Having had many bike helmet mirrors disintegrate over the miles and years, I’ve had a background project bubbling along to build something more durable. Whether that’s feasible or not remains to be seen, but here’s another go at it.
A full-up ball joint seems to be more trouble than it’s worth and, in any event, requires far too much precision to be easily duplicated. That renders those doodles, mmm, inoperative.
These doodles aren’t workable, either, but they convert the ball joint into two orthogonal rotating joints that could be 3D printed with some attention to detail.
The general idea:
- An ordinary inspection mirror has most of the tricky bits
- An azimuth-elevation mount aligns the shaft relative to the helmet
- The mirror shaft extends to put the mirror forward of your eye
- The existing mirror ball joint aligns the mirror relative to your eye
What’s not to like:
- Exposed screw heads
- Off-center, hard-to-grip adjustments
- Probably not printable without support due to all the bearing surfaces and cutouts and suchlike
A few more days of doodling produced something that seems better. The az-el joint axes and the mirror shaft axis now meet at a common point, so the mirror shaft moves as the radius of a sphere. The elevation screw hides behind the azimuth mount, out of the way, which makes it awkward to adjust the tension.
The helmet mount plate must be concave to more-or-less match the helmet curvature. I’ve been securing mirrors using double-sided foam tape to good effect, but it requires a fairly large pad to provide enough adhesive force.
Two glue joints make everything buildable and should have basically the same strength as the parts themselves. The helmet plate builds concave face up. The az and el mounts build with the bearings upward, as do the mating surfaces on the other parts. Maybe the screws need actual nuts embedded in the mating parts, in which case there may be problems.
The setscrew holding the mirror shaft can crush the tube; I think they’re thin brass, at best. Putting a stud screw on the end will hold the shaft in place, leaving the setscrew to prevent rotation. Perhaps the stud can reinforce the tube.
What’s not to like:
- Many parts (but all buildable at once)
- It sticks out too far from the side of the helmet
- Ugly on a stick
But maybe something will come of it.
The Smell of Molten Projects in the Morning 
Take a look at the helmet mounted cameras. Perhaps they have a mounting system that could be adapted.
gm
The last time I looked, helmet cameras used either sticky pads or straps through the vent holes; I think the former adhere to solid-shell mountaneering lids.
Both seem adapted to a wide-field camera that you point more-or-less forward. They can’t adjust the camera orientation in precise little increments, which is exactly what I need for a mirror.
Perhaps the newest cameras have delicate spidery mounts?
I do want a helmet camera, but … that’s a whole different project! [grin]
We had at one point talked about a bendy wire solution, rather than articulated joints. With your newfound printing expertise, it seems like you could print a mirror-to-wire bracket and a wire-to-helmet-surface-area bracket allowing for industrial double-stick tape or the like.
I still like bendy wire, but this one has the advantage of easy fore-and-aft adjustment, although I don’t know if that’s compelling enough to make up for all the gimcrackery. This might turn out to be a template for the bendy version, after we figure out exactly where the mirror must sit and how it must be angled on our helmets.
The whole notion of rotating joints with enough friction to remain in place forever more, except when the helmet falls off the bike seat, seems iffy at best.
I must also come up with a mirror-to-boom interface that doesn’t have so many fiddly parts as the inspection mirrors I have on hand. More on the fiddly parts problem over the next few days…
Printing a hemisphere that goes over the back of the mirror makes it aerodynamic! (and provides a fair volume for boom attachment.)
I have brief thoughts about putting an ultrabright LED in there as well, for directional illumination, but that runs into swiss-army-knife-or-el-camino territory.
Aye, with cable management issues… but I like it anyway!
Another back burner project is a decent rear light which might also become a front running light.
So many projects, so little time…
On the subject of hemispheres, why not print a ball joint with two hemispheres with a spring in between. Then the female part is a printed collar matching the diameter of the ball at about 3/4 extension. Can’t say whether or not it will survive a fall from a bike seat, but should be easy enough to at least attempt?
The printer really doesn’t have enough resolution for small-ish spherical objects; the filament winds up about 0.33 mm tall x 0.66 mm wide.
The top of those little octopodes (sounds fancier than octopi, I’m told) shows the problem. They’re about thumb-sized and the top few layers get rather sketchy.
I think cleverly crossing the filament orientations in the ball and socket so they didn’t lock together could almost make it work, but … this particular 3D really is the wrong hammer for that job.