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.

Tag: Repairs

If it used to work, it can work again

Bird Box Entrance Reducers: Round 2

One of the bird box entrance reducers I installed nigh onto a decade ago is still on duty, although downy woodpeckers definitely want a larger hole:

Bird Box - gray PVC pipe reducer - woodpecker damage — Bird Box – gray PVC pipe reducer – woodpecker damage

Another reducer had gone missing over the years, so I made one from a length of PVC pipe:

Bird Box - PVC pipe reducer - shaping — Bird Box – PVC pipe reducer – shaping

It started as 1-½ PVC pipe, 1-⅞ inch actual OD and should fit into a 1-½ hole, so I measured 1.5 × 3.15 around the circumference, bandsawed out the excess, draped it over a 1-½ Forstner bit, toasted it with a heat gun, and squashed it so it’s just a little bit bigger than the (enlarged!) hole in the box.

Now the entrance is 1-¼ (-ish), just like it should be:

Bird Box - PVC pipe reducer - installed — Bird Box – PVC pipe reducer – installed

The bird box in the front yard has been attracting starlings, in addition to serving as a hawk perch:

New Coopers Hawks - bird box takeoff whoops — New Coopers Hawks – bird box takeoff whoops

The oblong hole required advanced manufacturing techniques:

The front face should be too slick for larger birds and the little ones will zip right into the hole:

Bird Box - 3D printed entrance reducer — Bird Box – 3D printed entrance reducer

The two starlings who’d been evaluating the box seem to have moved on; we doubt they’re now homeless.

The OpenSCAD source code as a GitHub Gist:

	// Bird Box – oval entrance reducer
	// Ed Nisley KE4ZNU 2020-02-12

	//- Extrusion parameters must match reality!
	// Print with 3 shells and 3 solid layers

	ThreadThick = 0.25;
	ThreadWidth = 0.40;

	HoleWindage = 0.2;

	function IntegerMultiple(Size,Unit) = Unit * ceil(Size / Unit);

	Protrusion = 0.1; // make holes end cleanly

	inch = 25.4;

	//———————-
	// Dimensions

	EntranceID = 1.25 * inch;

	BoxHole = [1.5,2.25] * inch;
	BoxWall = 0.75 * inch;

	HoleOC = BoxHole.y – BoxHole.x;

	FlangeWidth = 5.0;
	FlangeThick = 5*ThreadThick;

	$fn = 12*4;

	//——————-
	// Build it

	difference() {
	union() {
	linear_extrude(height=BoxWall + FlangeThick)
	hull()
	for (j=[-1,1])
	translate([0,j*HoleOC/2])
	circle(d=BoxHole.x);
	linear_extrude(height=FlangeThick)
	hull()
	for (j=[-1,1])
	translate([0,j*HoleOC/2])
	circle(d=BoxHole.x + 2*FlangeWidth);
	}
	translate([0,0,-Protrusion])
	cylinder(d=EntranceID,h=2*BoxWall);
	}

view raw Oval Entrance Reducer.scad hosted with ❤ by GitHub

2020-02-19

CNC 3018XL: Arduino + Protoneer CNC

If the truth be known, I wanted to do this as soon as I discovered the CAMtool V3.3 board hardwired the DRV8825 PCBs in 1:32 microstep mode:

CNC 3018XL – Protoneer atop Arduino – installed

The Protoneer CNC board has jumpers, so selecting 1:8 microstep mode is no big deal.

As before, I epoxied another row of pins along the I/O header for Makerbot-style endstops:

Protoneer endstop power mod

I’ll probably regret not adding pins along the entire row, but, unlike the MPCNC, the CNC 3018XL won’t ever have hard limit switches. I plugged the Run-Hold switch LEDs into an unused +5 V pin and moved on.

I modified the DRV8825 driver PCBs for fast decay mode:

DRV8825 PCB – Fast Decay Mode wire

Then set the current to a bit over 1 A:

3018XL – Protoneer setup – Z 1 mm

Six hours later I hauled the once-again-functional CNC 3018XL to my presentation for the ACM:

Spirograph – intricate sample plot – detail

Memo to Self: Time to get another Prontoneer board …

2020-01-31
CAMtool V3.3 vs. The Fat Fingers of Death

As is my custom, the day before showtime I talked my way through a final full-up dress rehearsal, with the HP 7475A plotter and the CNC 3018XL running their demo plots. As if to justify my attention to detail, the 3018 refused to home, with its X axis motor grinding in a manner suggesting something had gone terribly wrong with its driver.

OK, I can fix that™.

Turn off the power, verify the leadscrew turns smoothly by hand, check all the connections & connectors, then pull the DRV8825 PCB to see if anything looks obviously wrong. It didn’t, so I carefully re-plugged the driver and moved the whole affair to the Electronics Workbench for further study.

I turned on the scope and Tek current probes, then turned on the 3018 power supplies, whereupon a great cloud of Magic Smoke emerged from the CAMtool board and filled the Basement Laboratory with the acrid smell of Electrical Death.

It seems I carefully and meticulously re-plugged the DRV8825 PCB into its socket exactly one pin too high, which, among other Bad Things, connects the +24 V motor power supply to the driver GND pin.

Obviously, this did not end well:

CAMtool V3.3 – blown stepper fuse

The fuse, put under considerable stress, vented smoke & debris in all directions across the board; note the jets above the white motor connector. Surprisingly, the 1 kΩ resistor just below it is in fine shape, as is the rather blackened electrolytic cap.

The fuse measures the same 150-ish mΩ as the fuses in the other two axes, but I doubt it’s actually a fuse any more.

Astonishingly, the Arduino clone on the board worked fine, so I could extract the GRBL configuration.

Memo to Self: Never plug things in with your head upside down!

2020-01-30
Anonymous Bike Taillight Current

Along with the (defunct) Blackburn Flea, the bike pack also disgorged an anonymous taillight with a battery resistant to recharging through the USB port. Gentle suasion cracked the solvent-glued joint around the case:

Bike taillight – cracking case

As with most modern electronics, a battery occupies most of the interior volume:

Bike taillight – opening case

For posterity, the connections:

Bike taillight – connections

I unsoldered the cell and charged it from a bench supply:

Bike taillight – external recharge

The voltage started out low with the current held to about 100 mA, eventually rose to 4.1 V, and stayed there while the current dropped to zero. Unlike the Blackburn cell, it appears not too much worse for the experience, although I haven’t measured the actual capacity.

Clipping the Tek current probe around the LED supply wire produced this waveform for the “dim” setting:

Anonymous Taillight – Low – 200 mA-div

Adding a voltage probe across the LEDs and clicking to the “high” setting:

Anonymous Taillight – High – 200 mA-div

The intense ringing at the start of the pulse seems an artifact of the measurement setup, but ya never know; these days, RFI can come from anywhere.

In any event, the COB LED strip draws 800 mA from a fully charged battery, about 26 mA for each of the 30 LEDs. The 5% duty cycle in the “dim” setting is decently bright and 18% in “high” is entire adequate.

A trio of blinks works for daytime rides, although the fastest one seems seizure-inducing.

I’ve strapped it around a rack strut and run it at the slowest blink, on the principle you can never have too many blinky lights …

2020-01-29
LibreOffice Impress vs. NFS Network Shares vs. Caching
Whenever I put together a presentation, LibreOffice Impress gradually grinds to a halt with images in the slide thumbnails repeatedly updating and never stabilizing; eventually, LO crashes and sends a crash report to whoever’s watching. This may be due to my enthusiastic use of images to get my point(s) across, although I’m just not gonna back down from that position:

LibreOffice Impress – Thumbnail thrashing

That’s a screenshot of a small thumbnail, enlarged for visibility, so it doesn’t look that crappy in real life.

Perhaps the problem arises because I insert the images as links, rather than embedding them to create a monolithic presentation file roughly the size of all outdoors?

Searching with the obvious keywords produces tantalizing hints concerning LO’s file locks clashing with NFS network share locking, which seems appropriate for my situation with all the files living on the grandiosely named file server (a headless Optiplex) in the basement.

The suggestions include making sure the NFS locking daemon is active, but I have NFC about how that might work in practice. The lockd daemon is running, for whatever that’s worth.

Seeing as how I’m the only one editing my LO presentations, disabling LO’s locks has little downside and requires tweaking one character in one line inside /usr/bin/libreoffice:
```
# file locking now enabled by default
SAL_ENABLE_FILE_LOCKING=0
export SAL_ENABLE_FILE_LOCKING
<<< blank line to show off underscores above >>>
```
After a brief bout of good behavior, Impress resumed thrashing and stalling.

Copying the entire presentation + images to the SSD inside my desktop PC didn’t improve the situation.

More searches on less obvious keywords suggested disabling the Impress “background cache”, whatever that might be:
```
Tools → Options → Impress → General → Settings
```
Then un-check the ☐ Use background cache item, which may be the last vestige of the now-vanished memory usage and graphics cache size settings from previous versions.

In any event, disabling the cache had no effect, so it’s likely a problem deep inside LibreOffice where I cannot venture.

It autosaves every ten minutes and I must restart it maybe once an hour: survivable, but suboptimal.

There seems to be an improvement from Version 6.0.7 (Ubuntu 18.04 LTS) to Version 6.2.8 (Manjaro rolling release), although it’s too soon to tell whether it’s a fix or just different symptoms.
2020-01-26
Subaru Forester Rear Wiper Disassembly

You’re supposed to just rotate the wiper blade holder and have it pop out of the mount on the end of the arm:

Subaru Forester – rear wiper blade mount

The blade holder has two opposed pegs fitting into those curved notches to the right of the hook for the holder’s pivot, with the intent of preventing it from rotating too far and sliding out. I was unwilling to apply sufficient force to disengage those pegs, as the penalty for breaking the wrong piece of plastic seemed very high. Apparently, the pegs should ride up over the slightly lower edge of their notch, bending the holder’s sides outward as they do.

So I jammed a little screwdriver beside one of the pegs, managed to encourage it out of its notch, repeated the treatment on the other side, and the blade holder popped right out.

The front wiper arms have J-hooks on their ends and disengage easily, at least after you realize the flat panel on the blade holder is actually a latch you’re suppose to pull up-and-out to release the hook. This goes more easily when assisted with the aforementioned small screwdriver.

The blades were in good shape after five years, mostly because the Forester spends most of its time in the garage. A trio of silicone wipers should last the rest of its life, with the OEM wipers tucked into the spare tire well Just In Case.

Back in the day, one could replace just the blades, not the entire holder, but I suppose this is progress.

2020-01-25
MPCNC Drag Knife Holder: Lock Screw

While calibrating the MPCNC’s probe camera offset for the drag knife holder, this happened:

Drag Knife – vertical escape

Well, at least it’s centered on the target:

Drag Knife – vertical escape – detail

This happened a few times before, because my fingers don’t fit neatly inside the drag knife holder to tighten the lock ring:

Drag Knife – LM12UU ground shaft – assembled

[Update: The lock ring keeps the holder at a fixed position inside the 12 mm shaft and doesn’t affect the blade directly. When the ring works loose, the threaded holder can rotate to expose more blade and, in this case, stab deeper into the target. ]

So I turned & knurled an aluminum ring, then tapped a 3×0.5 mm hole for a lock screw plucked from the Drawer o’ Random M3 Screws:

Drag Knife – lock screw – side

A view looking along the screw shows a bit more detail around the spring:

Drag Knife – lock screw – front

The general idea is to set the blade extension, then tighten the lock screw to hold it in place, without relying on the original brass lock ring, shown here while cutting a boss for the spring:

Drag Knife – turning spring recess

The lock screw’s knurled handle just barely kisses the NPCNC’s black tool holder ring, so my guesstimated measurements were a bit off. Clamping the knife holder one itsy higher in the tool holder solved the problem.

I cranked on 300 g of spring preload and, squashed like that, the spring’s rate is now 75 g/mm. Cutting at Z=-1 mm should suffice for laminated paper slide rule decks.

The original sizing doodle:

Drag Knife Holder – lock screw ring doodle

The short 18 mm section clears the inside of the LM12UU bearing, although it could be a millimeter shorter. The 19 mm section comes from the 3/4 inch aluminum rod I used, skim-cut to clean it up.

If I ever remake this thing, it needs a major re-think to get all the dimensions flying in formation again.

2020-01-21