Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.
One of the ribs in the six-passenger umbrella we keep in the van snagged on something and snapped its fitting on the spreader strut:
Umbrella strut – broken connector
This being wonderful engineering plastic that cannot be solvent-bonded, epoxy is the only adhesive that will work. However, those joints undergo tremendous stress in a deployed umbrella, so a bare epoxy joint won’t have enough strength for the job. What to do?
Wonder of wonders, when I got the umbrella into the Basement Laboratory Repair Wing, I discovered:
The not-quite-round strut fitting stub slipped right into a short brass tube from the heap and
Just enough of the fitting remained on the rib to anchor the tubing
A silicone tape wrap kept most of the epoxy inside while it cured:
Umbrella strut – epoxy curing
Clearing off a few blobs made it all good:
Umbrella strut – brass tubing splint
We don’t play golf, but such a big umbrella keeps most of the rain off two people; it’s a tchotchke from back when Mary worked at IBM (hence the color scheme). We call it our “six-passenger” umbrella because it looks about that big when we deploy it…
The bikes stand upright inside the van and the helmets ride on the floor with all their stalks sticking up. This usually works out well, but on our last trip my helmet rolled under my bike and rubbed the foam ball surrounding its mic against the chain, producing a result so awful that I had to install new foam.
The foam comes from a sheet of Sonex acoustic foam baffle, snipped into a reasonable approximation of a ball, with a slit deep enough to surround the mic, and a cable tie holding it closed:
Foam mic ball on bike helmet boom
For what it’s worth, I’ve found that excessive wind noise correlates with too much mic gain. The mic rides about a finger’s width from the corner of my mouth, I talk at a normal volume, the amp supplies about 20 dB of gain, and we have no trouble with wind noise. The amp gain depends on the mic sensitivity, so your results will certainly differ; these mics came from the heap with no specs whatsoever.
I suppose wind noise also depends on the bike’s speed, but when I’m going that fast I don’t have enough brain or lungs left over to hold a conversation…
The absurdly heavy pump nozzle atop the kitchen sink soap dispenser seemed more wobbly than usual. Some investigation suggested the fitting atop the plastic pump has gotten smaller, which may be due to having the nozzle wobble around on it.
In any event, a wrap of Kapton tape snugged it up just fine:
Sink soap dispenser pump
We’ll see how long that lasts; this thing may be nearing the end of its useful life.
Our Toyota Sienna arrived with a blank cover plate where a fancier model would have a switch. It seemed a shame to let that space go to waste, so I popped the plate out, rummaged around in the heap, found a small circuit board with a blinky LED that just exactly fit the space available, and drilled a suitable hole:
Sienna anti-theft blinker – inside
When it’s installed in the van, it looks and acts just like the security system we don’t have. For all I know, that plate was for the security system control, so perhaps it’s an exact match!
Sienna anti-theft blinker – bezel
The batteries last about two years, a few months later I notice the lack of blinkiness (it’s hidden behind the steering wheel in my normal driving position), and eventually I replace the corroded batteries. This time, I had to replace the entire battery holder; things got pretty nasty in there.
As I recall, the PCB came from a fancy “greeting card” mailed to me by the Business Software Alliance, with the implied threat that if all my paperwork wasn’t up to par, my use of potentially unlicensed software would blow up in my face. That was back in the day when mailing something that pretended to be a bomb was considered a cute joke and when I actually ran more than one Windows PC.
This little utility knife lives in my belt pack for those occasions when I need a sharper blade than my long-suffering Swiss Army Knife can provide. Alas, I must have sliced up something awful in the recent past:
Corroded mini utility knife blade
Although the knife frame is slightly magnetic, it’s also reasonably noncorrosive and cleaned up pretty well after some Topsaver and wire-wheel brushing activity:
Mini utility knife – somewhat cleaned
Whew!
For whatever it’s worth, this is the same knife that made it through the Washington DC “security scans” that eventually confiscated Mary’s Swiss Army Knife. She didn’t look any more guilty than I did…
It seems that Wouxun KG-UV3D HTs require nearly 0 V to activate the PTT input, which I discovered after the radio on Mary’s bike began acting intermittently. The TinyTrak3+ would transmit correctly, but the PTT button on the handlebar began to not work at all / work intermittently / work perfectly. The switch and cable were OK, pushing the button produced nearly 0 Ω at the 3.5 mm plug, the connections seemed solid, but the radio didn’t transmit reliably.
I finally got the thing to fail on the bench, which led to the discovery that:
Shorting the PTT input to the GPS+voice adapter PCB to ground didn’t make the radio transmit and
Data bursts from the TinyTrack3 worked perfectly
Gotcha!
TT3 PTT In-Out
The TT3+ pulls its PTT OUT pin down from +5 V using a 2N2222A NPN transistor (off to the right in the schematic snippet), but, for reasons having to do with ESD, the input from the PTT switch on the handlebars goes through a 100 Ω series resistor, then passes to the TT3 board through PTT IN to D6 before joining the TT3 transistor collector. The low-active diode-ORed signal heads off through PTT OUT to a 10 Ω series resistor, thence to the KG-UV3D PTT input. D6 is an ordinary 1N4148, with the net result that the PTT input voltage at the radio dropped to 630 mV with the PTT button pressed.
Not finding anything else wrong, I replaced D6 with a BAT54 Schottky diode that pulled the PTT voltage down to 300 mV and the radio worked fine.
Of course, a BAT54 is a surface-mount diode, so I clipped off the unused no-connect lead (it’s the only way to be sure it doesn’t do anything) and tacked it down slaunchwise between the PCB thru-hole pads. If I had a BAT54C with common cathodes, I could replace both D5 and D6 in one shot, but D5 just pulls down a PIC input that has an ordinary logic-level threshold voltage.
I don’t know why the KG-UV3D PTT is so fussy, although it may really be a current-driven signal that requires more current than can flow through the 110 Ω + diode forward drop in series with the PTT button. Wouxun presents no specifications that I can find.
The identical circuitry on my bike works fine with the stock D6 diode and a presumably identical KG-UV3D. I should replace that diode before it gives me any trouble, but I’ll wait until I must take the box apart for some other reason.
The slop sink in the rental house developed a drip and, unlike our kitchen faucet, required only a new washer. Of course, choosing the right size from that assortment posed a bit of a problem:
Slop sink valve with washers
The old washer is in the upper right; you can see the indentation from the valve seat.
There’s a variety of sizes & shapes; these represent just the closest matches. I have no idea what 3/8, 3/8R, and 3/8L might signify, but they’re all slightly different, some with conical cross-sections that may also be slightly different. Worst case, of course, you can sand down the rim of a too-large washer to make the diameter come out right.
The washer just in front of the old one has information molded right into the back: GOLDEN STATE 10¢ 1/2. Now there’s a show of confidence in price stability that you don’t see much any more!
I found one that fit snugly in the recess of the valve stem, turned the screw tight, and it’s all good.
The Smell of Molten Projects in the Morning 