Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.
A few days after epoxying a replacement WS2812 RGB LED into the base of the 21HB5A and, en passant, soldering a 3.5 mm plug-and-jack into the plate lead for EZ removal, the top LED failed.
21HB5A – Audio plug cable
In this case, it also failed the Josh Sharpie test with bad encapsulation sealing:
WS2812 LED failure – ink test patterns
Here’s a view from another angle, with a warm-white desk lamp for a bit of color:
WS2812 LED failure – ink test patterns – 2
Those patterns took a few days to appear and also showed up in some, but not all, of the previous failing LEDs.
Although I have no idea what’s going on, it’s certainly distinctive!
An envelope of RGBW LEDs, allegedly with SK2812 controllers, has arrived from a different eBay supplier, so it’s time for an upgrade.
After months of attempts and (occasionally) spectacular failures, one of the backyard squirrels managed to climb aboard the bird feeder:
Squirrel on bird feeder
The shutter closes when more than two cardinals and a titmouse perch on the wood bar, so the squirrel didn’t get anything. However, back in 2008, one of that critter’s ancestors mastered the trick:
Not a Squirrel-Proof Feeder
Since then, I’ve raised the feeder about five feet and inverted a big pot over two feet of loose PVC pipe around the pole.
Given the number of squirrel-training videos on Youtube, however, it’s only a matter of time until the critters put all the tricks together!
This big branch must have landed with a mighty thump across the Maloney Road entrance to the Dutchess Rail Trail:
DCRT – Maloney Rd – tree down – 2017-02-20
Yeah, some jerk ran a snowmobile up the slope around the tree, leaving a pile of dirt on the ramp. So it goes.
We took an alternate route, I emailed The Right Folks, and (most of) the tree vanished two days later; evidently, the property owner gets to deal with everything to the left of the line of trees.
The really bright LED worklights I added to the MicroMark bandsaw produced plenty of glare from the raw aluminum table top:
USB Gooseneck Mount – on bandsaw
No good deed goes unpunished, I suppose.
While rooting around for something else, I rediscovered my bottle of Birchwood Casey Aluminum Black (basically selenium dioxide) that’s intended for touchup work on small parts, not blackening an entire aluminum plate. Well, having had that bottle forever, it’s not like I’ll miss a few milliliters.
If this didn’t work, I could always sand the table down to the original aluminum finish.
So I applied a sanding block in hopes of smoothing the tooling marks:
MicroMark Bandsaw – sanded table
Looked pretty good, I thought, so:
Wipe it down with alcohol (per the bottle instructions)
Well, in terms of metal finishing, that blackening job looks downright crappy. Aluminium Black is intended for decorative work and will surely wear quickly on the bandsaw table, but it’s entirely good enough for my simple needs: the glare from those lights is gone.
After I took the picture, I blackened the brass screw in the slot. Came out a weird mottled green-bronze, might look antique in a different context, suits me just fine.
Now I know the Forester’s TPMS icon blinks on 1000 feet from a cold start with 12 psi in the offending tire. I returned home and pulled this from a sipe in the left rear tire:
Road debris – blade fragment
It’s atop a 0.1 inch grid.
The flat side on the right rode tangent to the tire surface, recessed slightly below the tread, and pretty much invisible inside the sipe. Of course, the point punched through the tire’s steel belt and let the wind out, ever so slowly.
I initially thought it was a utility knife blade fragment, but under the microscope it looks more like a saw blade tooth. It’s obviously been kicking around on the road for quite a while; back in the day, they occasionally swept the roads, but that was then and this is now.
Makes me glad I didn’t buy four new tires after the last flat. I suppose installing two plugs in the same tire counts as a net loss, but they’re small, widely separated injuries and that’s how it’ll roll.
For the record: with 14 k miles on the tires, tread wear = 2/32 inch of the original 6/32 inch depth.
Those tires should last another 30 k miles at our current pace, although I expect more random debris will kill one stone cold dead before that.
Five bucks delivered three sets of five warm-white LED filaments from halfway around the planet:
LED Filaments – 3×5 sets
Unfortunately, the “Top Rated Plus” eBay seller just popped three ziplock baggies into an unpadded envelope and tossed it in the mail:
Unpadded LED Filament Envelope
Which had pretty much the result you’d expect on the glass substrates within:
Broken LED Filament 1
Turns out every single filament had at least one break:
Broken LED Filament 3
Indeed, some seemed just as flexy as the silicone cylinder surrounding the pulverized substrate.
I reported this to the seller, with photographs, and got a classic response:
can you use?
No, I cannot imagine a use for broken LED filaments.
The seller proposed shipping replacements that would might arrive just after the eBay feedback window closed. I proposed refunding the five bucks. The seller ignored that and sent the replacements in an untracked package “as it is an economical shipping, we have to reduce our loss, so is it ok?”.
No, it’s not, but he / she / it didn’t actually intend that as a question.
Were the filaments intact, they’d pass 15 mA with 50 to 60 V applied in one direction or the other, for 1 W average dissipation. That’s probably too high for prolonged use in air (spendy bulbs with similar LEDs have argon / krypton fill for better heat transfer), but I can surely throttle them back a bit.
Perhaps the replacements will arrive before the feedback window closes?
I did order another batch from a different seller that might arrive intact before then. We shall see…
Starting the show takes 17 seconds from clicking the Restart button (second from right, top row) to APL’s Clear WS prompt. I have no idea how that compares with a Genuine IBM 5100.
I distinctly remember writing APL programs, but that’s about as far as my memory will take me. [sigh]
The Smell of Molten Projects in the Morning 