Wednesday, January 3, 2018

Model vs. Reality

Reality 1: the deep freeze that many (most?) of you have been enduring is making its way into middle Florida, modified by the warm water everywhere, of course.  The temp this afternoon was 46 (I think it was 2PM).  It has been raining since Monday, stopped after 3 or so, and tonight is forecast to possibly touch freezing.  With winds 15mph gusting 20, I don't think "cold pockets" are stable, so we'll get whatever the area gets.  Certainly not record-shattering, but the coldest night in a year or two. 

Reality 2: given the cold rain until late afternoon, instead of building my antenna on the back porch, I focused on building the box for the AC transformer that I mentioned Sunday.  The model is pictured in that post.  Here's what the looks like in reality after a few hours in the shop.  Only the top is cut out.
No, it's not done, but my main worry, the cutout for the variac came out OK although the initial CNC cut required modification. 

To get to this point, I draw the parts in my 3D CAD program, Rhino3D, then used my CAM program Deskproto to generate a simple tool path, based on cutting it with a 4-fluted, 3/8" end mill. 
The light purple line is where the center of the cutter goes.  No, I didn't need the outer rectangle to cut the panel to size; it came that size.  I just needed the circle and rectangle cutout.  There's a potential GOTCHA here.  I can compensate for the cutter's diameter in either this program or in the CNC controller with a tool table.  I have to remember to Never Do Both.  

The measurements for the big circular cutout and the rectangular cutout for the meter were all taken by hand, and used to make the layout.  Notice that I said I was using a 3/8" diameter cutter?  That means none of those sharp corners you see are actually sharp - they have a 3/16" radius (the cutter's radius).  That means I need a secondary cutting operation, or even finer with a couple of cutters.  I chose to use a single cut with a 1/8" diameter cutter, because that has worked for me before.  Now the corner radius goes down to 1/16" and I can cut the corner square with a file.  You can see those radiused corners in this shot, right after the first pass with the 3/8" cutter. 
To do the last cut, I use the CAD program again to get the X/Y coordinates of where the 1/8" cutter is when the cut is complete.  To do that, I draw a cylinder the diameter of the cutter and push it into the corner until it's tangent at both sides.  Then I take the coordinates of its center at that point, and use those values in the code that moves the mill.  It's a pretty easy cut and takes off very little metal.  You might notice that I drilled and tapped my "tooling plate" for the four screw holes the cover has built into it, and put scraps of 1/8" aluminum under it to save cutting into my plate.

When I said, "the initial CNC cut required modification", what needed to be changed was the depth of the rectangular cutout on the variac; it was too short.  It was right at the top of the part, but to get the box cover onto the variac, I had to slip it over some terminals which are taller than that rectangle allowed.  I extended it 1/10" and it went together easily.  Similarly, the rectangular cutout for the AC panel meter was too small to pass a feature it has on the sides, so The rectangle had to be opened up about 1/8" as well. 

And that's how I spent my day today.