escapes in the circular movement.
In the past I ordered a endmill from sorotec which broke after some drilling operations. That left me confused what the error might be. I also tried a shorter endmill originally for aluminum which transported the shavings quiete nicely but was too short.
I ordered new endmills again and hopefully all the possible errors caused by the machine are gone and I get my drill patterns and contours out of one endmill.
I founded a little company called raise-UAV. The main advantage is the increased time for my projects and that was totally necessary when I look at my roadmap. 🙂
It is primarily focused on the development of open drone systems manufactured with 3D printing technology.
The printer and the mini ghost I´ve started here will be hosted on the company side to keep things easier. Other things not drone and printer related will be posted here on the kleingeist blog in detail. Next bigger projects are a printed and easy to build kajak and an „earth rover“ for research purposes.
There will also be a shop on raise-uav.com where you can buy kits and ready made articles to keep me funded for further development.
I also began a video series picturing the builds and other useful content with its first episode right here
Has been tested intensively now. Works ok, with a little practice. I guess I can´t expect more from a printed design due to the weight shift with 500 g class DSLMs. Nevertheless thats enough to do some close slides for products etc.
See it slide here:
There might be some documentation necessary before releasing it on thingiverse as the mechanism needs to be adjusted correctly.
Inspired by one of my readers I had a look at the dust shoe for the Kress again. I redesigned it completely and got rid of the additional tubes and clamps. I thought that it would be a good idea to have a highly flexible part between vaccum and CNC. If I had ordered the 20mm silicon tubes it may have worked as well but reducing parts is allways priority.
So in the end we got a pretty classic dust shoe here. It doesn´t have a brush at the bottom as it seems to work pretty good at the moment and I really like to see the endmill. However I included three holes at the bottom for a later brush design.
It works pretty good with lighter materials but GFK, aluminum or similar are not fully sucked up.
It has been printed in 0,3mm draft quality for testing purposes.
I had the opportunity to try out at least 5 different printers. All of them had sifgnificant backdraws like missing heatbed, price and noise. They all did there job anyway and my Fabrikator is working pretty much „fire and forget“ all the time. Nevertheless I´m tempted to create my own printer just for the sake of adding it to my collection of valuable experiences.
I came up with T-Bot first iteration closely related to prusa´s I3 framewise. I didn´t liked the Z leadscrews standing out that much and too far away from the linear slides so the second version focused on making this a little more aerodynamic.
I am aiming for a precise as possible head and bed positioning mechanics with dual extrusion the bowden way. I have good experience with bowden so I don´t see any significant disadvantages at the moment.
The second version with the new z and x mechanics.
There are E3D V6 mounted on the gantry at the moment but it should be stable enough for light milling operations.
The frame is designed with 16mm MDF in mind but I probably change to a different material (not OSB). Thats why I have to redesign most of the frame parts as the new material is somewhat thinner.
I already have most of the electronic components in mind. The first prototype might be powered by components I have laying around but the finished version is meant to be state of the art 🙂
Watching Thomas Sanladerer´s channel really helped me out making decisions at some points in the design process so thank you for your effort in making great videos, if you ever read this.
I took another iteration to make it move like it should. As you can see, in comparison to the previous entry, the case and the mechanics have been modified. The belt system used is a bit more complicated than I initially thought and I couldn´t get it to perform „super smooth“ like its contestants.It is best suited for actioncams as you can see in the picture below. I tried a DSLM and it works somehow but it flexes under the load generating a somewhat unstraight path.
I think it´s okay anyway, taking the money spend into account versus performance. At least it was a good exercise. In the video below are some cam transitions made with it so get the idea of the achievable smoothnes. It´s more like basecap than edelkrone 😛
This project is currently in long term testing and refinement stage.
Inspired and a bit fascinated by the glide arm style camera sliders I decided to build my own version of it.
The parts are all standard you got laying around either from a MPCNC build or if you like modifying or building 3D printers. Anythin else can be sourced from your local hardware store or other sources.
It is not working yet as I´m wating on some hardware parts that are announced for arrival next week. After that it will be tested in various conditions and then maybe released into the wild 🙂