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.
Few enough folks stop by to make the bump from Hackaday’s mention of the Wasp Blower stand out:
Apparently Subaru fuses remain elusive and that old Thermador wall heater retains its hotness:
The WordPress AI came up with this prompt and image based on the minimal text above:
Gotta get me a workshop like that …
The AI Assistant says this post needs improvement:
Which is why I don’t use AI for anything …
Unbelievably, the ankle zipper tab broke off in my hand:
It’s one of those zippers where the tab releases a lock preventing the zipper from coming unzipped. Mary noped out of removing and replacing the entire zipper.
Trimming a snippet of aluminum miniblind from the Small Box o’ Flat Stuff and two dots of JB Kwikweld epoxy seemed appropriate:
Ugly, but serviceable:
The stray epoxy scraped off under fingernail pressure over the next two days and the pants are ready for the next snowfall.
A conical (a.k.a. bullnose) center in the tailstock simplifies supporting cylindrical objects:
The spring-loaded tailstock bearing has a 5 mm bore. The bullnose rests against a small spacer on its 5 mm shaft to hold it away from the bearing’s mounting screws with some bearing spring compression. I turned the spacer from aluminum rod because lathe work is satisfying, but a printed spacer would work fine.
The bullnose is a cone with steps encouraging the cylinder to sit properly:
With both ends centered, the cylinder sits concentric with the chuck axis:
The chuck grabs the OD and the bullnose supports the ID, so removing crud from both ends is in order.
The bullnose won’t work for a solid rod, so a negative cone = cup center may come in handy:
Stipulated: A CO₂ laser will bounce right off a solid aluminum rod. Imagine I chucked up a wood dowel, OK?
A cup center is what remains afteryoinking a bullnose out of a cylinder:
Looks like I did exactly that:
Somewhat surprisingly, the two parts nest perfectly:
That’s without the shaft installed on the cup, so they won’t sit quite so neatly on the shelf.
Aligning the rotary axis along the laser’s X axis and setting the focus requires attention to detail, but a decent tailstock center makes that effort meaningful.
The OpenSCAD code as a GitHub Gist:
| // Ortur Rotary Conical centers | |
| // Ed Nisley – KE4ZNU | |
| // 2025-12-27 | |
| include <BOSL2/std.scad> | |
| Style = "Bullnose"; // [Build,Cone,Bullnose,Cup,Cone] | |
| MinDia = 10.0; | |
| MaxDia = 50.0; | |
| /* [Hidden] */ | |
| LayerThick = 0.2; // should match slicer thickness | |
| Ramp = 1.0; | |
| ID = 0; | |
| OD = 1; | |
| LENGTH = 2; | |
| HoleWindage = 0.2; | |
| Protrusion = 0.1; | |
| NumSides = 8*3*4; | |
| $fn=NumSides; | |
| Gap = 5.0; | |
| WallThick = 2.0; | |
| TailBearing = [5.0,7.0,10.0]; // tailstock shaft, LENGTH = insert depth | |
| StepHeight = 2*LayerThick; | |
| NumSteps = (((MaxDia – MinDia)/2) / Ramp); | |
| ConeOAH = NumSteps * (Ramp + StepHeight); | |
| //—– | |
| // Bullnose shape | |
| module Bullnose() { | |
| difference() { | |
| union() | |
| for (i = [0:NumSteps – 1]) | |
| up(i*(Ramp + StepHeight)) hull() | |
| cyl(StepHeight + Protrusion,r=(MaxDia/2 – i*Ramp),anchor=BOTTOM) position(TOP) | |
| cyl(Ramp,r1=(MaxDia/2 – i*Ramp),r2=(MaxDia/2 – (i+1)*Ramp),anchor=BOTTOM); | |
| } | |
| } | |
| module Cone() { | |
| difference() { | |
| Bullnose(); | |
| down(Protrusion) | |
| cyl(TailBearing[LENGTH] + Protrusion,d=TailBearing[ID],circum=true,anchor=BOTTOM); | |
| } | |
| } | |
| module Cup() { | |
| difference() { | |
| cyl(ConeOAH + TailBearing[LENGTH],d=MaxDia + 2*WallThick,anchor=BOTTOM); | |
| up(ConeOAH + TailBearing[LENGTH] + Protrusion) | |
| yrot(180) | |
| Bullnose(); | |
| down(Protrusion) | |
| cyl(TailBearing[LENGTH] + 2*Protrusion,d=TailBearing[ID],circum=true,anchor=BOTTOM); | |
| } | |
| } | |
| //—– | |
| // Build things | |
| if (Style == "Bullnose") | |
| Bullnose(); | |
| if (Style == "Cone") | |
| Cone(); | |
| if (Style == "Cup") | |
| Cup(); | |
| if (Style == "Build") { | |
| right(MaxDia/2 + Gap) | |
| Cone(); | |
| left(MaxDia/2 + WallThick + Gap) | |
| Cup(); | |
| } |
Stipulated: A chuck rotary doesn’t need a home switch.
With that in mind, a home switch seemed like it might come in handy and this is the simplest workable design:
The cover mimics the size & shape of the Ortur cover, minus the stylin’ rounding & chamfering along the edges:
It has a certain Cybertruck aspect, doesn’t it?
Two beads of hot melt glue hold the switch flush along the cover’s inside surface:
One might argue for a tidy cover over those terminals.
While contemplating the layout by holding the switch here & there, seeing the switch roller neatly centered on the pulley hub told me the Lords of Cosmic Jest favored this plan:
A simple cam lifts the roller:
That’s obviously laser-cut acrylic sitting on double-sided tape.
Edit: The pulley ratio is 1:3, so the step/rev value is three times the DIP switch setting on the stepper driver.
Some finicky repositioning put the #1 chuck jaw on top after homing:
A more permanent adhesive under the cam may be in order.
Update: The switch triggers more reliably with a simple setscrew standing proud of the pulley hub:
Wiring the normally open switch contacts in parallel with the existing Y axis home switch lets both the gantry and the rotary trigger the controller. The front-panel switch ensures only one of those two can move:
With all that in place and the switch flipped, the chuck rotates happily and homes properly with the controller in normal linear mode.
Spoiler: A Ruida-ish KT332N controller ignores the Y-axis Home enable setting with Rotary mode enabled, because everybody knows a rotary has no need for a home switch.
The OpenSCAD code as a GitHub Gist:
| // Ortur Rotary belt cover | |
| // Ed Nisley – KE4ZNU | |
| // 2025-12-23 | |
| include <BOSL2/std.scad> | |
| Layout = "Show"; // [Show,Build,Block,Shell] | |
| /* [Hidden] */ | |
| ID = 0; | |
| OD = 1; | |
| LENGTH = 2; | |
| HoleWindage = 0.2; | |
| Protrusion = 0.1; | |
| NumSides = 2*3*4; | |
| $fn=NumSides; | |
| Gap = 5.0; | |
| WallThick = 1.6; // OEM wall | |
| CoverOA = [81.5,50.5,23.0]; // open side down | |
| CoverRadius = 4.0; | |
| CoverTrimZ = 6.0; | |
| CoverTrimAngle = 45; | |
| BreakX = (CoverOA.z – CoverTrimZ)/tan(CoverTrimAngle); | |
| ScrewOC = [51.0,38.0]; | |
| ScrewHoleID = 3.5; | |
| ScrewHeadRecess = [ScrewHoleID,7.0,1.8]; | |
| ScrewOffset = 8.0; // cover edge to hole centerline | |
| SwitchOA = [21.0,20.0,6.5]; // X = body + roller, excludes terminals | |
| SwitchOffset = [0,0,17.0]; // nominal end = roller at centerline | |
| //—– | |
| // Overall cover shape | |
| module CoverBlock() { | |
| cuboid([CoverOA.x,CoverOA.y,CoverTrimZ],anchor=BOTTOM) position(TOP+LEFT) | |
| prismoid(size1=[CoverOA.x,CoverOA.y],size2=[CoverOA.x – BreakX,CoverOA.y], | |
| height=CoverOA.z – CoverTrimZ,shift=[-BreakX/2,0],anchor=BOTTOM+LEFT); | |
| } | |
| // Cover shell | |
| module CoverShell() { | |
| difference() { | |
| CoverBlock(); | |
| down(Protrusion) | |
| resize(CoverOA – [2*WallThick,2*WallThick,WallThick – Protrusion]) | |
| CoverBlock(); | |
| } | |
| } | |
| // The complete cover | |
| module Cover() { | |
| difference() { | |
| union() { | |
| CoverShell(); | |
| left((CoverOA.x – ScrewOC.x)/2 – ScrewOffset) | |
| for (i = [-1,1], j=[-1,1]) | |
| translate([i*ScrewOC.x/2,j*ScrewOC.y/2,0]) | |
| cyl(CoverOA.z,d=ScrewHoleID + 2*WallThick,anchor=BOTTOM); | |
| } | |
| left((CoverOA.x – ScrewOC.x)/2 – ScrewOffset) down(Protrusion) | |
| for (i = [-1,1], j=[-1,1]) | |
| translate([i*ScrewOC.x/2,j*ScrewOC.y/2,0]) { | |
| cyl(CoverOA.z + 2*Protrusion,d=ScrewHoleID + HoleWindage,anchor=BOTTOM); | |
| up(CoverOA.z – ScrewHeadRecess[LENGTH]) | |
| cyl(ScrewHeadRecess[LENGTH] + 2*Protrusion, | |
| d1=ScrewHeadRecess[ID] + HoleWindage,d2=ScrewHeadRecess[OD] + HoleWindage, | |
| anchor=BOTTOM); | |
| } | |
| translate(SwitchOffset) left(CoverOA.x/2 – WallThick – Protrusion) | |
| cuboid(SwitchOA,anchor=RIGHT+FWD); | |
| } | |
| } | |
| //—– | |
| // Build things | |
| if (Layout == "Block") { | |
| CoverBlock(); | |
| } | |
| if (Layout == "Shell") { | |
| CoverShell(); | |
| } | |
| if (Layout == "Show") { | |
| Cover(); | |
| } | |
| if (Layout == "Build") { | |
| up(CoverOA.z) | |
| xrot(180) | |
| Cover(); | |
| } | |
Because the focus pen worked on the bench, I was certain this had to be true:
There is a break somewhere along the blue wire carrying 24 V to the focus pen. The signal and 0 V wires are fine.
I updated the original post, because I’m going to use that picture a lot whenever the subject of laser machine wiring comes up.
A box of air filters that Came With The House™ (and fit nothing therein) surfaced during a recent heap probe and prompted a quick-n-dirty project:
It replaces a tired box fan (barely visible at the top) that’s been shoving air around the basement to equalize the humidity.
The quintet of 140 mm fans seems quieter, although they don’t move quite as much air. Given that I have no way to know how much air circulation is enough, it’s likely sufficient.
The strip of black tape covers a hole for the knob on the fan power / speed control, although I cranked it up to full throttle and expect to leave it there:
The controller sits on a platform cut from 1.5 mm cardboard:
The 3D printed holder came with the controller. I cannot imagine how they have enough time to print a holder for each controller; maybe it’s a QC check for a 3D printer manufacturer.
I intended the controller to sit on the other side of the middle fan, but realized I had to cut the opening after mounting the fans and got the chirality wrong; the wiring in there layout leaves something to be desired.
The fans mount on a sheet of cardboard cut from one side of a Home Depot Extra Large Box and the bottom of the filter box comes from the other side. Because I don’t have a deep emotional attachment to the filters, they’re attached to each other (and the bottom sheet) with hot melt glue. I do have a slight attachment to the fans, but four dabs of glue hold each one in place. More gaffer tape holds the fan sheet to the front of the assembled box, in the unlikely event I must get in there again.
Hey, it’s Christmas: good things come in boxes, right?
The focus “pen” = switch on the OMTech laser stuck out far below the nozzle:
The nozzle is 18.5 (-ish) mm above the surface with the laser beam focused to a tight spot. The brass (-ish) tip of the pen flew about 5 mm above the material, requiring considerable attention to the placement of magnets, clamps, and similar accoutrements around the material on the platform.
Having dismantled the pen while replacing its wiring, this seemed like a good time to figure out how to get more clearance under its tip.
Removing the pen nose shows the tip on its 3 mm screw inside the spring pushing the tip downward:
I replaced the original spring (on the bottom) with a softer spring, mostly because the tip exerted what seemed like entirely too much force on the material. That makes no difference for acrylic & plywood, but anything squishier required deploying the focus gauge after I remembered the problem.
The other end of the screw is impossible to photograph in situ, but the tapered head seats in a recess leaving several millimeters of air below the proximity sensor. I made a little steel slug to reduce the pretravel by filling that gap:
The spigot on the slug (turned from 7/32 inch steel rod) aligns it with the screw head, with high-viscosity cyanoacrylate adhesive holding it in place:
The surface finish of my slug matches their tapering, so I figure it’s about right.
A setscrew near the top of the pen clamps the proximity sensor with a few millimeters of adjustment:
The slug reduces the pretravel to nearly zero with the sensor at the bottom of its range.
The brass tip had been twisted onto the screw as far as it would go, so I cut a few millimeters off the screw to put the tip closer to the pen nose:
Even with reduced pretravel, the tip nearly vanished into the pen body before tripping the sensor, so I unscrewed it two turns = 1.4 mm.
With the pen back in the machine and plugged in, measure the switch travel with a step gauge:
Protip: Measure the as-cut height of those steps, then either shim the bottom of the gauge with tape of a suitable thickness or add that much to the layout and cut another set.
With a good step gauge in hand:
ABFigure B-A, round up to the next millimeter, then set that value as the Home Offset for whatever axis moves the platform. My tweaked pen had 2.5 mm of travel, so I used 3.0 mm:
Adjust the pen position to put the tip more than the Home Offset below the nozzle (I picked 5 mm) to ensure the switch will trip before the nozzle contacts the platform, then do an Autofocus.
Measure the distance from the nozzle to the platform (mine was 5.5 mm), subtract that from 18.5 mm (the known focused distance for my laser head, as above), and set that as the Focus Distance:
Another Autofocus should then put the nozzle exactly 18.5 mm (or whatever your machine needs) off the platform / material.
This shows the pen now flies 5 mm below the nozzle:
The step gauge shows it’s 13.5 mm above the platform, much better than the previous 5 mm.
The switch trips juuuust before the nozzle hits the material:
I should lower the pen a millimeter, but that’s in the nature of fine tuning.
Happy Dance!
The Smell of Molten Projects in the Morning 