Making a Clock Colon: Post Milling

Finished colon dots
Finished colon dots

I used a pair of blue LEDs for the colon in the Totally Featureless Clock. Each one has a brass tube to define the dot and a white plastic diffuser to eliminate hotspots.

Some rummaging in the brass cutoff assortment produced a pair of tubes with a 0.300 inch ID that closely matched the width of the LED segment bars. The catch is that I don’t have a core drill that spits out 0.300 inch slugs…

Milling the dots
Milling the dots

So I taped a chunk of translucent acrylic to some plywood scrap and milled the dots. Helix milling on the lesser of a 4% slope or 1/5 of the cutter diameter, 15 inches/min, no cooling, maybe 1500 rpm.

The resulting disks were snug slip fits into the tubes, although I added a dot of cyanoacrylate to ensure they didn’t get any ideas about perpetrating an escape.

It took two disks to remove all the hotspots, which reduced the light intensity to the point where I had to increase the LED current, which really heated up the linear regulator driving the dots. Fooey! In retrospect, I think frosting the LED lens would eliminate the need for a second diffuser without decreasing the intensity much at all.

The code is available as an OpenOffice file there, too.

(Post milling)
(Ed Nisley KE4ZNU - Feb 2010)
(Origin = center of post at surface)
(Double-stick tape holding acrylic sheet to sacrificial plate)

(-- Dimensions)

#<_PostDia>	= 0.300				(post OD)
#<_PostRad>	= [#<_PostDia> / 2]

#<_Thickness>	= 0.120			(sheet thickness)

#<_MillDia>	= 0.250				(cutter diameter)
#<_MillSpeed>	= 15				(cutting speed)

#<_MaxCutDepth>	= [#<_MillDia> / 5]	(max cutting depth)
#<_MaxCutSlope>	= 0.04			(max cutting slope)

#<_TraverseZ>	= 0.300				(safe travel height)
#<_TraverseSpeed> = 25			(safe traverse speed)

G20					(inches!)

(-- Figure cut depth per helix pass)

#<_PassCut> = [#<_MaxCutSlope> * 3.142 * [#<_PostDia> + #<_MillDia>]]		(limit max cut for each pass)

O9000 IF [#<_PassCut> GT #<_MaxCutDepth>]
#<_PassCut> = #<_MaxCutDepth>		(limit max cut for each pass)

(-- Set up cutter comp)

G0 Z#<_TraverseZ>

G0 X[0 - 3 * #<_PostRad>] Y0		(get to entry point)

G42.1 D#<_MillDia>
G2 X[0 - #<_PostRad>] I#<_PostRad> F#<_TraverseSpeed>

(-- cut down through sheet)

#<CurrentZ> = 0.0

G0 Z#<CurrentZ>


O1000 DO

#<NextZ> = [#<CurrentZ> - #<_PassCut>]	(figure ending level)

G3 I#<_PostRad> Z#<NextZ>		(once around)

#<CurrentZ> = #<NextZ>

O1000 WHILE [#<CurrentZ> GT [0 - #<_Thickness>]]

G3 I#<_PostRad>					(clear final ramp)

G40			(comp off)		

G0 Z#<_TraverseZ>
G0 X0 Y0


3 thoughts on “Making a Clock Colon: Post Milling

  1. Hi,

    I’m working on a numitron clock and grappling with how to make a nice colon w/o needing high voltage for something like a neon indicator or an INS-1. I don’t have your awesomely cool machining tools :). Wondering if you’d consider making me a set of these? I live in south central PA.


    1. a numitron clock

      I really like the way those things look!

      If I were going to do the colon again, I’d do it differently… and you probably should, too. Those diffuse white plastic disks weren’t diffuse enough: all that fancy milling just put the wrong thing in the brass tube.

      Start with the same tubing to define the crisp outer border. You can get brass tubes at a local hobby shop or, failing that, the usual online suppliers. Cut it with a fine saw and file the edge to be perpendicular… or, even easier, saw the ends from a single tube, knock off the burrs, and put the LED in the cut end with the original ends outward.

      Then I’d stack up disks cut from a translucent plastic milk bottle. File a knife edge on the outside of the remaining brass tubing and use that as a hollow punch to cut out the disks. Maybe you could heat it a bit to soften the plastic, but that’s in the nature of fine tuning. If those don’t provide enough diffusion, drink a bottle of Lactaid milk: the white plastic is amazingly thick.

      You probably need an amber LED to match the incandescent filaments. Sand the LED dome to a rough surface; I think a roughened dome will actually produce a more uniform pattern than a flat end, but you should certainly test that assumption.

      All that sounds rather low-tech, but I think it’ll provide better results with a whole lot less hassle. Honest!

Comments are closed.