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

Sears Sewing Table: Foot Repair

While arranging the venerable Sears Sewing Table in its new abode, we found the casters underneath didn’t roll nearly as well as they should, which turned out to be due to an accumulation of damage:

Sears sewing table – torn MDF

As far as I can tell, all four casters have been displaced upward, probably because they have no support directly above their stems and any force applied to the wheel has plenty of lever arm against those screws.

The MDF panels on the outside of the table have pictures of wood laminated to their surface, but lack enough structural integrity to keep the screws in place. The plywood, however, survived largely unscathed, although the screws were pulling out.

I poked as much wood glue into the gaps as possible, then applied as many clamps as possible, with wood strips on both sides of the bulge squashing the MDF into a flat sheet. Over the course of two gluing sessions (I need more clamps!) spanning three days, while Mary really wanted to start sewing, the glue cured. I had plenty of time to unbend the brackets and put a more-or-less right angle between their two screw plates.

Rummaging in the box of laser scraps (after finding said box) produced disks cut from various projects that fit between the plywood bottom of the cabinet and the stems:

Sears sewing table – repaired foot – side

The brackets deliberately don’t match their original shape, because their new squareness put the screws into undamaged spots in the MDF and plywood:

Sears sewing table – repaired foot – bottom

The MDF will never be quite the same, but it’s flat on the visible side and the glue (seems to have) consolidated the fragments well enough.

Although those wheels look terrible, the bracket now holds the stem vertically and all four of them roll easily and pivot smoothly.

The laser-cut disks are held in place by pure faith and the overwhelming weight of all the MDF in the table, so they’re not going anywhere. Because the table’s weight now rests on the caster stems, as distributed across the plywood cabinet bottom through the disks, the brackets shouldn’t be subject to excessive upward force.

She started therapy right away:

Sears sewing table in use

And we continued moving in …

2024-04-08
SJCam M50 Trail Camera: Battery Wire FAIL

My SJCam M50 Trail Camera has had its share of problems, including water making it past the seals to corrode some connections:

M50 Trail Cam – contact corrosion

I thought cleaning that mess up would solve an intermittent power problem, but the camera continued to fail immediately after being deployed and finally refused to work at all.

The camera case has eight (!) AA cells in one half connected to the electronics in the other half by a pair of wires that pass through the hinge between the halves:

M50 Trail Cam – pivot wire route

The steel rod is the hinge pivot, with the battery half wearing brown and the electronics half in lighter plastic. As you’ll see in a bit, the rod is fixed in the electronics half and the battery half pivots around it.

The two short case sections on the right contain the two wires carrying the 6 V battery power. Some gentle manipulation suggested the fault lay inside those hinge sections, which meant I had to figure out how to get them apart.

The other end of the steel rod has a knurled section jammed firmly into the electronics half, but I managed to carve away just enough plastic to expose just enough of the knurl to get just enough of a grip (yes, with a pair of genuine Vise-Grip 10WR Locking Pliers, accept no substitutes) to yoink the rod out:

M50 Trail Cam – extracted pivot

With the hinge released, the problem became immediately obvious:

M50 Trail Cam – failed hinge wires

Yes, those are wire strands poking out of the hole in the left hinge section.

A tedious needle-nose tweezer session extracted the remains of the wires from the hinge and cleaned out the adhesive:

M50 Trail Cam – extracted OEM PVC wires

Although those two hinge sections are hollow with plenty of room for the wire, it seems the assembler squirted adhesive into both sections to glue the wires in place. As a result, every time I opened the case to charge the batteries, maybe two millimeters of wire twisted 180° degrees. The wonder is that it lasted as long as it did.

I snaked a pair of 20 AWG silicone-insulated wires through the hinge sections:

M50 Trail Cam – silicone rewiring

The OEM wires had PVC insulation, which is a terrible choice for wires that will undergo lots of flexing, but that’s what SJCam used.

Two untidy blobs of acrylic caulk do at least as good a job of sealing the case openings as the black gunk visible in the earlier pictures:

M50 Trail Cam – new caulk

I left all of the wire in the hinge un-stuck, hoping the twist will distribute itself over maybe 5 mm of wire and last longer.

In anticipation of future repairs, however, I left enough of the knurled end of the hinge rod exposed to get an easy grip:

M50 Trail Cam – restaked pivot

Solder the new wires to the old pads, assemble in reverse order, and it works as well as it ever did:

The alert reader will note I did not reset the camera clock after charging the batteries, a process requiring the janky SJCam app.

The two finches on the right have been constructing a nest in the wreath hanging at our front door. They tolerate our presence, although they’d be happier if delivery folks dropped packages elsewhere.

2024-04-05
Bobbin Rock: Bigger and Tighter

While loading bobbins for her next quilt project, Mary found another one that just wouldn’t fit on the winder:

Bobbin with polishing rock

Knocking the rock out of the bore required a pin punch and more than a gentle tap, but it eventually left without damage. The little stick in the slot looked organic, although it vanished without a trace during the operation.

I originally thought the bobbin factory’s final vibratory polisher used walnut shells, but the evidence certainly suggests ordinary gravel!

2023-11-27
Kitchen Sink O-ring Seat Polishers

My long-running battle with the American Standard kitchen sink O-rings continues, but this time I tried polishing the seats with a strip of 3M 2000 grit sandpaper:

Sink O-ring seat polisher – overview

Although it’s technically sandpaper, the effect is more like lapping than sanding and the O-rings now ride on a very smooth surface.

The knurled half-ring is ¼-inch = 6.3 mm acrylic with an ID precisely fitting the pillar + sandpaper:

Sink O-ring seat polishers

The one on the right has an OD matching the surface inside the spout, but it turned out to be easier using fingertips, even if that isn’t quite how one should do lapping.

The LightBurn layout shows the “knurls” are half-circles either added or subtracted from the arcs, as LightBurn’s Circular Array tool is my copilot:

O-ring Polishers – LB layout

You’ll want to measure the ID and OD of your sink faucet, as well as the thickness of your sandpaper, before making make your own.

Imagining / laying out / building those took less time than writing this up; I loves me some quick laser cutter action.

2023-11-24
Laser Cutter Camera: USB Mystery
The USB camera stuck inside the lid of the laser worked fine:

OMTech Laser – camera mount

Until a month or two ago, when it began disconnecting randomly.

The camera cable has a standard USB A connector on one end and what looks like a 1.5 mm JST ZH connector on the other:

Laser cutter camera cable

Of course, it’s not quite long enough, so it plugs into a good-quality 1 meter USB 3.0 extender to the PC sitting atop the laser cabinet.

Some low-effort tweaks were unavailing:
- Different USB ports
- Different USB extension cable to the ports
- Hub vs. direct
Eventually, some rummaging in the Box o’ USB Cables produced a cable from a different camera and, as you might expect, swapping the two identical cables solved the problem.

I have no idea what’s going on, but I’d lay significant money that when this cable gets flaky, swapping the original cable back in will solve the problem once again.
2023-11-03
Nuheara IQbuds² MAX Battery Replacement
Nuheara predicts two to three years of battery lifetime for their IQbuds² MAX not-really-hearing-aids and, indeed, after 2-½ years of more-or-less steady use, the right bud developed a bad case of not charging fully and discharging quickly. The batteries are not, of course, customer-replaceable, so one can:
- Buy a single bud
- Buy a complete new pair + case + accessories
- Ask about their repair service
Unsurprisingly, a single bud costs more than half the cost of the full set and the repair service is a complete mystery. Given that the left bud’s battery will likely fail in short order, let’s find out what’s inside.

Your ear sees this side:

Nuheara IQbud – bottom view

The dark oval is a (probably IR) sensor telling the bud when it’s jammed in your ear.

Everybody else sees this side:

Nuheara IQbud – top view

The small slit over on the right and the two holes around the top seem to be for various microphones.

Jamming a plastic razor blade into the junction between the two parts of the case, just under the mic slit, and gently prying around the perimeter eventually forces the adhesive apart:

Nuheara IQbud – case splitting

Do not attempt to yank the two pieces apart, because a ribbon cable joins the lower and upper PCBs:

Nuheara IQbud – ribbon cable

The metallic disk in the lower part is the lithium battery.

Ease the upper part away, being very careful about not tugging on the ribbon cable:

Nuheara IQbud – raising battery

The battery has moved upward, revealing the lower PCB.

Rolling the upper part toward the ribbon cable eventually produces enough space to extract the battery:

Nuheara IQbud – battery freed

Note the orientation:
- The rebated end is the negative terminal and faces outward
- The wider end is the positive terminal and faces inward
With the battery out, you can admire the PCBs and ribbon cable:

Nuheara IQbud – interior view

What is not obvious from the picture: two pairs of spring-loaded pogo pins contacting the battery. There is no actual battery holder, as it’s just tucked into the structure of the bud, with the perimeter adhesive providing the restraining force for the pogo pins.

The battery seems a variant of a standard 1654-size lithium cell:

Nuheara IQbud – OEM ZJ1654A lithium cell

The 1654 cells I got came with wire leads welded to the cell and a complete Kapton enclosure; apparently other devices use soldered connections rather than pins. They proudly proclaim their “Varta” heritage, but I have no way to prove they actually came from Germany.

I snipped off the wires, carved a pair of holes through their Kapton for the contact pins, tucked the cell in the bud, pressed the halves together, applied a clamp, then wrapped a strip of Kapton tape around the perimeter:

Nuheara IQbud – reassembled

It seems remarkably easy to wrap the tape over the front microphone, but don’t do that. Conversely, sealing the entire perimeter is the only way to prevent acoustic feedback, so I added a snippet of tape just under the front mic opening.

Do that for the other bud and declare victory.

That is, fer shure, not the most stylin’ repair you’ve ever seen, but I was (for what should be obvious reasons) reluctant to glue the halves together. I expect the tape to peel off / lose traction after a while, but I have plenty of tape at the ready. Worst case, I can glop some adhesive in there and hope for the best.

Because the buds lost power during their adventure, they required a trip through their charging case to wake them up again. After that, they work as well as they did before, with consistently longer run time from both buds.

Whew!
2023-10-10
Moonlander Keyboard vs. Board Chow

The Moonlander keyboard has per-key LEDs that I’ve denatured enough that most show a pale gray, with a few others highlighted in orange. A few weeks ago the LEDs on the right-hand thumb cluster and the N key went nuts, cycling through a surprising assortment before settling on bright red; the obvious resets / firmware reflashing / tapping were all unavailing.

ZSA’s tech support recommended taking the thumb cluster apart to check the ribbon cable connecting it to the main keyboard half:

Moonlander thumb cluster – PCB bottom

Come to find out my unclean personal habits lodged a particularly corrosive nugget of board chow on the cable:

Moonlander – corroded ribbon cable

It’s a more-or-less standard 0.5 mm pitch cable, but only 20-ish mm long, much shorter than the cables carried by the usual sources. ZSA sells them for $2 each, plus $25 courier shipping, so I bought three; they arrived in two days from halfway around the planet.

Because I don’t foresee my personal habits changing any time soon, I tucked a Kapton tape snippet in the gap to serve as a gutter:

Moonlander thumb cluster – tape shield installation

That’s with the two hinge screws out and the cluster eased down-and-away from the keyboard enough to get the tape pressed against the keyboard.

With the screws installed and the cluster at its normal most-downward angle, the gutter closes up:

Moonlander thumb cluster – tape shield folded

With the cluster in its normal operating position (for me, anyway), the gutter is nearly invisible:

Moonlander thumb cluster – normal position

For the record, I tucked the remaining ribbon cables inside the left-hand thumb cluster against future need.

2023-10-09