Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.
The MDF signs I made last year disintegrated pretty much on the expected schedule, so it’s time for something more durable:
Please Close The Gate – acrylic engraving
The idea is to engrave both sides of a 3 mm orange acrylic sheet, shoot it with rattlecan black paint, and declare victory. The second step awaits warmer weather, but at least I’m doing my part to prepare for the new gardening season.
Vaporizing that much acrylic produces a fair bit of debris:
Please Close The Gate – acrylic dust on laser head
Some dust / vapor accumulates / condenses on the honeycomb platform beyond the orange sign, but most of it gets through to the baffle on the exhaust duct:
Please Close The Gate – acrylic dust on exhaust port
A closer look shows it really does grow out from the perimeter of each hole:
Please Close The Gate – acrylic dust on exhaust port – detail
Now, if that doesn’t trip your trypophobia, nothing will …
A few passes with the trusty Electrolux vacuum’s dust brush brought the visible surfaces back to normal.
By now, the duct fan blades have surely layered on a good coating, too, which shall remain undisturbed until I find a better reason to open the duct.
A LightBurn forum discussion about problems making Z-scale (1:220) bricks led me to trying a few ideas on the way to figuring out what was going wrong.
Each brick is about 1.0×0.5 mm, so an entire wall doesn’t cover much territory:
Z-scale bricks – assortment
Yes, those are millimeters along the scale.
The kerf on my 60 W CO₂ laser seems slightly wider than the “mortar” lines should be, so I made a layout with the vertical lines slightly inset from the horizontal ones:
Z Scale Brick Wall – LB layout
That let the kerf complete the lines without burning into the adjacent bricks:
Z Scale Brick Wall – laser lines
The cuts are obviously too wide (and deep!), but just for fun I colored the chipboard with red marker and rubbed a pinch of flour into the lines:
Z Scale Brick Wall – color and flour
Which looks chunky, but not terrible, for what it is. Maybe concrete blocks would look better?
The next attempt started with a raster bitmap scaled at 254 dpi = 10 pix/mm, so that single-pixel “mortar” lines between 10×5 pixel bricks would be 0.1 mm wide:
Raster Z-Scale Bricks
Scanning the image at 100 mm/s makes each pixel 1 ms “wide” and, because the power supply risetime is on the order of 1 ms, the laser won’t quite reach the 10% power level across the vertical lines:
Raster Z-Scale Bricks – LB layer settings
The raster lines come out lighter and (IMO) better looking:
Z Scale Brick Wall – raster lines
The horizontal lines are darker because the beam remains on at 10% across their full length, but the overall result seems much closer to the desired result.
The original poster will use a diode laser and, combining all the ideas we came up with, now has a path toward making good, albeit invisibly small, bricks.
Sharpereyes than mine pointed out I misspelled Poughkeepsie, so I took advantage of the opportunity to make the whole thing look better:
Library card tag – revised front
It turns out the low-surface-energy tape stuck like glue to the acrylic tag (because that’s what it’s designed for) and peeled right off the laminating film on the printed paper. So I stuck some ordinary adhesive film to the back of the new paper label, left its protective paper on the other side, cold laminated the film+paper, laser-cut the outline, peeled off the back side of the laminating film with the protective paper, and stuck the new adhesive to the LSE tape still on the tag.
I have no idea how well this will work out in the long term, what with two adhesive layers bonded to each other, but this whole thing is in the nature of an experiment.
During the course of diagnosing and fixing the latest oven igniter failure, an unrelated series of events produced a flood under the kitchen sink and across the floor. After cleaning up the mess and determining the floor under the cabinet was merely damp, rather than wet, I drilled a hole suitable for another PC cooling fan from the Box o’ MostlyFans, installed the fan to pull air upward, and let it run for a couple of days while watching the humidity drop.
Fortunately, I had a hole saw exactly the right size for an 80 mm case fan:
Kitchen sink – fan cover plate
I will lay big money on a bet saying your kitchen cabinets don’t have Real Wood like that, nor are the interiors painted bold Chinese Red. This place really is a time capsule from 1955.
While the drying happened, I made a hole cover from 1.5 mm black acrylic and, there being no style points involved, rounded up a quartet of black-oxide self-drilling sheet metal screws to hold it in place.
Although it’s not obvious, there’s a layer of transparent plastic “shelf paper” in there. It covers the fan hole under the cover, so any future spills will have approximately the same difficulty reaching the floor as this one did.
The LightBurn layout produces both the fan cover and a template to mark the four screw holes around the fan opening:
Kitchen Sink Fan – LB layout
The blue tool layer lines serve as a guide for the rest of the cover layout; the matching orange square on the right marks the fan outline on the drill template as a quick size check.
No need for an SVG version, because now that you have the general idea, it’s easy to recreate it for your own fan.
Another LightBurn forum discussion helped me work through how the Z Axis motions should work. This is a lightly edited mashup of several of my comments and builds off a discussion concerning the proper setup of the axis homing / direction settings; the starting point concerns whether the “up” jog arrow should make the platform go up or down.
It may be a matter of definitions and the consistency thereof.
An earlier comment in the thread said “The Machine’s Z-Axis operates in the wrong direction!”, so I had (erroneously) suggested flipping the Direction Polarity control to reverse it, which made it move in the other (wrong) direction when homing.
Knowing that, I suggested restoring Direction Polarity for the correct homing direction, then flipping Invert Keypad Direction reverse both the keypad and LightBurn directions.
If that does not sort the directions out the way you want, then it’s not clear how to proceed.
The Focus Distance parameter determines how far the U (or Z) axis moves from its default position after the focus pen / switch trips: it adds distance and can only be positive. Mine arrived at 0.0 mm and remains that way.
That default position comes from the U (or Z) axis parameter Home Offset controlling the backoff distance from the switch trip point. Mine is at 10.1 mm, which positions the nozzle 18.5 mm from the material and puts the focal point at the surface.
I think the intent is to have the vendor determine Home Offset to make the focus switch work correctly with a minimum mechanical backoff, then add Focus Distance to match the actual lens focal distance. The settings on my machine came from OMTech, but I don’t regard them as unalterably correct.
For my Sherline CNC mill, jogging “up” increases the distance between the tool and the table by raising the spindle.
For my MakerGear M2 3D printer, jogging “up” increases the distance between the nozzle and the platform by lowering the platform.
In both cases, the “up” button corresponds to an increasing distance.
I think the idea behind the Ruida’s setup parameters is to put the just-homed Z axis origin at the platform surface, with the jog buttons (and LightBurn’s motions) then raising the focus point to the surface of the material by lowering the platform: positive numbers increase the distance.
With that in mind, picking the Invert keypad direction setting so that the up button makes the platform go down is correct: it increases the distance from the initial home position. That should also make positive Z steps increase the distance (away from the work) and negative steps decrease it (into the work), which seems sensible.
A discussion on the LightBurn forum prompted an investigation of how Ruida laser cutter controllers handle their homing operation. This is a lightly edited version of (some of) my comments with added pix.
There being nothing quite like a good new problem to take one’s mind off all one’s older problems, I just did some tinkering with the X axis settings…
This is on a Ruida-ish KT332N, but the manual’s verbiage is pretty much the same as the 6445G.
Changing the settings requires a controller reset, which will trigger a homing operation if it’s enabled.
Limiter Polarity inverts the sense of the limit switch input signal. My OMTech has Normally Open proximity sensor switches:
OMTech Fotek PL-05N proximity limit switch
So the default setting of False makes it inactive in the middle of the platform.
I set it to True, which made it active in the middle of the platform:
Ruida KT332N Controller – inverted X limit switch
So True would be the correct setting for a Normally Closed switch.
Direction polarity sets the direction the axis moves during reset, which must be toward the home switch. Mine was set to False and it homes toward the right side where the limit switch sits.
I set it to True and it started homing toward the left, where the limit switch isn’t. Hitting Esc on the control panel stopped the homing process and left the head in the middle of its travel, unhomed.
Apparently Direction Polarity directly controls the DIR signal to the stepper motors, because both the control panel keys and LightBurn’s motion buttons then moved “the wrong way”.
That does not affect LB’s machine coordinate origin, which remains at the top right of the layout, or the orientation of the X axis, with values increasing toward the left as usual.
Invert keypad direction inverts the motion direction with respect to the homing motion. Mine was False, but flipping it to True reverses both the keypad and LB directions.
I think Limit Trigger refers to “hard limits”, which requires a limit switch at each end of the axis, but my controller does not have the + inputs that would allow me to test that. The controller doc says it uses “soft limits” to prevent out-of-bounds motion by using the home position and maximum axis travel value.
So the available options allow all possible combinations of:
NC or NO limit / home switches
Homing direction
Keypad / LB motion commands
A reasonable sequence seems to be:
Set the switch polarity to match the hardware
Set the homing direction toward the switch
Set the “keypad” direction to match your expectations
With that in mind, careful experimentation on the actual controller in hand seems prudent …
The improved Holly Mirror Coaster looks pretty good:
Holly Coaster – overview
Until you realize some of those specks aren’t surface dust and take a closer look:
Holly Coaster – mirror speckles 1
The surface scratches are doubled by their reflection in the bottom mirror. The little dots that aren’t doubled reveal marks in the mirror surface itself.
In this case, they cause defects in the mirror coating allowing alcohol from the fat-tip permanent markers coloring the engraved areas to hit the acrylic. The starbursts come from stress cracks around the punctures.
Peering even closer shows similar cracks along the edges of the colored areas:
Holly Coaster – mirror speckles tight detail
Not much to do about the random speckles, but it’s obvious I must up my coloring game.
Which would be significantly easier if rattlecan spray paint sprayed at winter temperatures …
The Smell of Molten Projects in the Morning 