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.

Category: Machine Shop

Mechanical widgetry

123 Block Links: Cap Screw Head Pins
Contemplating a project using a small saw in the Sherline suggested that attaching the workpiece to the side of a 123 block would simplify the machining. My blocks have a centered quintet of 3/8-16 tapped holes through the 2×3 side, all the remaining holes are untapped, and it has no smaller holes. The hole spacing doesn’t match the Sherline tooling plate, but the T-nut slots in the underlying table would suffice.

Rather than run long 10-32 screws through the entire block, It Would Be Nice to use short screws from, say, the nearest holes:

123 Block Links – assembled

I cannot possibly be the first person to have this idea, but the obvious keywords don’t produce any useful results on The Intertubes, other than a link to a different (and far more complex) block with counterbored holes of various sizes.

Update: Jason found a video about building those blocks and somebody else built some pins similar to mine. Nope, I’m definitely not the first person to have this idea!

Further doodling produced some useful dimensions:

123 Block Links – SHCS head pin doodle

The holes through the blocks probably came from a 5/16 inch drill, the 75% thread depth diameter for the 3/8-16 taps used on the threaded holes. They’re distorted, full of debris, and hardened enough to kill a file, so I eventually settled on 8.2 mm pins that pass through most of the holes.

The socket head screws seat at the pin axis, because the pin diameter is scary close to the counterbore diameter and I didn’t see much point in finesse. I started with a half-inch aluminum rod and peeled it to size, because it simplified the clamping and I have a bunch of them.

The pins are 3/4 inch long to leave a little space on either side of the 1 inch deep holes. I started with comfort marks along the length of the rod:

123 Block Links – laser alignment

Center-drill so the clearance drill doesn’t skitter off the top:

123 Block Links – center drilling

The counterbore calls for a 0.204 inch = #6 drill, just slightly larger than the #7 clearance drill for a 10-32 screw:

123 Block Links – counterbore

I touched off the counterbore flutes on the sides of the hole, then drilled downward half the 12.8 mm actual rod diameter:

123 Block Links – 10-32 SHCS test fit

Lower the counterbore into the hole again, relax the vise enough to let the rod slide, jog the spindle to X = -25.4 mm, and tighten the vise again:

123 Block Links – index setup

I figured I needed four pins, tops, so make half a dozen to be sure:

123 Block Links – all c-bored

Stick the rod in the mini-lathe chuck, add some comfort marks, and prepare to peel it down to 8.2 mm:

123 Block Links – lathe setup

Having done the lathe work during a Squidwrench remote meeting, I have no pictures of the process, but it goes a little something like this:
- Peel off 0.5 mm at a time, stopping just beyond the mark on the left
- Mark 3/8 inch on each side of the hole center
- Face the end
- Chamfer the rim with a file
- Clean up the body hole and counterbore
- Part the pin off a bit to the left of the mark
- Remove the rod
- Chuck the pin with the cut off end outward
- Face to the mark
- Chamfer
- Repeat for all six pins
- Done!
It’s tedious, but not particularly difficult.

Futher doodling suggested the need for threaded pins to join two blocks together.
2021-02-08
Tour Easy: Extended V-brake Noodle

Although our Tour Easy recumbents use ordinary (*) V-brakes, their frame geometry doesn’t route the rear cable quite the way the brake designers expected. Mary’s Medium-Small frame always had its rear brake cable resting against the frame tube, where it bent slightly as she applied the brakes:

Tour Easy rear V-brake layout

That’s looking up from under the rear wheel, with the bike on a workstand, and, yeah, it’s pretty grubby down there.

The squashed rubber boot suggests the brake arms are too close together, but that’s where they must be to hold the brake pads in the proper position, even with new pads and big spacer washers. As a result, the cable stop over on the right at the end of the noodle rests against the frame and dings the paint.

My first thought was to add some length to the end of the noodle inside the stirrup, so I made an adapter with the ID on the noodle end matching the OD on the fitting end:

V-brake – larger noodle – end stop adapter

Which worked poorly, because the noodle has a straight section leading up to the fitting inside the stirrup; any additional length pushes the noodle curve against the stirrup pivot and cants it out of line:

Tour Easy rear V-brake noodle

I’ve been avoiding the fallback plan of building a bigger noodle for years, but finally combined a foot of 3/32 inch brass tubing, a tube bender spring, and various large-diameter round-ish objects from the Basement Warehouse Wing:

V-brake – larger noodle – bending

I annealed the tube by running a torch along its length until the color changed to the obvious “I’m hot enough” copper color, then let it air-cool while I did something else. Brass work-hardens quickly and required two more annealings while finishing that smooth curve; as far as I know, brass doesn’t harden with the heat-and-quench cycle used for steel.

A little more lathe work produced a replacement fitting:

V-brake – larger noodle – end stop

The hole is barely one diameter deep, but I think it’ll align the tube well enough for my simple needs. The failure will most likely involve having the cable chew through the inward side of the mis-aligned tube, which should become obvious in short order.

The fitting on the OEM noodle seems to be crimped in place, but I figure my version is unlikely to fall off in normal use:

V-brake – larger noodle vs OEM

Lined up thusly, you can see the reduced straight section behind my fitting and the much larger sweep out to the cable stop.

The OEM noodle had a (presumably) PTFE liner, so I adapted a length of PTFE brake cable liner by mashing the end with various conical objects until it kinda-sorta looked like the cable stop might capture the ragged flange:

V-brake – larger noodle – PTFE liner

Reassembling in reverse order produces a comforting sight:

V-brake – larger noodle – installed

Despite appearances, the new noodle sits below the frame and well above the chain in normal use. In the most extreme small-small cross gearing position the chain barely clears it, but the takeup arm on the rear derailleur starts clattering enough to remind us not to do that.

Brass is certainly not as strong as stainless (?) steel, although I think it ended up in a reasonably hard condition after all the bending. I’m certain neither of us can squeeze the brake lever enough to come anywhere close to causing a problem.

Making a noodle was easier than I expected and, in a month or so, we’ll see how it behaves under actual riding conditions.

(*) “Ordinary” as of many decades ago, because the design dates back to the mid-70s, when Fast Freddy Markham broke 65 mph on a rather customized Easy Racers Gold Rush.

2021-02-05
LED Bulb: Mechanical FAIL

Replacing the second torchiere lamp shade required unscrewing its 100 W equivalent LED bulb, which required far too many turns and eventually felt sufficiently wrong to reveal the problem:

LED Bulb – unscrewed base

The entire metal base shell unscrewed from the plastic housing and twisted off the lead from what looks like a PTC fuse in series with the center contact; the cute little pigtail effect suggests I’ve wrecked the epoxy-to-wire seal.

It had a five year warranty which, alas, expired three years ago. This style of bulb has fallen out of favor, so I may as well get some Quality Shop Time out of it.

I don’t know how the factory machinery attached the lead to the contact button, but I’m going to go primal on it with some solder. The trick will be soldering it after assembly, so the first step is to drill through the middle of the button.

Grab it nose-down in the Sherline’s three-jaw chuck, flip it over, grab the chuck in the drill press vise, line it up, center-drill the button, then drill right through that sucker:

LED Bulb – base drilling setup

Of course, the contact came loose from the base, because I pretty much drilled right through the rivet flange holding it in place:

LED Bulb – removed center contact

Nothing a dab of epoxy can’t fix, though. I scuffed up the outside of the contact to remove the nickel (?) plating and expose the underlying brass to improve its solderability.

After the epoxy cured, align wire with hole, screw the base onto the lamp shell, and it’s ready for soldering:

LED Bulb – base ready for solder

The hole is way too large for the wire, but I wasn’t about to wreck a tiny drill on what might have been a weld nugget. In any event, the bigger the blob, the better the job:

LED Bulb – soldered base

Just like light bulb bases used to look, back in the day.

With a bit of luck, it’ll sit in that socket for another seven years.

It could happen, ya never know.

2021-02-03
Torchiere Lamp Shade 2

Three and a half years later, the shade on the living room’s other torchiere lamp crumbled at a touch:

Torchiere Lamp Shade 2 – crumbled

Because I live in the future and had solved this problem in the past, eight hours of print time produced a second shade:

Torchiere Lamp Shade 2 – on platform

I sliced the same STL file with PrusaSlicer to get G-Code incorporating whatever configuration changes I’ve made to the M2 over the years and include any slicing algorithm improvements; the OpenSCAD code remains unchanged.

The as-printed shade had pretty much the same crystalline aspect as the first one:

Torchiere Lamp Shade 2 – no epoxy

Smoothing a layer of white-tinted epoxy over the interior while spinning it slowly in the mini-lathe calmed it down enough for our simple needs, although the picture I tried to take didn’t show much difference.

That was easy …

2021-02-02
Dial Test Indicator Mount Finishing

While pondering a project requiring a slitting saw, I discovered the clamp on the dial test indicator magnetic mount I’d picked up a while ago didn’t quite fit the 5/32 inch = 4 mm stem on the indicator. The clamp ring is obviously punched from sheet, then formed into its final shape, as the holes are somewhat un-round. Running the proper drill through the holes removed a minute sliver of steel:

Dial test indicator mount – redrilling

And then it fit just fine:

Dial test indicator mount – in use

Although it looks like I’m in the process of sawing the ball off the indicator, I’m really measuring the runout, which turned out to be maybe 5 mils = 0.13 mm. The blade is likely too small for what I’m thinking of using it for, so the pondering continues.

The two bigger holes in the clamp fit the equally standard 3/8 inch = 9.5 mm stems just fine, so it’s just another one of those tools where I get to finish the last few percent of their manufacturing.

2021-01-31
Tour Easy Seat Hatchery

Removing the seat from Mary’s Tour Easy revealed an unexpected sight:

Tour Easy seat – bottom view

A closer view:

Tour Easy seat – pupal remains

An insect, most likely a rather large butterfly or moth, decided to pupate on the underside of the seat, tucked inside the old seat cover. We can’t fault the critter’s logic!

Mary is sewing up new seat covers for our Tour Easy ‘bents in preparation for the new riding season. Who knows what we’ll find under there in a few years?

2021-01-30
Photo Backdrop: Wingnut Upgrade

You’re supposed to secure the photo backdrop’s top crossbar to the uprights by fiddling with a wingnut, which you must do while reaching over your head. Emart apparently realized this operation was fraught with peril, because the package contains four wingnuts. After setting it up once, I replaced the wingnuts with finger-friendly knobs containing acorn nuts:

Photo Backdrop – thumbscrew vs printed knob

The upright pole ends in an M10×1.5 stud, which fits the biggest acorn nuts in the Warehouse Wing.

The knobs come from Thingiverse, although the OpenSCAD program required a bit of rework to make it compatible with the current version. Fiddling around with the Customizer parameters produced a Good Enough knob:

M10x1.5 Acorn Nut knob – solid model

I pulled the acorn nut into the knob using the upright pole hardware to keep it aligned. Spin the wingnut on the stud “backwards”, add the washer, push the nut slightly into the knob to get it started, then thread it onto the stud:

Photo Backdrop – knob nut seating – 1

Turn the knob to pull the nut inward until the stud hits the inside of the nut:

Photo Backdrop – knob nut seating – 2

Unthread the nut a bit, run the wingnut out to meet the bottom of the knob, and repeat the operation until the nut bottoms out inside the knob:

Photo Backdrop – knob nut seated

Toss the wingnuts into the Warehouse Wing against later use.

Bonus project: on the other end of the upright, you’ll find it impossible to actually lock the leg carrier against the pole:

Photo Backdrop – tripod leg lock

The plastic fitting is … generously … sized around the 25 mm OD upright pole and requires more compression than I could produce with my puny fingers. It turns out the 18 mm OD leg tube exactly fills the space available inside the fitting, so you (well, I) must squash the steel tube in order to close the fitting on the pole.

Remove the wingnut + screw to free the end of the leg, stick an inch of the leg into the bench vise’s soft jaws, and mash gently to about 16 mm across the holes; it’ll expand slightly in the other direction. Reassemble in reverse order and discover the thumbscrew now squeezes the fitting exactly as it should.

There might be more finishing to do when we actually hang a quilt from the stand, but at least it’s now usable.

Sheesh and similar remarks.

2021-01-29