Posts made by o_lampe

@oliof I think, I try to use a microswitch, because a precise filament position isn't critical for me. I'll have to purge the nozzle after the tool change anyway.
It may even work without any switch, just by measuring the required retraction length.

@mrehorstdmd My understanding is, when using polar printer kinematics, RRF internally converts regular gcode (XY based moves) into theta/rho values.

Brushed metal always looks great, but wouldn't it kill the brush, when it moves sideways while pressed on the sheet?

Guess what?!
My Boss has seen your Arrakis videos and now I have to build a sand table first

I've generated a testfile with Sandify for a polar machine. But the file looks like classic cartesian coordinates with 0,0 at the bottom/left end instead of center.
Is RRF able to digest that on a polar bed?

<edit> according to the Wiki XY-offset is not supported yet
Is it possible to use a G1.1 macro instead which does the post processing on the fly?

;
; File name: 'PolarSandify'
; File type: gcode
;

G1 X117.776 Y111.451
G1 X87.768 Y122.172
G1 X67.855 Y95.324
G1 X88.664 Y67.022
G1 X123.451 Y78.957
G1 X122.412 Y117.191
G1 X84.209 Y127.838
G1 X62.033 Y93.102
G1 X89.794 Y60.376
G1 X130.841 Y77.746
G1 X126.383 Y123.577
G1 X79.901 Y133.128
G1 X56.414 Y90.128
G1 X91.762 Y53.782
G1 X138.485 Y77.431
G1 X129.625 Y130.544
G1 X74.879 Y137.970
G1 X51.070 Y86.420
G1 X94.566 Y47.321
G1 X146.295 Y78.037
G1 X132.081 Y138.024
G1 X69.184 Y142.294
G1 X46.072 Y82.003
G1 X98.194 Y41.072
G1 X154.184 Y79.584
G1 X133.701 Y145.943
G1 X62.864 Y146.034
G1 X41.488 Y76.908
G1 X102.629 Y35.115
G1 X162.063 Y82.081
G1 X134.440 Y154.223
G1 X55.972 Y149.130
G1 X37.386 Y71.172
G1 X107.846 Y29.527
G1 X169.842 Y85.534
G1 X134.258 Y162.782
G1 X48.566 Y151.522
G1 X33.828 Y64.837
G1 X113.815 Y24.386
G1 X177.429 Y89.936
G1 X133.125 Y171.536
G1 X40.709 Y153.160
G1 X30.874 Y57.950
G1 X120.498 Y19.763
G1 X184.735 Y95.276
G1 X131.016 Y180.395
G1 X32.470 Y153.995
G1 X28.579 Y50.565
G1 X127.851 Y15.730
G1 X191.671 Y101.533

@mrehorstdmd said in Spirograph emulator with Duet2:

What are you actually trying to do?

Good question.
If it's confirmed that Sandify's polar code is supported by RRF, I'd like to build a polar laser engraver or light grinder. (infinite rotation required and supported?)
It's definitely a step up and even easier to build than the SCARA-like dual arm, shown in my opening post.
I spent the whole Sunday afternoon watching your Arrakis playlist and then played with the polar version of Sandify. I created some interesting and doable patterns. Even some basic wiper pattern have their charm.
The fractal spirograph algorythm doesn't seem to produce the same patterns as the original, but much more complex and mesmerizing images.

@mrehorstdmd These pics are amazing! But they are the result of pushing sand around and back, over and over again.
I'm not sure if it would look as good with a dremel?
Maybe if I use a thick layer of wooden dust and later fix it with clear epoxy?

Nevertheless, it would require a known kinematic for RRF to reproduce those sandify patterns.
Right now I only have a mixing tool that moves three motors monotonously...
I think, I try wooden dust with steelball and magnet and see if it behaves sandy-like

Hi guys,
do you remember the VDE100 screw extruder?
It was meant as a DD-extruder on a smarteffector without fouling the rods.
Now the Bondtech INDX toolchanger is coming up and that inspired me to redesign the VDE100.
I added a pair of bevel gears which allows me to use a NEMA-08 motor without hollow shaft.
The gears are also the point where I split the extruder.

• Fixed to the linear rail: the motor, filament switch and the hotend
• a set of toolhousings each hosting a modified VDE100.

Here's a rough first teaser

A shout out to @tombrazier and @rqthree who did the major work on screw extruders especially the VDE100

@tas Quite the eye-opener! Thanks for posting.
I always thought that a crankshaft would create a sine wave. Now checking the above double-crank again, I'm sure the toolhead will draw an even weirder curve.

Maybe I should invest some time and try to simulate it?
Would OpenScad be able to do it, or do I have to learn Blender or the like?

@droftarts That was the vid that got me started. All the 3D printed parts must have taken a long time...
and I thought: bike-parts?! Another guy had the same idea a while ago. It's a good starting point but not CNC-able.
The version I posted above is a better candidate. It could even work as a FDM printer.
Like a short range 5-bar SCARA with a rotary bed. I wonder if RRF supports such weird kinematics?

@dwuk3d Almost forgot the thermistor....If you can, double up the lines for it, to reduce false readings from bad slip ring contacts.
If you already use an SSR, you know the possible downsides of bang-bang bed heating. I did that for a long time and it didn't bother me.

A more general concern is your project-philosophy, if I may critize you.
As a former project manager, I know it's important to set three things before you even start:

• define a goal (and stick with it)
• define a timeline (not so important here)
• define a budget

Especially ignoring the first one often leads to unfinished projects...just saying

At work I started building a spirograph machine based on bicycle parts similar to this:

I then began to work out the basic math behind the pattern, which I now regret not have done earlier.
Turns out, the outer toothcount should better be a binary number. That way only the small sprockets define the number of revolutions to complete a pattern.
My chain counts 124 teeth, which leads to 31 * n revolution patterns.
I'd grind a wooden surface to dust...especially at the center.

Good thing, the emulated CNC version can have any number of teeth. Plus I can vary the arm lengths and emulate non_circular sprockets.

If you're interested, there is this nice spirograph simulator. Just hit the demo button and get hypnotized

I wouldn't use a slip ring for a heated bed. At least I wouldn't trust the Chinese specs too much.
Maybe you can get away with it by using a 230V heater with a SSR? But that opens another rabbit hole.

@dwuk3d What puzzles me is: he prints without a heated bed and he's able to print at the center of a rotating bed.
And he doesn't even mention this, like it's a common thing.

I'd like to see some prints with single colour filament, just to see how clean the surface looks. IMHO, there are way to many artifacts, but they're obscured with pixie dust filament

@jay_s_uk Hmmm,

@dwuk3d said in Sovol SV08 Multiple Motion System Upgrade.:

Can't find the @CNCModeller post

Hmm, me neither. Maybe he was so succesful with it, that he went closed source ?

I only found this early photo of polar printer

Parallel Vase mode would be cool, but when I made things in vase mode I already had to slow down a lot, because of the minimum layer time/part cooling.
With multiple heads you'd also have to sort out a precise method to sync layer times, unless the part is super symmetric. But with such a part, you can just split it in vertical halfes (or quarters) and use mirror mode on a stationary bed.

I'd need a small and reliable filament switch to mount it under a screw extruder (VDE100) and thought of an optical switch or the IR probe.
Has anyone done that and wants to share the experience?

@dwuk3d As long as you don't go the rotating bed route, I'm sure you'll be succesful.
It has been proven (by @CNCmodeller ) that the accelerations required to print near the center of a rotating bed are impossible to achieve. The center itself is a No-Go zone.

Awesome print video
I'd give it a like, but I don't watch YT while being logged in.
I've noticed, the red tool was pretty impatient priming and wiping the nozzle, while blue was still printing. I guess that's a matter of postprocessing.
Changing from red to white or vice versa is a different act, because they move on the same y-beam and the unused tool is up in the air.
Maybe you can lower it for wiping and raise it again until it's on duty?

@dwuk3d said in Sovol SV08 Multiple Motion System Upgrade.:

My current plan for the next phase is occasionally moving bed in the Y direction

It's obvious how much print area you loose with the blue tool looking in the same direction...but for tool change-reasons you have to look for Plan B
You might not even need a bigger heater for the bed, since the print area is still the same.
PS: If you take the bed-slinger route anyway, why not reconstruct the gantry to move only in X-direction and let the bed do Y. Just like many Prusa I3 clones still do.
It'll save a ton of motors/belts/idlers on the gantry

Forget that, you want to print with 2 heads simultaneously later on, which is impossible with a bed slinger...

@droftarts Your input just started a long train of thought, but it's not related to this thread.
[OT]
The almost forgotten screw-extruder was build as DD extruder (hollow shaft a.s.o.) but with a bit of modification it could become a super compact tool changer, similar to the INDX. Plus it wouldn't need hollow shaft motors anymore.(no filament softening issues! )

@dc42 said in Spirograph emulator with Duet2:

the final pen position not quite lining up with the initial position

Depending on how good the build will be, it will look wobbly anyway. Especially the joints and wrist bearings are giving me an headache. I want to use a Dremel grinder at least...
I thought of replacing the circular motion of the two arms with ballscrews going back and forth.
IMHO, the pattern is mostly related to different starting positions ( or different start-angles in the original toy-mechanic)

But how could I mimic a sinusodial motion with a linear axis?
Do we have "s-curve" acceleration?

@Notepad In my theory, 20 teeth is the worst, since if syncs "perfectly" with the 200 steps/turn torque curve.
21 isn't a prime number, so I'd go for 19 or 23