The Smell of Molten Projects in the Morning

Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.

Author: Ed

  • MPCNC: Spirograph Generator with Tool Changes

    An improved version of my GCMC Spirograph pattern generator, now with better annotation and tool changes:

    Spirograph pattern - overview
    Spirograph pattern – overview

    The GCMC code sets the stator and rotor gear tooth counts, the rotor diameter, and the pen offset using a pseudo-random number generator. This requires randomizing the PRNG seed, which I do in the calling script with the nanosecond of the current second: rnd=$(date +%N).

    The G-Code file name also comes from the timestamp:

    ts=$(date +%Y%m%d-%H%M%S)
    fn='Spiro_'${ts}'.ngc'
    # blank line to make the underscore visible
    

    Which means you must call the Bash script slowly to generate a pile o’ plots:

    for i in {1..60} ; do sh /mnt/bulkdata/Project\ Files/Mostly\ Printed\ CNC/Patterns/spiro.sh ; sleep 1 ; done
    

    Sift through the heap with drag-n-drop action using an online G-Code previewer. There seems no clean way to convert G-Code to a bitmap on the command line, although you can do it manually, of course.

    The GCMC program spits out the G-code for one plot at a time, so the Bash script calls it four times to fill a sheet of paper with random patterns:

    for p in $(seq 4)
    do
      rnd=$(date +%N)
      gcmc -D Pen=$p -D $Paper -D PRNG_Seed=$rnd $Flags $LibPath -q "$Spirograph" >> $fn
    done
    

    The -q parameter tells GCMC to not include the prolog and epilog files, because the calling script glues those onto the lump of G-Code for all four plots.

    The -D Pen=$p parameter tells the GCMC program which “tool” to select with a Tn M6 tool change command before starting the plot. Although plotter pens have a well-defined position in the holder and a pretty nearly constant length, you must have a tool length probe installed and configured:

    MPCNC Tool Length Probe - Plotter Pen
    MPCNC Tool Length Probe – Plotter Pen

    Set the overall sheet size in inches or millimeters to get a plot centered in the middle of the page with half-inch margins all around:

    Paper='PaperSize=[16.5in,14in]
    

    With all that in hand, those good old black ceramic-tip pens give impeccable results:

    Spirograph pattern - black ceramic pen - detail
    Spirograph pattern – black ceramic pen – detail

    The surviving ones, anyhow. I must apply my collection of Sakura Micron pens to this task.

    The other three colors come from fiber pens with reasonably good tips:

    Spirograph pattern - central details
    Spirograph pattern – central details

    They’re a lot like diatoms: all different and all alike.

    The GCMC and Bash source code as a GitHub Gist:

    # Spirograph G-Code Generator
    # Ed Nisley KE4ZNU – December 2017
    Paper='PaperSize=[16.5in,14in]'
    Flags="-P 2"
    LibPath="-I /opt/gcmc/library"
    Spirograph='/mnt/bulkdata/Project Files/Mostly Printed CNC/Patterns/Spirograph.gcmc'
    Prolog="/home/ed/.config/gcmc/prolog.gcmc"
    Epilog="/home/ed/.config/gcmc/epilog.gcmc"
    ts=$(date +%Y%m%d-%H%M%S)
    fn='Spiro_'${ts}'.ngc'
    echo Output: $fn
    rm -f $fn
    echo "(File: "$fn")" > $fn
    cat $Prolog >> $fn
    for p in $(seq 4)
    do
    rnd=$(date +%N)
    gcmc -D Pen=$p -D $Paper -D PRNG_Seed=$rnd $Flags $LibPath -q "$Spirograph" >> $fn
    done
    cat $Epilog >> $fn
    view raw spiro.sh hosted with ❤ by GitHub
    // Spirograph simulator for MPCNC used as plotter
    // Ed Nisley KE4ZNU – 2017-12-23
    // Spirograph equations:
    // https://en.wikipedia.org/wiki/Spirograph
    // Loosely based on GCMC cycloids.gcmc demo:
    // https://gitlab.com/gcmc/gcmc/tree/master/example/cycloids.gcmc
    // Required command line parameters:
    // -D Pen=n pen selection for tool change and legend position
    // -D PaperSize=[x,y] overall sheet size: [17in,11in]
    // -D PRNG_Seed=i non-zero random number seed
    include("tracepath.inc.gcmc");
    include("engrave.inc.gcmc");
    //—–
    // Greatest Common Divisor
    // https://en.wikipedia.org/wiki/Greatest_common_divisor#Using_Euclid's_algorithm
    // Inputs = integers without units
    function gcd(a,b) {
    local d=0;
    if (!isnone(a) || isfloat(a) || !isnone(b) || isfloat(b)) {
    warning("Values must be dimensionless integers");
    }
    while (!((a | b) & 1)) { // remove and tally common factors of two
    a >>= 1;
    b >>= 1;
    d++;
    }
    while (a != b) {
    if (!(a & 1)) {a >>= 1;} // discard non-common factor of 2
    elif (!(b & 1)) {b >>= 1;} // … likewise
    elif (a > b) {a = (a – b) >> 1;} // gcd(a,b) also divides a-b
    else {b = (b – a) >> 1;} // … likewise
    }
    local GCD = a*(1 << d); // form gcd
    // message("GCD: ",GCD);
    return GCD;
    }
    //—–
    // Max and min functions
    function max(x,y) {
    return (x > y) ? x : y;
    }
    function min(x,y) {
    return (x < y) ? x : y;
    }
    //—–
    // Pseudo-random number generator
    // Based on xorshift:
    // https://en.wikipedia.org/wiki/Xorshift
    // http://www.jstatsoft.org/v08/i14/paper
    // Requires initial state from calling script
    // -D "PRNG_Seed=$(date +%N)"
    function XORshift() {
    local x = PRNG_State;
    x ^= x << 13;
    x ^= x >> 17;
    x ^= x << 5;
    PRNG_State = x;
    return x;
    }
    //—–
    // Spirograph tooth counts mooched from:
    // http://nathanfriend.io/inspirograph/
    // Stators has both inside and outside counts, because we're not fussy
    Stators = [96, 105, 144, 150];
    Rotors = [24, 30, 32, 36, 40, 45, 48, 50, 52, 56, 60, 63, 64, 72, 75, 80, 84];
    //—–
    // Start drawing things
    // Set these variables from command line
    comment("PRNG seed: ",PRNG_Seed);
    PRNG_State = PRNG_Seed;
    // Define some useful constants
    AngleStep = 2.0deg;
    Margins = [0.5in,0.5in] * 2;
    comment("Paper size: ",PaperSize);
    PlotSize = PaperSize – Margins;
    comment("PlotSize: ",PlotSize);
    //—–
    // Set up gearing
    s = (XORshift() & 0xffff) % count(Stators);
    StatorTeeth = Stators[s];
    comment("Stator ",s,": ",StatorTeeth);
    r = (XORshift() & 0xffff) % count(Rotors);
    RotorTeeth = Rotors[r];
    comment("Rotor ",r,": ",RotorTeeth);
    GCD = gcd(StatorTeeth,RotorTeeth); // reduce teeth to ratio of least integers
    comment("GCD: ",GCD);
    StatorN = StatorTeeth / GCD;
    RotorM = RotorTeeth / GCD;
    L = to_float((XORshift() & 0x3ff) – 512) / 100.0; // normalized pen offset in rotor
    comment("Offset: ", L);
    sgn = sign((XORshift() & 0xff) – 128);
    K = sgn*to_float(RotorM) / to_float(StatorN); // normalized rotor dia
    comment("Dia ratio: ",K);
    Lobes = StatorN; // having removed all common factors
    Turns = RotorM;
    comment("Lobes: ", Lobes);
    comment("Turns: ", Turns);
    //—–
    // Crank out a list of points in normalized coordinates
    Path = {};
    Xmax = 0.0;
    Xmin = 0.0;
    Ymax = 0.0;
    Ymin = 0.0;
    for (a=0.0deg ; a <= Turns*360deg ; a += AngleStep) {
    x = (1 – K)*cos(a) + L*K*cos(a*(1 – K)/K);
    if (x > Xmax) {Xmax = x;}
    elif (x < Xmin) {Xmin = x;}
    y = (1 – K)*sin(a) – L*K*sin(a*(1 – K)/K);
    if (y > Ymax) {Ymax = y;}
    elif (y < Ymin) {Ymin = y;}
    Path += {[x,y]};
    }
    //message("Max X: ", Xmax, " Y: ", Ymax);
    //message("Min X: ", Xmin, " Y: ", Ymin); // min will always be negative
    Xmax = max(Xmax,-Xmin); // odd lobes can cause min != max
    Ymax = max(Ymax,-Ymin); // … need really truly absolute maximum
    //—–
    // Scale points to actual plot size
    PlotScale = [PlotSize.x / (2*Xmax), PlotSize.y / (2*Ymax)];
    comment("Plot scale: ", PlotScale);
    Points = scale(Path,PlotScale); // fill page, origin at center
    //—–
    // Set up pen
    if (Pen > 0) {
    comment("Tool change: ",Pen);
    toolchange(Pen);
    }
    //—–
    // Plot the curve
    feedrate(3000.0mm);
    safe_z = 1.0mm;
    plot_z = -1.0mm;
    tracepath(Points, plot_z);
    //—–
    // Put legend in proper location
    feedrate(500mm);
    TextSize = [3.0mm,3.0mm];
    TextLeading = 1.5; // line spacing as multiple of nominal text height
    MaxPen = 4;
    line1 = typeset("Seed: " + PRNG_Seed + " Stator: " + StatorTeeth + " Rotor: " + RotorTeeth,FONT_HSANS_1);
    line2 = typeset("Offset: " + L + " GCD: " + GCD + " Lobes: " + Lobes + " Turns: " + Turns,FONT_HSANS_1);
    maxlength = TextSize.x * max(line1[-1].x,line2[-1].x);
    textpath = line1 + (line2 – [-, TextLeading, -]); // undef – n -> undef to preserve coordinates
    if (Pen == 1 || Pen > MaxPen ) { // catch and fix obviously bogus pen selections
    textorg = [PlotSize.x/2 – maxlength,-(PlotSize.y/2 – TextLeading*TextSize.y)];
    }
    elif (Pen == 2) {
    textorg = [-PlotSize.x/2,-(PlotSize.y/2 – TextLeading*TextSize.y)];
    }
    elif (Pen == 3) {
    textorg = [PlotSize.x/2 – maxlength, PlotSize.y/2 – TextSize.y];
    }
    elif (Pen == 4) {
    textorg = [-PlotSize.x/2, PlotSize.y/2 – TextSize.y];
    }
    else {
    Pen = 0; // squelch truly bogus pens
    textorg = [0mm,0mm]; // just to define it
    }
    if (Pen) { // Pen = 0 suppresses legend
    placepath = scale(textpath,TextSize) + textorg;
    comment("Legend begins");
    engrave(placepath,safe_z,plot_z);
    }
    if (Pen == 1) { // add attribution along right margin
    attrpath = typeset("Ed Nisley – KE4ZNU – softsolder.com",FONT_HSANS_1);
    attrpath = rotate_xy(attrpath,90deg);
    attrorg = [PlotSize.x/2,5*TextLeading*TextSize.y – PlotSize.y/2];
    placepath = scale(attrpath,TextSize) + attrorg;
    comment("Attribution begins");
    engrave(placepath,safe_z,plot_z);
    }
    goto([-,-,25.0mm]);
    view raw Spirograph.gcmc hosted with ❤ by GitHub
  • A Spirograph for Christmas

    Gotta play with my new toy:

    Spirograph - liquid ink - ceramic tip
    Spirograph – liquid ink – ceramic tip

    That’s with a set of liquid ink and ceramic tip plotter pens. They’re unbelievably cranky, but produce wonderfully fine lines:

    Spirograph - liquid ink pen - detail
    Spirograph – liquid ink pen – detail

    Text comes out exactly the way vector lettering should look:

    Spirograph - liquid ink pen text - detail
    Spirograph – liquid ink pen text – detail

    There’s a slight shake visible at 500 mm/min = 8.3 mm/s, but it’s Good Enough.

    All the pen-and-ink traffic around the center produced a ring of damp green fuzz:

    Spirograph - liquid ink - ceramic tip - center detail
    Spirograph – liquid ink – ceramic tip – center detail

    The artsy part of the plot ran at 1800 mm/min = 30 mm/s, with little of the wobbulation at 6000 mm/min = 100 mm/s. None of that would matter with a router, of course.

    It’s a nice, Christmasy design in kinda-red and sorta-green.

    From the stack of plots accumulating near the MPCNC bench:

    This slideshow requires JavaScript.

    Plots 7 and 9 show the tape sutures required to produce a 26×18 inch sheet covering the MPCNC’s full work area. The squat plots fit on B-size sheets and the rest come from 17×14 inch artist’s sketchpad sheets.

    I used Google PhotoScan to capture and rectangularize paper sheets from the floor or atop the bench, then battered the contrast and crushed the file size with a one-liner:

    i=1 ; for f in 1* ; do printf -v dn "Spiro %02d.jpg" $(( i++ )) ; convert $f -level '10,80%' -density 300 -define jpeg:extent=300KB tweaked/"$dn" ; done
    

    The plots look great in person (modulo some incremental software improvements), but the slideshow images look awful because:

    • Google PhotoScan produces surprisingly low-res images
    • I’m overly compressing the results

    They’re not (yet) art and there’s no point in a high-quality workflow.

    Enjoy the day …

  • Wandering Washroom Water

    You’d think architects would know better:

    Fancy washroom counter
    Fancy washroom counter

    The washroom has in-the-wall towel + trash stations at each end of the counter, but they’re obviously inadequate for the purpose. Fortunately, the counters slope away from the attractive stacks of paper towels.

  • Critters on the Patio

    A light snowfall revealed plenty of overnight traffic on the patio:

    Small animal tracks in the snow
    Small animal tracks in the snow

    I should set up an IR camera to watch what’s going on out there!

  • Epson R380: Re-re-routed CISS Tubing

    Alas, what seemed like a better tube route didn’t work any better and, in fact, the generous loop snagged crosswise between the print head box and the R380’s frame. So I deployed the big diagonal cutters and a nasty end cutter to chop a channel through the side of the box:

    Epson R380 - modified print head box
    Epson R380 – modified print head box

    As far as I can tell, the thin section above the reinforcing gridwork exists specifically to get in the way of routing CISS tubes, but I suppose it could be just for pretty.

    With the tubes coming directly off the top of the tanks and folding neatly as the print head moves under the frame, I could rearrange the supports to hold the tubes in a nearly straight line throughout their motion:

    Epson R380 - straight CISS tube route
    Epson R380 – straight CISS tube route

    So far, so good.

    Although the yellow ink now feeds properly and all the nozzles appear up on the test page, the printer output has an overall cyan tinge that gave the Annual Christmas Letter a gloomy aspect. Maybe the latest bottle of cheap Light Cyan ink isn’t quite as light as it should be?

  • Cycliq Fly6: Battery Replacement

    After verifying my defunct Cycliq Fly6 has a dead battery, I ordered a handful of 18650 cells from Batteryspace (a known-reputable source):

    Fly6 - battery replacement - Samsung label
    Fly6 – battery replacement – Samsung label

    The transplanted protection PCB goes between the tabs, with a nickel strip snippet because I didn’t cut the old strip in the right place:

    Fly6 - battery replacement - PCB
    Fly6 – battery replacement – PCB

    The PCB goes under a manila paper layer, the ends get similar caps, and the whole affair receives an obligatory Kapton tape wrap:

    Fly6 - battery replacement - endcap
    Fly6 – battery replacement – endcap

    Reassembly is in reverse order. I now know the Fly6 will reset / start up when the battery connector snaps into place, but, because it emits identical battery-charge beeps when it starts and shuts off, there’s no way to tell what state it’s in. I don’t see any good way to install the ribbon cable from the LED PCB before plugging in the battery, so just blindly press-and-hold the power button to shut it off.

    After an overnight charge, it makes videos of my desk just fine and will, I expect, do the same on the bike.

    Now that I’ve taken the thing apart, I should open it up and tinker with the (glued-down) camera focus adjustment to discover whether:

    We’ll find out when the weather warms up in a week or two.

  • MPCNC: GCMC Text vs. Speed

    GCMC includes single-stroke fonts derived from Hershey fonts, so I added a legend to the Spirograph Shakedown generator:

    MPCNC - GCMC Text - 3000 mm-min
    MPCNC – GCMC Text – 3000 mm-min

    Obviously, plotting 2.5 mm tall characters at 3000 mm/min = 50 mm/s isn’t a Good Idea on a less-than-absolutely-rigid CNC machine.

    Slowing down to 250 mm/min = 4.2 mm/s produces much better results:

    MPCNC - GCMC Text - 250 mm-min
    MPCNC – GCMC Text – 250 mm-min

    A closer look, albeit with less-than-crisp focus:

    MPCNC - GCMC Text - 250 mm-min - detail
    MPCNC – GCMC Text – 250 mm-min – detail

    This isn’t a conclusive test, but it reminds me that Speed Kills.

    The green plotter pen started life with a standard 0.3 mm felt nib, but it’s worn somewhat wider over the intervening years decades. Those 2.5 mm characters would look better coming from a narrow ceramic pen, which would require a pen change before doing the legend; using 4 mm characters would produce better results.

    The line spacing is 110% of the font X height, which obviously isn’t quite enough. Something on the order of 150% should look better.

    This GCMC code (including those mods) produces the legend:

    feedrate(250mm);
    
    textsize = [4.0mm,4.0mm];
    textat = [0.5*PlotSize.x/2,-PlotSize.y/2 + 2*textsize.y];
    
    textpath = typeset("Seed: " + PRNG_Seed + "  Stator: " + StatorTeeth + "  Rotor: " + RotorTeeth,FONT_HSANS_1);
    scalepath = scale(textpath,textsize);
    placepath = scalepath + textat;
    engrave(placepath,1.0mm,-1.0mm);
    
    textpath = typeset("Offset: " + L + "  Lobes: " + Lobes + "  Turns: " + Turns,FONT_HSANS_1);
    scalepath = scale(textpath,textsize);
    placepath = scalepath + textat + [-,-1.5*textsize.y];
    engrave(placepath,1.0mm,-1.0mm);