If you have a little cnc mill, then a natural thing
to do is get a little injection molder to go along with it. This will scratch
the itch of your novelty seeking genes for a good long while!
Here we're gonna make an injection mold for a shaft encoder. That will fit
on the end of a motor shaft and detect the position of rotation. 20 holes
is what the design calls for in this case, cuz it's a nice number for dividing
to the ultimate step resolution when the motor runs the axis of our laptop
cnc mill. (...more on all that as we go...)
What is needed is to detect rotation by means of an optical detector. We'll use a disk with 20 holes that are aligned between infrared LED emitter and the detector. The model only has 16 holes. It will use one emitter and a pair of detectors just for testing.
So, first thing is make a model of this shaft encoder and put it up on an o-scope and see what kind of responses we get.
A model was made by injection molding ABS in a poured epoxy mold...good enuff for a model as shown:
You can see the effects of misalignment of the individual holes, even.
Next, i test a model using 2 emitters and find that the response (tested at 10Krpm) is identical, indicating the detectors are saturated either way. But using a pair of LEDs lets the package profile be a bit flatter, lets both halves be identical, plus i got a deal on a load of opto-interruptors (600 sitting here waiting) so let's go with the paired emitter/detectors.
Now that the general scheme is arranged, let's design a mold for the shaft encoder. It will consist of a front, back and the optical disk with the holes in it.
Calculating spacing of the 20 1/16" diameter holes determines the overall size. Let's draft it up...
2 detectors packed face to face with detection at edge on joint...
LED powered a 5v with 240 ohm resistor in series ...each detector sinking 5V thru 4.7K resistor
detectorA peak output dropped @ 0.2V to 4.8V
detectorB peak output dropped @ 0.3V to 4.7V
output variation up to .05V was observed, due to individual radial misalignment of each disk hole
phase is nice quadrature- see above.
CONCLUSION: test a second model which will use 2 leds and 2 holes to increase the signal by directly aligning a beam with each lens instead of using one led and shooting in from the edge to hit both.
This looks ok. Let's figure out the cutters and toolpaths we need and make some g-codes to cut this!
Here's toolpath for the 1/8 mill to cut the main
pockets and bore holes.
This will be the first path. To make code for the nice spirally pockets i
made an applet, which you'll find on the Applets
Page.
This is just the first part of a series on project that is expected to add up to a high speed laptop cnc mill. Stay tuned for continuing developments...
OK...now it's time for the 1/16 end mill tool path. Let's see how many cutters we can break as we go as much as 1/10 deep into the work. It's all about sharpness. I also lube the cutting by flooding it with a mixture of charcoal lighter and high stearate wax. The evaporation of the fluid cools, then the wax will flow onto anything warm enuff to melt it. I learned that stearates are well know for lubrication. Oh yah, don't forget the applet!
Finally, we need to make little lips that will space
and align the mating parts using a 1/32 end mill.
When the 2 parts mate, the lips will mirror and interlock. The package will
be glued closed and sheet stock will cover the outsides. Might cut that on
the mill too...
Next path is the center drill. I didn't spec the
size cuz it doesn't matter a whole lot.
To do the bolt circle g-codes i made another applet, also on the Applets
Page...
Now the 1/8 drill for alignment pins and for thru
holes on the disk shaft.
I didn't need to make applets for this...lol.
Making those arcs was pretty easy using the arc.bore
applet...
All together now:
