The plastic-ball-in-plastic-socket joint found in bicycle mirrors seems to fail after a year or two of constant use. These are some doodles & thoughts about building a small, robust, adjustable joint.
A bike mirror needs two ball joints:
- at the helmet mount to put the mirror in the proper spot
- at the mirror to align the image
A flexy boom can replace the helmet joint, although rotation around X (pitch) is still handy.
A flexy mirror mount can replace the mirror joint, but it must also be compact.
Without heroic measures, the range of travel for a ball joint isn’t all that much.
How to make a ball? Anneal & drill a standard ball bearing for a wire shaft? Solder onto chrome steel? CNC mill the end of a bar in a rotary table?
How to make a socket? Some of that low-temperature themoplastic might be useful. Mold it around the ball, slit radially, and squash it in a circ clamp?
How to adjust? Circumferential clamp around the socket or pull the whole socket into a wedge? Radial cuts through the socket to allow compression or depend on plastic/elastic deformation?
How much friction? You want it stiff enough to hold position in a strong wind and easy enough to reposition. You definitely don’t want grub screws or fiddly knobs!
The doodles are all far too complex, some are absurd, one can’t be built (at least by me), and I’ll probably end up using some bendy wire anyway.
Something of this may be useful in another project … and now I can throw out that scrap of paper.
The Smell of Molten Projects in the Morning 
If I had to make a socket for a ball-and-socket, I’d make it in two halves to make the machining easier, and have them thread together with a locknut to hold them at the appropriate tightness.
But a bendy wire seems so much more robust. Aviation headset mic booms all used to be ball-and-socket and they *sucked*. They’ve all moved to bendy flexible solutions save for the very cheapest ones.
used to be ball-and-socket and they *sucked*
That’s pretty much the conclusion I came to.
I made the mic booms from 12 AWG wire and they work reasonably well, so I’ll likely do something just like that for the mirrors. Somewhere in the heap I have some non-work-hardening armature wire (the sculpture kind, not the motor kind) that could be the right hammer for the job.
I’m waiting for my next shipment of Round Tuits… I’ve had some on backorder for months!
I’d rather make it as:
Part A: take a cylinder, through-drill it the size of your shaft, drill/bore a spherical-bottomed hole the OD of your ball, drill and tap the remainder oversize.
Part B: essentially a setscrew with a spherical recess in the tip. Recess sized to the OD of the ball, shaft/thread sized to match the hole in Part A.
Drop your ball-on-a-stick through from the back side, install Part B and tighten to taste. The socket will be far stronger, eliminating the tendency for the ball to pop out, and you can easy adjust the friction. And no sharp shoulders in tension to act as stress risers.
I like it! Much less busywork and neater on the outside, too.
The only trick is the spherical machining, which pretty much restricts me to the diameters of standard ball-end mills. But, eh, I could shotgun a few sizes and get a lifetime supply of mills out of the order.
Likely the bendy wire option is even easier.
Still haven’t figured out how to stick a rod into a ball. I’m sure I can turn it into a CNC project: start with a cylinder and mill off anything that isn’t a ball. That’d surely be easier than building a ball-turning fixture for the lathe, right? [grin]
Well, in my case, when you have a hammerCNC lathe, &c &c. :)
Blog ate my strikeout tags. Bad blog, no blogbiscuit.
Blog ate my strikeout tags.
Dunno how the comment editor is supposed to work; other folks have had their tags eaten, too.
I seem to be able to use tags, but that may be due to my exalted status…
I got some cheap clamp lamps with ball-and-socket joints that actually work. The socket is formed out of two punched pieces of sheet metal bent into roughly hemispherical shape. The ball itself is knurled to provide some drag without too much sticktion. This arrangement does have a tension wingnut, which is pretty much set and forget for a clamp lamp, but would be a lose on vibrating machinery like a bike. These clamp lamps are brand new production, obviously cost optimized, but they actually work fairly nicely. Possibly worth stealing ideas from.
The bash-to-fit idea appeals to me, but I doubt it’d scale to mirror-strut size.
Much to my surprise, the mirror I lashed to Mary’s visor continues to work fine: I knew it’d pop off by now.