Category: Machine Shop

Mechanical widgetry

HQ Sixteen: Bobbin Winder Split Shaft Tweak
The HQ Sixteen has much larger bobbins than Mary’s Kenmore and Juki sewing machines. It also came with a dedicated bobbin winder:

HQ Sixteen bobbin winder – overview

That thing has a distinct Industrial Revolution aspect compared to the BarbieCore bobbin winder I laid hands on a while ago.

Out of the photo on the right:
- The thread cone and guide tower
- The thread tension disks
Mary had been having trouble winding the bobbins, as the tension seemed entirely too low and the thread did not lay smoothly across the bobbin, so she asked me to take a look.

The motor shaft has an O-ring for friction drive against the large wheel driving the shaft with the bobbin on the other end. The small silver lever over on the left flips an over-center lock pressing the wheel against the O-ring and tripping the microswitch in the aluminum housing, thus turning the motor on. The bobbin fills until a small finger monitoring the thread level flips the lock back over center, the wheel disengages, the switch turns the motor off, and a spring drives the wheel against the rubber rod in the upper left.

Which worked well, but all the bobbins had a loose-to-sloppy fit on the shaft, to the extent that the shaft really couldn’t drive them against any thread tension.

Loosening the screw holding the drive wheel on the shaft lets it slip off and the shaft slides out to the front:

HQ Sixteen bobbin winder – split shaft

The sides of the split shaft should press firmly against the bobbin core, but that just wasn’t happening.

Measuring a dozen bobbins showed most had an ID of 6.04 mm, with a few around 6.01 mm; unsurprisingly, the latter had the best, albeit still loose, fit. Conversely, the split shaft had two isolated points 6.01 mm apart across a diameter, with the remainder around 5.95 mm. Those are not large differences, but it was obvious why the bobbins didn’t wind correctly.

I filed some graunch off the split edges, then gently pushed the Designated Prydriver into the end of the split to spread the sides juuuust a little bit, until all the bobbins pushed on firmly and fit snugly:

HQ Sixteen bobbin winder – split shaft test fit

It reassembled in reverse order and we’ll see how it behaves during the next marathon bobbin-filling session.
2025-04-15
Champion Hose Nozzle: Refreshed Seal Attempt

The battered Champion hose nozzle came into play last fall, leaked profusely when turned off, went to a Safe Place for the winter, and recently emerged:

Champion hose nozzle – disassembled

The conical surface (to the right of the tip) must make perfect contact with the edge of a perfect cylindrical hole in the outer shell to shut off the water, which was obviously no longer happening.

There is no reason why that hole should still be concentric with the outside of the shell, but centering the latter in the four-jaw chuck put the hole within about 0.2 mm of where it should be:

Champion hose nozzle – lathe centering

I defined that to be Close Enough™ and made the hole smooth & concentric with a teeny boring bar and sissy cuts. A drill would likely have worked well enough, too.

Gently filing the nastiness off the cone showed it wouldn’t suffice, so center it while noting the irregular diameter all around:

Champion hose nozzle – lathe centering cone

A skim cut revealed the need for more attention:

Champion hose nozzle – scarred cone

Another tenth of a millimeter improved its disposition:

Champion hose nozzle – improved cone

Gentle touchup with a fine file reserved for special occasions may have been a further improvement:

Champion hose nozzle – finish filed

Add a dollop of silicone grease to encourage the shell to turn much more easily on the O-ring, reassemble in reverse order, and top it off with a new hose washer.

A quick test on a reasonably warm day showed the cone met the cylinder poorly enough to consign this nozzle to the brass recycling box.

It was fun trying, though …

2025-04-14
Inkscape: LightBurn Layer Color Palette
Inkscape is not a CAD program (neither is LightBurn), but for my simple needs it works well enough, with the compelling advantage that OpenSCAD can import named layers and extrude them into solid models.

LightBurn can import Inkscape SVG images to define the patterns for laser cutting / engraving and will automatically put the vectors into layers corresponding to their colors if and only if the SVG image uses colors from the LightBurn palette. Regrettably, picking those colors from the default Inkscape palette is essentially impossible, but you can have Inkscape use a palette file that displays only the LightBurn colors corresponding to its layers.

I conjured this GIMP / Inkscape palette file based on the table in a LightBurn forum post, plus tool layer colors from another post:
```
GIMP / Inkscape Palette
Name: LightBurn Layers
#
  0   0   0 BLACK
255 255 255 WHITE
  0   0   0 LBRN #0
  0   0 255 LBRN #1
255   0   0 LBRN #2
  0 224   0 LBRN #3
208 208   0 LBRN #4
255 128   0 LBRN #5
  0 224 224 LBRN #6
255   0 255 LBRN #7
180 180 180 LBRN #8
  0   0 160 LBRN #9
160   0   0 LBRN #10
  0 160   0 LBRN #11
160 160   0 LBRN #12
192 128   0 LBRN #13
  0 160 255 LBRN #14
160   0 160 LBRN #15
128 128 128 LBRN #16
125 135 185 LBRN #17
187 119 132 LBRN #18
 74 111 227 LBRN #19
211  63 106 LBRN #20
140 215 140 LBRN #21
240 185 141 LBRN #22
246 196 225 LBRN #23
250 158 212 LBRN #24
 80  10 120 LBRN #25
180  90   0 LBRN #26
  0  71  84 LBRN #27
134 250 136 LBRN #28
255 219 102 LBRN #29
243 105  38 LBRN T1
 12 150 217 LBRN T2
```
Plunk that file (which I named Lightburn.gpl) into /home/ed/.config/inkscape/palettes/, restart Inkscape, then select it (the Name line defines its mmm name):

Inkscape – selecting LightBurn palette

Which lays a row of the LightBurn layer colors along the the Inkscape window:

Inkscape – LightBurn palette

The text after the RGB triplet in each file line appears as the tool tip for the color swatch:

Inkscape – LightBurn palette

Because LightBurn uses only the vector Stroke and ignores its Fill, you (well, I) must become accustomed to Shift-clicking palette colors.

You can fetch a similar palette file directly from the LightBurn doc, although minus the tool tips. GIMP and Inkscape have many palettes available, should you make artsy drawings where subtle color shading matters.

I generally use only a few cheerful primary colors, because I have trouble distinguishing (heck, in some cases even seeing) the more subtle colors against LightBurn’s light (or dark) workspace background. I assign the layer cut settings using the Material Library: reds for cutting, blues for marking, and grays for engraving.

When I need more than maybe half a dozen colors, I (eventually) realize I’m trying to be too clever and split the project into separate LightBurn files.
2025-04-11

HQ Sixteen: Front Horizontal Spool Adapter

Mary wanted a horizontal spool adapter mounted closer to the front of her HQ Sixteen, in the M5 threaded hole where the Official Horizontal Adapter would go:

HQ Sixteen - front spool adapter - installed — HQ Sixteen – front spool adapter – installed

Yes, the pin through the spool is fluorescent edge-lit orange acrylic that looks wonderful in sunlight and is much more amusing than the black rod in the adapter atop the power supply pod.

The top of the machine case is not flat, level, or easy to model, so I deployed the contour gauge again, with some attention to keeping the edge pins parallel & snug along the machine sides:

HQ Sixteen – machine profile measurement

Tracing the edge of the pins onto paper, scanning, and feeding it into Inkscape let me lay a few curves:

HQ Sixteen – top profile curve – Inkscape fitting

The laser-cut chipboard test pieces show the iterations producing closer and closer fits to the machine.

Importing the final SVG image into OpenSCAD and extruding it produced a suitable solid model of the machine’s case:

HQ Sixteen - machine solid model — HQ Sixteen – machine solid model

Subtract that shape from the bottom of the adapter to get a perfect fit atop the machine:

HQ Sixteen - horizontal thread spool adapter - front pin - solid model - show — HQ Sixteen – horizontal thread spool adapter – front pin – solid model – show

Early results are encouraging, although the cheap polyester thread Mary got from a friend’s pile and is using for practice untwists itself after passing through the tension disks on its way to the needle. She’ll load much better thread for the real quilt.

The OpenSCAD source code and SVG of the HQ Sixteen’s top profile as a GitHub Gist:

	// HQ Sixteen – horizontal thread spool adapter for front pin
	// Ed Nisley – KE4ZNU
	// 2025-04-07

	include <BOSL2/std.scad>

	Layout = "Show"; // [Show,Build,Base,Wall,Frame]

	/* [Hidden] */

	Protrusion = 0.1;
	HoleWindage = 0.2;

	ID = 0;
	OD = 1;
	LENGTH = 2;

	WallThick = 8.0;
	BaseThick = 12.0;
	Washer = [5.0,10.0,1.0]; // M5 washer

	Spool = [0.25*INCH,50.0,55.0]; // maximum thread spool
	SpoolClearance = [2.0,5.0,5.0]; // spool pin pointed to +X axis
	SpoolPin = [Spool[ID],Spool[ID],Spool[LENGTH] + WallThick + SpoolClearance.x];

	BasePlate = [WallThick + SpoolClearance.x + 13.0, // X flush with side of machine
	Spool[OD]/2 + 2*SpoolClearance.y,
	BaseThick];

	BaseOffset = [-(BasePlate.x – Washer[OD]),-Washer[OD],0.0]; // left front corner w.r.t. pin


	SpoolOC = [0, // relative to left front top of Base
	BasePlate.y/2,
	SpoolClearance.z + Spool[OD]/2 + BaseThick/2];





	//———-
	// Construct the pieces

	// HQ Sixteen top frame profile
	// Aligned with hole somewhere along X=0, front edge at Y=0
	// Lengthened slightly to cut cleanly

	module MachineFrame(Length=BasePlate.y + 2*Protrusion) {

	back(BasePlate.y + Protrusion) xrot(90)
	linear_extrude(height=Length,convexity=5,center=false)
	import("HQ Sixteen – top profile curve.svg",layer="Top Profile");

	}

	// Baseplate
	// Aligned with hole one washer diameter in from corner

	module Base() {

	$fn=18;

	difference() {
	fwd(Washer[OD])
	difference() {
	right(Washer[OD])
	cuboid(BasePlate,anchor=RIGHT+FRONT+CENTER,rounding=BaseThick/2,edges=RIGHT);
	MachineFrame();
	}
	down(BasePlate.z)
	cylinder(d=SpoolPin[OD] + HoleWindage,h=2*BasePlate.z);
	up(BasePlate.z/2 – Washer[LENGTH])
	cylinder(d=Washer[OD] + HoleWindage,h=2*Washer[LENGTH]);

	}

	}

	// Wall holding spool pin

	module Wall() {

	$fn=36;

	translate(BaseOffset) {
	difference() {
	union() {
	translate(SpoolOC)
	right(WallThick)
	cylinder(SpoolClearance.x,d=Spool[OD]/2,orient=RIGHT);
	hull() {
	translate(SpoolOC)
	cylinder(WallThick,d=Spool[OD]/2,orient=RIGHT);
	up(BasePlate.z/2 – 1)
	cube([WallThick,BasePlate.y,1],center=false);
	}
	}
	translate(SpoolOC) left(Protrusion)
	cylinder(SpoolPin[LENGTH],d=SpoolPin[OD],orient=RIGHT);
	}
	}

	}

	module Adapter() {
	Base();
	Wall();

	}


	//———-
	// Show & build the results

	if (Layout == "Base")
	Base();

	if (Layout == "Wall")
	Wall();

	if (Layout == "Frame")
	MachineFrame();

	if (Layout == "Show") {
	Adapter();

	color("Gray",0.5)
	MachineFrame(60);
	color("Green",0.75)
	translate(BaseOffset)
	translate(SpoolOC)
	cylinder(SpoolPin[LENGTH],d=SpoolPin[OD],orient=RIGHT,$fn=18);

	}

	if (Layout == "Build")
	up(-BaseOffset.x)
	yrot(-90)
	Adapter();

view raw HQ Sixteen – horizontal thread spool adapter – front pin.scad hosted with ❤ by GitHub

Sorry, something went wrong. Reload?

Sorry, we cannot display this file.

Sorry, this file is invalid so it cannot be displayed.

view raw HQ Sixteen – top profile curve.svg hosted with ❤ by GitHub

2025-04-09

HQ Sixteen: Padded Table Shims
The HQ Sixteen has been running at higher speeds as Mary practices using its stitch regulator and the vibrations shook several of the table shims (blocks, whatever) onto the floor. I hope a layer of EVA foam provides enough compliance to keep them in place:

HQ Sixteen – padded table shim – installed

The foam is 2 mm thick, so subtracting that from the nominal thickness makes the new blocks come out right.

A short module extracts the footprint for export as an SVG image to laser-cut both the foam and the adhesive sheet required to stick it in place:
```
module ShimPad(Thickness = PadThick) {

    if (Thickness)
        linear_extrude(height=Thickness)
            projection(cut=true)
                ShimBlock();
    else
        projection(cut=true)
            ShimBlock();

}
```
It turns out linear_extrude() chokes on a zero height.

When handed a nonzero Thickness, the code generates a simulated foam sheet:

HQ Sixteen – table shims – solid model – padded

The footprint looks about like you’d expect:

HQ Sixteen – padded table shim – installed

Import into LightBurn, duplicate it sufficiently, set the speed & power & kerf for EVA foam, then cut ’em out:

HQ Sixteen – table shims – padding cuts

Ditto for the adhesive, stick together, and upgrade the fleet.

If these shake loose, snippets of adhesive film will stick them firmly to the underside of the table panels.

Update: Yeah, they needed sticky snippets. Whole lotta shakin’ goin’ on with that machine!
2025-04-08
HQ Sixteen: Stitch Regulator
The stitch regulator on our Handi-Quilter HQ Sixteen uses a pair of encoder wheels running along the tracks supporting the machine:

HQ Sixteen – stitch regulator sensor – rear

This must be HandiQuilter’s very first encoder version, because a ribbon cable connects the encoders to the control pod:

HQ Sixteen – stitch regulator sensor – rear

I stuck an adhesive cable clamp under the machine to rein in some of the slack, but the jank is strong with that arrangement and I must figure out a better arrangement with supple cable and better support. We’ll run this lashup for a while.

Anyhow …

The stitch regulator uses signals from the wheels to measure the distance the machine travels across the fabric and controls the motor speed to produce a fixed number of stitches per inch at that travel speed, as set in the control panel:

HQ Sixteen – display – stitch mode

Close inspection shows the LCD module came from an early 2000s mobile phone, but there’s no shame in repurposing cheap & readily available hardware.

When the stitch regulator is not active, the machine runs at a fixed speed set on the control panel:

HQ Sixteen – display – speed mode

The controller can set the speed between 100% to 10% of the motor’s 1500 stitch/min full speed, with 1% steps that seem too large on the low end and too small for the high end. Aiming my laser tachometer at a retroreflective tape snippet on the handwheel shows the machine runs at the correct fractions of its actual 1492 stitch/min = RPM.

The stitch regulator uses the same motor speed range, which sets corresponding limits on the maximum and minimum speeds across the fabric, with the ratio set by the stitch/inch value.

At the 10 stitch/inch setting Mary has been using, the travel speed range is:
- 15 inch/min = (150 stitch/min) / (10 stitch/inch)
- 150 inch/min = (1500 stitch/min) / (10 stitch/inch)
When you stop moving the machine, the controller will shut off the motor after a few stitches in the same place, which turns out to be convenient for tying off the end of a stitched line on a quilt. When you move too fast, the machine will top out at 1500 stitch/min while producing too-long stitches until the travel speed drops below 150 inch/min.

What’s not obvious is how slow those speeds are:
- 0.25 inch/s = 15 inch/min
- 2.5 inch/s = 150 inch/min
As an exercise, fire up the metronome app on your phone at one tick per second, then try drawing intricate patterns within those speed limits. You will inevitably move too fast, even without the soundtrack of a frantically accelerating motor topping out at 1500 RPM.

We think the surprisingly low upper speed limit accounts for much of the trouble Mary’s compadres report while using the stitch regulator.

After laying down a few square yards of practice quilt patterns while measuring the results and becoming accustomed to the sound and feel of the machine running at high speeds, Mary’s producing good results:

HQ Sixteen – stitch regulator – counting stitches

I definitely hit the knees in gratitude when the stitch regulator Just Worked™ after plugging it in, because that ribbon cable did not inspire any confidence whatsoever.
2025-04-07
Vole Trap Boxes: Deluxe Edition

The larger vole trap boxes didn’t survive the early spring rainfall, so we decided to upgrade the fleet with more durable boxes:

Vole Box – finished

I obviously need a larger light box.

The trap boxes come in 7 quart and 3.5 quart sizes, although we expect either will comfortably accommodate a single vole.

They’re made of polypropylene plastic eminently suited for laser cuttery, so I borrowed the holes from the cardboard box setup:

Vole Box – hole cutting

The clamps on the knife bars held the angle block and boxes in pretty much the same position, so I didn’t realign anything after figuring out a pair of magnets would hold the lid to the angle:

Vole Box – lid fixture magnets

The box side is slightly sloped, so I probably should have angled the block to tilt the lid, but this isn’t a precision job:

Vole Box – lid fixture

The white smudges on the lid come from vaporized polypropylene:

Vole Box – fume deposits

The body count thus far is just one field mouse, but the season is yet young.

2025-04-04