banner perso

Some links about me. Many of my 3D designs are free. I also post on Google+ and in another blog, oz4.us

Monday, March 30, 2015

CNC fail? 3D printed design blows my ears instead of the dust

I bought a well-built Chinese CNC machine last year, namely a 4 axis CNC3040Z-DQ router/engraver, for about €1000 (as far as I remember).

Milling a PCB with a low cost CNC router/engraver controlled by LinuxCNC.
My initial need was to "etch" electronics PCB boards at home, to get rid of so called proto- and perf boards. This machine is overkill for the job, but I do not regret my purchase since it can do much more than PCB milling...

But only recently did I spend enough time to use it. Even though the reliability and consistency of a CNC milling machine is way better than that of a 3D printer, the whole process is cumbersome. It takes a lot of time to get used to the software, which is a decade behind the ergonomics of more intuitive 3D printing software. Worst, the controlling software like linuxCNC or Mach3 still require a parallel port and a special real-time distribution of linux. Both are a real pain because you need a dedicated and obsolete machine just to drive the CNC. Recent projects like GRBL are knocking at the door though, as an alternative to parallel port hardware: they rely on Arduinos, but are limited by the power of this platform. ARM-based boards are more promising, especially for fast machines, or when you have more than 3 axis of freedom (e.g. see 6DoF)

Note: I may write more about CNC machining at home, its software and process in other posts. But I start here with a quick, funny, and miserable failure of mine, that was meant to be an improvement in the first place. And, well, it mixes printing and milling so it is a good transition.




And who thought a 3D printer was dirty?


A 3D printer is tremendously useful
for such a design. But it does not mean
the design is functional!
As a subtractive manufacturing, a CNC milling machine is way dirtier than a 3D printer. The latter only generates lots of strings and bits of plastic, but they are well contained in the bottom of semi-enclosed printers like the Ultimaker Original (the second version homes to the bottom, which is annoying in this regard imho!).

Reciprocally, CNCs generate an outstanding amount of dust, composed of very fine particles that easily fly and float in the air. Contrary to 3D printers, no way you can clean up after a job, it must be done during the milling process. The high speed spinning end mill throws the dust all around contrary to a mostly static 3D printer nozzle.

Moreover, raw PCB boards are often made of toxic epoxy: I do not want to inhale that dust at all. So even though a vacuum cleaner is the minimum you want to use, it still may not block the fine dust and may only throw them even farther in the room (in this regard, I was told about HEPA-grade filters).



The unsafe, boring and tiring way (hence triply unsafe!).
Keep your fingers away and never ever
touch the drill bit while it rotates!
Anyhow, it is both tiring, cumbersome and dangerous to stay and hold a hose close to the milling bit. The 20.000 RPM spinning bit would probably shatter at the very moment something unexpected touches it. The end bit may be small, but its kinetic energy is huge, and without eyes I would not be able to hack anymore...

So I soon started to think about a design to hold the hose for me, so I could step back from the job (and, well, work on something less boring, like staring at a 3D printer -- additive processes are way more interesting imho!)






A milled CNC extraction head, held by magnets (video),
and what looks like a hollowed broom below!
(this is a really smart design by Andy Shrimpton)

Of course, many designs already exists to hold a vacuum hose close to the end mill.

Some of them even are made with the CNC, like the smart one on the right (milled from polycarbonate).

But there are also many designs that are 3D printed, for more "complex" shapes (complexity is no goal of course, but a 3D printer allows "non-flat" shapes at least), such as duck-shaped adapters for existing vacuum hoses which are bolted on the CNC head.




3D printing a "too smart" end, to suck up the dust of the CNC end mill ?


I just find it terribly hard for me to recycle ideas and designs stuff when I think I can do it myself, especially when I think I can improve on existing devices.

A swirling vacuum head with internal
angled fins. Owning both a 3D printer
and a CNC is such a killer combo!
So after a few hours of designing in Openscad, I came to what I though would be a good idea: a "swirling sucking end". My goal was eventually to create a swirling air movement all around the mill bit, in order to suck the dust up and into the vacuum cleaner. The printed adapter would screw on the CNC spindle and stay somehow far from the milled material (the drill bit can plunge into it for about 3 cm, no way I could extend the adapter downwards -- and the chuck must remain accessible).

This is why the design on the left features some angled fins around the CNC chuck, that would eventually help to create a vortex down to the milled board and around the end mill...



A word of caution on using 3D printed parts for a CNC (again)

Making anything that will sit close to a CNC must be done carefully again.

Once again, for safety, you certainly do not want a 3D printed part to crack or break and it hits the end mill in the middle of a job. You would get the whole thrown in the air randomly with the metallic bit. This is why I chose the material carefully for the job (here I used black PET (Colorfab XT) as a tough material, but I could probably better switch to Nylon). Carbon-fiber reinforced filament are no option, as is way too stiff, hence it breaks without prior deformation.

Specifically, I do not trust 3D printed thermoplastics as much as milled materials. Even milled thick thermoplastic sheets would give a stronger object because the material is more homogeneous than with layers that may have bonding issues.


So? Did it work?

Alas, not only this sucker is a pure failure as a vortex-inducer but all the powerful vacuum energy seemed to be converted into an unbearable resonant noise!

This thing just sucks nothing but it blows my ears. This damned design probably exceeds 120dB, and the weaved resonant frequencies as it stops are really unbearable!

Here is it (video): a better vacuum sucking head? Not at all!
Warning: extremely noisy, check your audio volume first!
This thing just sucks nothing, but it blew my ears.

Conclusion: failure gives experience...

It teaches me five good lessons:

  1. (even painful) fun is always hidden behind where you do not expect it
  2. Field testing is the only way to check that a design works!
  3. Reciprocally I should learn more about fluid mechanics theory
  4. KISS (keep it simple, stupid): now I will retry without the small angled fans
  5. Always better protect myself, not only for the eye, but now also for the ear!
Update:: ...and a 6th lesson: never draw conclusions too early! The major source of noise actually came from the blue tube in relation to the adapter to the larger vacuum tube. Plugging the larger vacuum hose directly makes it much quieter (but still inefficient).


So now... I will retire deep in a cave for a few hours, to design something else while I heal my ears.

I deeply think that failure is a good opportunity
(discover! as in this nice post about hackers/makers)
(illustration by Carol Dweck and Nigel Holmes)

About Me

My photo

If you know me and you cannot tell exactly what my real job is, then you probably found the right Jeremie. Check zax.fr for some pointers.

I am self-employed and I help start-ups, research centers, small companies with their needs related to computers, sensors, data processing and mechatronics. If you have a project and know what "R&D" is, then you already sparked my interest ;)