Posts Tagged M2
Mary has been doing Ruler Quilting and wanted a pencil guide (similar to the machine’s ruler foot) to let her sketch layouts before committing stitches to fabric. The general idea is to offset the pencil by 1/4 inch from the edge of the ruler:
That was easy.
Print three to provide a bit of cooling time and let her pass ’em around at her next quilting bee:
Her favorite doodling pencil shoves a 0.9 mm lead through a 2 mm ferrule, so ream the center hole with a #44 drill (86 mil = 2.1 mm) to suit:
The outer perimeters have 64 facets, an unusually high number for my models, so they’re nice & smooth on the ruler. Even though I didn’t build them sequentially, they had zero perimeter zits and the OD came out 0.500 inch on the dot.
The chamfers guide the pencil point into the hole and provide a bit of relief for the pencil’s snout.
If I had a laser cutter, I could make special rulers for her, too …
The OpenSCAD source code as a GitHub Gist:
When I installed the new fine-tooth filament drive gear (wheel, whatever) in the M2, I ran some numbers that suggested replacing the fixed-position screw with a (more-or-less-)constant-force spring. Some recent discussions on the M2 forum suggest, at least to me, that the drive gear is, indeed, less forgiving of filament diameter variations, drive housing wear, and suchlike than the chunkier old gear.
Having recently bought an assortment of longer M4 screws, I finally got around to installing an appropriate spring from the Big Box o’ Springs and another washer to capture it:
Before doing anything, I measured the gap between the filament drive body (on the left) and the lever arm (on the right) holding the idler bearing: 21 mil = 0.53 mm.
I don’t have a number for the spring constant; it’s rather stiff.
After installing the spring, I cranked the screw to restore the same gap as before, which should mean the spring is exerting roughly the same force on the arm as the fixed-position screw.
The general idea: the spring allows the flexible arm to move as the filament diameter changes, while maintaining roughly the same pressure on the drive gear, thus producing nearly the same depth-of-engagement grooves in the filament. Maintaining “the same pressure” requires the motion to be relatively small compared to the spring preload distance, which seems reasonable with ±0.1 mm diameter variations and maybe 5 mm of preload.
The new filament drive gear hasn’t ever stripped out (after that initial finger fumble), so this will be more of a test to verify that the spring doesn’t make the situation worse.
All the sockets now sport channels in the bottom to capture the braid to the plate cap (whether or not the tube has a plate cap) and the wiring from the Arduino:
The Slic3r preview shows the detail a bit better:
The boss around the pins is now 25 mm OD and snaps neatly into the unpunched hub hole of a hard drive platter:
I moved the mounting holes to 42 mm OC to give the button heads on those screws a bit more clearance from the base.
Moving the knockoff Neopixel up to the top of the pipe leading to the tube base dramatically increases the amount of light going into the tube envelope:
You can just barely see a strip of foam tape holding the LED PCB (loosely) into the too-large hole.
The OpenSCAD source code also produces the improved base clamp; to get a socket, just set
Layout = "Socket" and away you go. It doesn’t yet have the reduced-diameter hole down the middle; that’s in the nature of fine tuning.
Now, with the 0D3 tube properly clamped and aligned in the Sherline mill:
I can slowly run an end mill down onto the spigot:
Eventually converting the whole post into black dust in the vacuum cleaner:
That was completely uneventful, which is pretty much the whole point of good fixturing, isn’t it?
Applying the vacuum cleaner while milling seems to have kept the dust out of the base, although I’m not sure I can pull that trick off every time.
In order to clamp the tube in a V-block, the clamp must position the tube’s centerline so the envelope will clear the V groove, thusly:
The clamp now extends into the V-block and surrounds the entire Bakelite tube base:
The little divot captures the clamp screw and the slot lets the whole affair compress just enough to firmly squeeze the entire tube base.
The tube data table now includes columns for the envelope OD and the base OD, although only the 0D3 (and similar) Octal tubes in my collection have a bulging envelope and a smaller base. You can build clamps for cylindrical glass tubes if you like; I don’t vouch for the accuracy of the table contents.
For whatever it’s worth, the 6SN7GTB tube I started with has a 32 mm Bakelite base and the 0D3 tube has a 29 mm base. That should probably justify two separate entries in the table, but I’m making this up as I go along.
The OpenSCAD source code as a GitHub Gist:
One of the octal tubes in my collection has a broken spigot / key post that lets some light in through the bottom of the normally opaque Bakelite base:
Perhaps drilling out the base would let more light pass around the evacuation tip, but that requires a shell drill to clear the tip. Some doodling suggested a drill with 12 mm OD and 8 mm ID, which was close enough to one of the smaller homebrew drills in my collection that I decided to see how it worked:
You (well, I) can’t freehand such a hole, particularly with a glass tip in the middle, so I needed a way to clamp the tube in either the drill press or the Sherline. A pad for the clamp screw in a V-block seemed appropriate:
The screw hole sits at the 1/3 point to put more pressure near the pin end of the base. Maybe that matters.
The setup looks like this, with a small red laser dot near the front of the base:
The tube rests on a random scrap of plastic, with the hope that the drill won’t apply enough pressure to break the glass envelope.
In normal use, the V-block would be oriented the other way to let you cross-drill the cylinder. In this end-on orientation, drilling torque can rotate the tube; compliant padding for more traction may be in order.
The OpenSCAD source code as a GitHub Gist now includes a module that spits out the clamp:
The original plate cap, even without fins, seemed entirely too large for the 21HB5A tube. There’s not much wasted space inside and, after trimming the outside a bit, this is about as small as seems possible:
PETG doesn’t bridge well and, after cleaning out the wire hole, the remaining shell didn’t hold the brass tube very securely. Epoxying tubes into two caps at once, with a longer brass tube holding them in alignment, worked well:
The tube eliminates vertical tilt and you (well, I) can eyeballometrically align the caps and tubes in azimuth. The thin ring of JB Kwik epoxy around the brass tube isn’t visible, so it’s all good:
This project may eventually force me to try epoxy coating, high-build primer, and good paint…