Things didn't go as planned.
Florescent light protector tubes have an unfortunate tendency to curve inwards when cut. I discovered this only after I had already cut a slot halfway down one side of the tube. The original design had only one support ring in the middle, which was made from part of the same 10-round vortex mag that provided the feed lips and cap, like so:
Making these support rings, and ensuring that stray epoxy didn't impede the motion of the follower, turned out to be a very tricky business. The guide tabs had to be ground to make them narrower. Cutting under the green plastic pieces was a little awkward, the space under each one had to be cleaned out with a file, and during this process the green plastic often broke away from the epoxy. Superglue was used to fix it each time.
Overall, this was a laborious, finicky, difficult build, and as such this design is not suitable for mass production.
Oh, and there is one more problem: the support rings don't actually help! They hold the tube in place enough to allow the follower to slide fairly smoothly (with some, but not unreasonably much, friction - note that the follower does not slide when the tube is at an angle, as it is in the first picture) - but vortex disks have grippy foam edges, and as such do not slide through the tube without the application of considerable force!
 |
| This tube has curves in all the wrong places. |
Finding another type of tube of the right size which is not heavy would be difficult - FL tubes are pretty much perfect for this application except for their tendency to curve inwards. Attempting to fabricate a structure from scratch would also be difficult - let's call that idea a last resort. I could use an intact FL protector tube and trim the guide tabs from the follower, and use magnets in the follower and the bottom of the magazine to suspend the follower at the right height. This would result in an dramatic increase in feeding force as the follower approaches the point where it stops - and excessive feeding force may result in a jam. Cutting lots of slots in the FL protector tube, such that no unbroken section of tube spans a large part of the tube's circumference, could reduce the tube's inwards curvature, but might make it fragile - this is at least worth a try. Finally, I could coat the FL protector tube in something which becomes rigid when it sets. A fiberglass repair kit might work well here (so long as the magazine can be made without having to grind fiberglass - fiberglass dust is nasty stuff). This seems to be the most promising idea, and this is what I intend to try next.
Being that this is strictly in the DIY realm I would say that overengineering is bad. Choosing a tubular material that is either more rigid and/or doesn't contain internal stress (that causes the tube to deform when cut into) or finding alternative means to fabricate the mag body is the proper answer here, not turning the mag body into a complex/expensive fabricated structure designed to rigidize and restrain the thin polycarb tubing.
ReplyDeletePVC pipe is always worth mentioning for very low cost, extreme availability, and ease of fabrication. If a pipe that will fit discs and remain sufficiently rigid when slotted exists (I would avoid class pipe for use in this manner for similar reasons to the lamp protectors), that is an answer, but possibly a slightly heavy one. Otherwise, the mag body lends itself to being made from two separate lengthwise halves, as opposed to a slotted single-piece tube, and each half would be easily made by cutting pipe lengthwise and thermoforming (using a heat gun) the pieces to the curvature of some suitable mandrel. This can be anything with an OD matching the ID of a mag body (or the T12 lamp sleeves you use now). The thermoforming would eliminate the tension found in pipe and tubing that causes deformation when cut, and result could be made even from thin class pipe. In either case, support bridges to cross the slots can also be made from PVC pipe pieces, either unmodded pipe of suitable size, or thermoformed to the curvature of the mag body to fit perfectly, and bonding problems can be solved with Devcon.
Admittedly my answer to nearly everything is to go for the PVC and the heat gun; this may be a case of "when you have a hammer, everything looks like a nail". There are many other possibilities of easy ways to make this part that I just can't think of.
Now on the subject of dodging the "non-circular cross section mag body" problem entirely: How about a nondirectional follower? Which part of the disc it pushes on or with what area it contacts the disc is not relevant and having a flat surface on the end is arbitrary; it is only necessary that it present a ramp profile to the pusher rod tip in whatever orientation it encounters the pusher in. A truncated cone with softened edges, and a spherical (not necessarily hemispherical, but some section of a sphere), strike me as possibilities.
I am interested in making a design which others can easily copy, so the fact that FL protector tubes are commonly available in a small number of standardized sizes is important. Springing for PVC and a heat gun would be less than ideal (but possible). A FL protector tube might make a good base for thermoforming PVC, provided that a cheap filling can be found which ensures rigidity under heat, and that it does not melt to the PVC.
DeleteI've tried cutting slits in the tube as described at the end of the post, and it worked - at the expense of weakening the tube. With 1 slit in the smaller and 3 in the larger lengthwise half, the disks were still gripped tightly. With 3 in the smaller and 7 in the longer lengthwise half, the problem was completely solved - in fact, the disks seemed to fall more reliably in this section than in a completely uncut tube!
Unfortunately, this weakened the tube to the point where an empty tube began to bend visibly under its own weight about 10 degrees away from the vertical and which collapsed at about 45 degrees. Still, this *could* work with external supports.
This is partially moot, as the current design is unsuitable anyway as it is too labor-intensive already (and this would only make it worse).
However, it does suggest an alternative design: fewer support rings, perhaps only the one in the middle, with lots of slits, and something to strengthen the tube (and/or external supports).
Your idea for altering the follower's shape is a good one. I foresee possible damage if the pusher hits the follower at high speed - but this is preferable to hitting it and jamming hard.
On the subject of Devcon: as a Canadian, I've yet to find the stuff (or an equivalent from another brand) without either a hefty shipping price or an absurd markup. So, in my case, bonding problems mean that the contact area needs to be increased.
What about magnets? One magnet in the follower, and another where it needs to stop.
ReplyDeleteNot a terrible idea, but as I mentioned in the article that could cause a jam.
DeleteNow, if you used magnets below where the follower needs to stop, which repel it, that might work better - but you would need darn strong magnets to avoid a risk of the follower falling though if the blaster is bumped hard.
I've also considered using a long string to stop the follower at the right place. There are a lot of potential solutions which could work here - any solution which allows the continued use of the guide tabs and slots already in the magazine will be preferable because this would automatically stop the follower at the right place. I'll look into other ways (magnets, string, etc.) of stopping the follower if and when I'm convinced that retaining the use of the guide table is not feasible.