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: Sherline

Sherline CNC mill

KG-UV3D GPS+Voice Interface: APRS Bicycle Mobile

Wouxun KG-UV3D with GPS-audio interface

Both of the GPS+voice interfaces for the Wouxun KG-UV3D radios have been working fine for a while, so I should show the whole installation in all its gory detail.

If you haven’t been following the story, the Big Idea boils down to an amateur radio HT wearing a backpack that replaces its battery, combines the audio output of a Byonics TinyTrak3+ GPS encoder with our voice audio for transmission, and routes received audio to an earbud. Setting the radios to the APRS standard frequency (144.39 MHz) routes our GPS position points to the global packet database and, with 100 Hz tone squelch, we can use the radios as tactical intercoms without listening to ~~all~~ much of the data traffic.

The local APRS network wizards approved our use of voice on the data channel, seeing as how we’re transmitting brief voice messages using low power through bad antennas from generally terrible locations. This wouldn’t work well in a dense urban environment with more APRS traffic; you’d need one of the newfangled radios that can switch frequencies for packet and voice transmissions.

So, with that in mind, making it work required a lot of parts…

Tour Easy – KG-UV3D GPS interface

A water bottle holder attaches to the seat base rail with a machined circumferential clamp. Inside the holder, a bike seat wedge pack contains the radio with its GPS+voice interface box and provides a bit of cushioning; a chunk of closed-cell foam on the bottom mostly makes me feel good.

The flat 5 A·h Li-ion battery pack on the rack provides power for the radio; it’s intended for a DVD player and has a 9 V output that’s a trifle hot for the Wouxun radios. Some Genuine Velcro self-adhesive strips hold the packs to the racks and have survived surprisingly well.

Just out of the picture to the left of the battery pack sits a Byonics GPS2 receiver puck atop a fender washer glued to the rack, with a black serial cable passing across the rack and down to the radio bag.

A dual-band mobile antenna screws into the homebrew mount attached to the upper seat rail with another circumferential clamp. It’s on the left side of the rail, just barely out of the way of our helmets, and, yes, the radiating section of the antenna sits too close to our heads. The overly long coax cable has its excess coiled and strapped to the front of the rack; I pretend that’s an inductor to choke RF off the shield braid. The cable terminates in a PL-259 UHF plug, with an adapter to the radio’s reverse-polarity SMA socket.

The push-to-talk button on the left handgrip isn’t quite visible in the picture. That cable runs down the handlebar, along the upper frame tube, under the seat, and emerges just in front of the radio bag, where it terminates in a 3.5 mm audio plug.

The white USB cable from the helmet carries the boom mic and earbud audio over the top of the seat, knots around the top frame bar, and continues down to the radio. USB cables aren’t intended for this service and fail every few years, but they’re cheap and work well enough. The USB connector separates easily, which prevents us from being firmly secured to a dropped bike during a crash. I’d like much more supple cables, a trait that’s simply not in the USB cable repertoire. This is not a digital USB connection: I’m just using a cheap & readily available cable.

All cables converge on the bag holding the radio:

Tour Easy – KG-UV3D + GPS interface – detail

Now you can see why I put that dab of white on the top of the knob!

The bag on my bike hasn’t accumulated quite so much crud, because it’s only a few months old, but it’s just as crowded:

KG-UV3D + GPS interface on Tour Easy – top view

This whole “bicycle mobile APRS system”, to abuse a term, slowly grew from a voice-only interface for our ICOM IC-Z1A radios. Improving (and replacing!) one piece at a time occasionally produced horrible compatibility problems, while showing why commercial solutions justify owning metalworking tools, PCB design software, and a 3D printer.

I long ago lost track of the number of Quality Shop Time hours devoted to all this, which may be the whole point…

In other news, the 3D-printed fairing mounts, blinky light mounts, and helmet mirror mounts continue to work fine; I’m absurdly proud of the mirrors. Mary likes her colorful homebrew seat cover that replaced a worn-out black OEM cover for a minute fraction of the price.

2012-09-11
Wouxun KG-UV3D Plug Plate

Based on those measurements that suggest spacing the plugs at 11.5 mm on center, I tweaked that parameter in the source code there and printed another one, just like the other one. Actually, I printed four of the fool things this time:

Wouxun plug plates – 11.5 mm fixture

With the plugs in the gluing fixture and the fixture in the vise, a ring of epoxy around the threaded sides holds them in place:

Wouxun plug plate – wired

A trial fit in the Wouxun KG-UV3D shows that the jacks prefer the 11.2 mm spacing I measured on the Wouxun headset, but they’ll accept plugs on 11.5 mm centers. I don’t know if that’s a real specification difference, a manufacturing tolerance, or what.

FWIW, I’ve been using snippets of that cable forever, because it’s perfect for this application: two unshielded conductors and three more inside a braid, supple as a snake. It’s surplus, of course, with a gorgeous push-lock plug (and the jack!) on one end that must have cost a fortune… and which I’ll never to use for anything. Got two of them, just in case.

Mushing an epoxy putty turd on the top anchors everything in place and protects the wires:

Wouxun plug plate – epoxy cap

In point of fact, the cable insulation isn’t anchored inside the blob and a minor tug could pull it loose. There will be a bit of slack at the case to allow for unlatching it from the radio, but the lashup will spend its entire life inside a snug pouch, so it shouldn’t come to any harm. We shall see.

2012-08-19
Kenwood / Wouxun Headset Jack Spacing

Wouxun plug mounting plate – overview

Try as I might, I cannot uncover a definitive answer to this simple question: What’s the center-to-center spacing of the mic and earphone jacks on the side of Kenwood and Wouxun HTs?

The usual searches produce answers like 11 and 12 mm, both of which are obviously wrong, as can be determined eyeballometrically just by holding a scale against the plugs.

Based on measurements I made on a Wouxun headset, the yellow plug mounting plate put the plugs on 11.2 mm centers and they fit into the KG-UV3D radio; it’s been working fine ever since.

However, having just measured a speaker/mic and a headset, both from Kenwood, I come up with 11.5 mm. Frankly, I trust the Kenwood hardware a bit more: the plugs seem more rugged and the overall production values are higher.

The calculation is simple: measure the pin diameters, then subtract half their sum from the outside distance across the pins. Cross-check by adding half the sum to the inside distance between the pins, which should give the same answer. It helps if the pins are actually round.

The jacks in the Kenwood and Wouxun radios have enough compliance to accept either a Wouxun or a Kenwood headset plug without complaint. Maybe it doesn’t matter?

Despite that, I made another gluing fixture with 11.5 mm spacing:

Plug alignment plate – 11.5 mm spacing

Those are 0.1 inch grids; it’s a little bitty block of smoke-gray polycarbonate from the scrap heap. The plugs are nominally 3.5 mm (which is not 1/8 inch in this universe) and 2.5 mm, with clearance drills #28 and #39.

Then I tried poking those 11.2 mm spaced plugs, now firmly epoxied in place in the yellow plate, and guess what: they don’t fit, no how no way. That’s not surprising, because there’s no compliance on either side of the joint and the plugs aren’t on the right centers for the fixture. Makes for a good No-Go gauge, I suppose.

However, I think I’ll tweak the solid model spacing to 11.5 mm and run off another plug mounting plate for the next radio.

FWIW, our ICOM IC-Z1A HTs use a sensible 10.0 mm spacing and that old fixture worked fine.

2012-08-17
HT GPS+Voice: Battery Contacts

For this version of the contacts (the old version is there) that make the GPS interface look like a standard Wouxun lithium battery, I left a bit more of the slot on the brass screw heads and increased the recess depth to compensate:

HT-GPS Case – Battery contacts

The nuts all have fancy nickel plating, with washers & ring lugs silver-soldered in place:

HT-GPS PCB – battery contact parts

The trial fit looks OK:

HT-GPS Case – PCB and battery contacts – end view

I even found the cutest little flat 1/4 inch wrench that fits 4-40 nuts, so I can do a better job of crunching the PCB between the nuts. That excess screw length has got to go, too…

2012-08-16
HT GPS+Voice Circuitry: Bare PCB

Drilling the PCB went fine, as did the etching & silver plating:

PCB with edge wrap – front

The rear side has a fine ground plane:

PCB with edge wrap – rear

The small spots scattered over the rear mark vias that stitch the front and back planes together; lacking plated-through holes, I solder nippets of 24 AWG wires to both sides. The wrinkly edge comes from solder on the copper foil binding the entire perimeter.

While I have no hard evidence that all of the fuss & bother matters, the most recent version of this circuit is the quietest yet: the machine noise from the TinyTrak3+ that plagued the first iteration has pretty much vanished.

I’ll grant you that the silver plating doesn’t look very silvery in these pix, but it’s quite different from the bare copper in person. Here’s the front just after rubbing it in with a vigorous circular motion:

HT-GPS PCB – raw plated – top

2012-08-14
Storing Sherline CNC Mill Leadscrews

After replacing the Y axis leadscrew, I decided that the X axis leadscrew was in fine shape, because it’s tucked under the table and not exposed to the swarf and grit that fell on the Y axis screw before I installed the bellows. Being that sort of bear, I couldn’t throw out the worn Y axis leadscrew, so I had two rather delicate rods that really needed more protection than a twist of paper.

So I sawed off a length of 1 inch PVC pipe, faced the ends in the lathe, and added two rubbery endcaps from the heap:

Sherline leadscrews stored in PVC pipe

That fits neatly into the big box alongside the rotary table, with the bag of assorted nuts so they’re all together.

Despite what you see there, the screws are wrapped in paper with a bit of oil, so it’s all good.

2012-08-06
Harbor Freight Slitting Saw Arbor
A three-pack of 100-tooth 2 inch cutoff saw blades followed me home from Harbor Freight a while ago. Although they’re intended for a craptastic HF tabletop saw, I thought they might come in handy on the Sherline for slicing lengths of brass tubing. The reviews for the saw indicate the blades are no good for steel, barely adequate for brass, and dandy for wood; they have nowhere near enough teeth for a screw cutoff blade.

None of the arbors in my collection fit a blade with a 3/8 inch hole, so a bit of lathe work produced one while the 3D printer cranked out a GPS+audio case:

Cutoff saw arbor in Sherline toolholder

The shaft is 3/8 inch drill rod and the collars are 3/4 inch drill rod, both of O1 oil-hardening steel that will remain forever unhardened, fitting into a Sherline endmill toolholder. I drilled-and-bored the collars to a slip fit on the shaft, then epoxied the rear one in place:

img_2156 – Cutoff saw arbor – parts

I drilled a 0.6 inch deep blind hole in the shaft and tapped it 10-32 all the way down for a 1/2 inch SHCS. A bag of assorted 10-32 taps produced a bottoming tap that came in handy, but I put tapping in the same category as parallel parking: I’ll walk half a mile to not parallel park the van. Couldn’t avoid it this time.

The flat on the shaft came from a bit of hand filing, which was easier than setting up the mill.

The front collar’s undercut ensures just the rim contacts the blade. The photo shows the vanishingly thin layer of epoxy on the rear collar that mooshed out as I clamped the stack together:
- Fixed (rear) collar
- Waxed paper with a 3/8 inch hole punched in the middle
- Cutoff blade
- Split lockwasher for a bit of space
- Loose (front) collar
- Socket head cap screw
After the epoxy cured, a pass through the lathe skimmed off that thin epoxy layer and trued up the fixed collar face to eliminate the last bit of wobble. The radial runout remains just enough so that one tooth tings before the others engage, but I’m not entirely convinced that’s due to the (minimal) shaft-to-blade clearance.

In use, putting the split lockwasher between the loose collar and the SHCS provides a little clamping compliance.

At some point, I’m sure this thing will come in handy…
2012-08-03