Motorizing my Adjustable Angle CNC Clamping Fixture

Using my adjustable angle clamping fixture is relatively easy.  One nagging "flaw" it has is the ease or perhaps unease of setting the angle on it.  Projects often expect a precise setting. The action of lifting the jig and holding it in position while setting the clamps is usually an awkward ballet.  If only I could motorize it. 

Setting the Angle

The desire to motorized my clamping fixture is an old one. I've looked into stepper motors with Arduino control, drive screws and rack and pinion gear drives, long counter-balanced lever arms, magnetic repulsion, etc.. Magic had only a fleeting consideration as I found it an unreliable force.  😉

This project strives to motorize the lifting of the fixture with a linear actuator.  It is also a test to see how fine of control I can achieve with momentary toggle switch up/down button presses to control the actuator.  I hope the actuator movement can be slowed down with a variable speed controller for more precise positioning of the fixture angle.  

Actuator Placement

The tough challenge was finding an actuator that could work with the geometry of my fixture and how it is mounted on the CNC relative to the base the CNC sits on. Position of one end of the actuator needed to be "exactly" where the end limits of a specific actuator would line up in both horizontal and vertical positions of the fixture.  An actuator with eight inches of travel would only work if mounted in a very specific position.

As the fixture rotates from vertical to horizontal, the actuator had to be entirely behind the vertical position.  Geometry dictates exactly where it could be mounted. Ideally the push of the actuator would be 90 degrees from the fixture plate, but that is impossible given the arcing transition from vertical to horizontal.

The actuator I ordered has 8" of travel.  It has built in limits to not go past 8" or under 0" when closed. In the diagram above I had to find a position under the bed of the fixture where when mounted the actuator moves exactly 8" from vertical to horizontal positions.  You can see that I've come very very close.  Within a few 1/1000ths from 8" .

Mounting brackets that came with the actuator are represented in the drawing.   

I took the fixture off the front rail and clamped the bed of it to my CNC so I could cut pockets and holes for the nuts and bolts needed to mount the actuator end brackets

I've now got one end bracket bolted to the underside of the fixture bed.  I've at least momentarily mounted the other end bracket on the CNC base.  I'm waiting for a variable speed control to arrive before wiring up the actuator to see if where it is mounted works.

Wiring the speed control took some contemplation.  The control box is a simple toggle relay that switches the current depending on which control button is hit.  It suggested the speed control should be between the power supply and the control relay box.

The speed controller arrived.

Speed Controller

The lead from the power supply needed to be cut off and wires stripped to attach to the speed controller. Done. The wires from the control box to the actuator need to be soldered together.   

I'd hoped the speed controller came with a way to bolt/screw it down, but that appears to be another puzzle to solve. I want to affixed it to CNC frame so it is secure and won't move when I raise or lower the actuator speed.  It appears to be an aluminum extrusion, capped with plastic ends and a front frame for a metal plate that has the speed dial and markings on it. There are a couple of screws on one end, but not the other end although there are holes for screws there. There are also two screws on one side of the extrusion. In line but not equally spaced from the ends.  I suspect I'll need to open it up to see if I can drill a couple holes in the bottom to put screws through for attaching to the CNC frame.

Soldering the actuator cable to the control box cable is done.  A system check verified that the speed control and remote control all work as expected.  Installed on my CNC it looks like it will do the job.  There is a bit of flex in the fixture, but with the speed control and a magnetic digital angle gauge attached to the metal blade of a tri-square it only took a few second to set the angle on the fixture.  Tightening the cam levers momentarily wiggles the CNC and make the angle on the gauge vary, but once done and settled down the angle returns to what the actuator set it to. 

Actuator in Place

The controller came with a wireless switch.  That is nice as finding a place to mount a wired switch would be a challenge.  I can keep the wireless control in the drawer of the work station. 

In practice I'll never need to bring the fixture to a horizontal position as I have a better setup on the bed for horizontal work.   For now though this linear actuator should do fine handling any or of he future angled or compound angle CNC jobs. 

One small shortcoming is that when beginning at the vertical position and actuator retracted, starting out the actuator force applied is at a shallow angle relative to the fixture bed. Due to play in the connection points precise setting of angles from 80 to 89 are best done manually.   The ideal angle between fixture and actuator would have been 45 degrees at both starting (vertical) and ending (horizontal) positions of the fixture. Unfortunately no such mounting point for the actuator exists. If I could mount the base of the actuator 45 degrees from the bracket when vertical I'd need a bit more than 8.5" of travel to push it to horizontal position. 

A revision to make it work at 45 degrees would require a new mounting platform for the bottom of the actuator, and new positions for both end brackets. The bottom end would extend much farther away from the CNC base. I'm looking for an indirect mechanical approach/solution, but complicating the mechanism is contrary to my personal design philosophy. A better strategy might be to start from scratch and design a new fixture that includes motorized control from the beginning.  

Questions? Suggestions?  Leave a comment.

4D

 

Polygons and Pattern Work

 

Spun Heptagons

There is an interesting geometric feature that basic polygons all have.  For any polygon you can take a chord (connecting opposing corners) and with one end as the center point spin and copy it twice the number of the polygon sides over 360 degrees and you'll find that every corner of the polygon is passed through by one of the spun vectors.   

Heptagon Chord

It works with Heptagons.  It'll work with Nonagons or Hexagons or Octagons too.

Octagon Chord

I suspect the fact that all corners of the polygon fall on a spun vector will be true for any number of sides. For pattern work using more than a 13 sided polygon the patterns created get pretty dense.  To carve them in wood using a V-bit the tinier details get lost or brittle. 

Nonagon Chord

To make an interesting pattern I used a feature in my CNC software to make copies of the polygon that attach to each of the spun vectors. 

Pattern from Nonagons

Before I could make toolpaths to carve the pattern I had to outline the sections I wanted to be cut.  Once done I selected just the outlined shapes and use the V-Carve toolpath with a 120 degree V-bit to produce the shape below.

Spun Nonagons

Even star shapes can be used. More detail can show up depending on what the inner radius is relative to the outer radius.

Spun Star Pattern

A nine tipped star was used to make the complex pattern above. A 7 tipped star was used make the pattern carved below.

7 Tipped Star Pattern

While I haven't actually carved any of the examples above, I am contemplating covering a small feature  wall in my house with an array of patterned tiles made using this trick. More Complex patterns can be made with a denser spun vector array.  Instead of just 14 vectors with a heptagon, 28 vectors could be used. 

Heptagon x 28

More spastic patterns can be made using a circular number of vectors that isn't a multiple of the polygon sides, or picking something other than a corner of the polygon to spin/copy it with. 

Heptagons offset

Above is a pattern made using a heptagon that was offset from where the center was spun about.  Given that there are 360 degrees in a circle to spin about, and a nearly infinite selection of positions you can spin different polygons about, the possibilities are endless.

For a furniture project the vectors could be flattened to fit on the face of a drawer front, or used on cabinet doors to add some interesting detail.  Here are a couple of drawer fronts:

Where necessary a circle area in the center could be left uncarved to allow for a place to screw in a handle/knob. The sectioned areas don't have to have straight sides. The top drawer front pattern was done using a circle spun 16 copies around a center point.   

Questions? Leave a comment. 

4D

 

Drop Leaf Table Design and the Square Root of 3.

The previous post of a table showed a top that when folded in half remained the same length to width ratio.  When folded up the table top took up 50% of the area it did when unfolded.  

This alternate design pushes the concept to a practical limit by using the square root of 3 to make a table that is 1/3 of it's unfolded size when folded up.  A table base that is inset just a little more than 20% but not more than 25%  will support 3 table sections when unfolded but allow the two outer sections to fold down when rotated 90 degrees.  The base length to width ratio is the same as the unfolded table top above it. 

This 1/3 design starts with a wood panel that is 24" x 96" (or 2' x 8').  Using 1/3 of the length as the maximum width of our table top, we can determine the largest possible table top we can make using the square root of 3 ratio.

1/3 of 96 is 32. The square root of 3 is 1.732. 32/1.732 = 18.48".  So the dimensions of each table top section will be 32" x 18.48" .  This will also be the maximum footprint of the table base. Three of these sections will make up the unfolded table top at 32" x (18.48 x 3) or 32" x 55.4".  This table top should seat 6. 

Top ends folded down

Top ends lifted up before rotating

Rotated over the legs for support

Square Rt. of 3 Proportions.

The top pivots around the center of the base.  The pinwheel stretcher configuration makes room for a center plate where the pivot can be placed.

Obviously the legs could be more interesting. The top would have a nice edge profile. A classic drop leaf connection and hinges would be used to allow the top ends to hinge down after rotating 90 degrees. Tops of the legs have a teflon t-peg in them to allow the top to rotate smoothly over them. The center section of the top has a metal post that slips into a bronze bushing in the wood plate between the stretchers. It is held in with a spring, washer and wing nut to allow some tensioned lift when rotating.

Final details needed can only be determined if I ever make such a table. This post is simply a parking place for ideas I've had over the past 4 decades.

Question?  Leave a comment.

4D

Another useful Geometry Application. Furniture Using The Square Root of 2.

When I taught furniture design to college students one project assignment was to design a table that could be reduced by 50% and still be useful when the starting dimension wasn't needed. 

This is one useful solution that my students have used several times over the years.  It uses the Square Root of 2 (1.414213562373095) to create a rectangle that remains the same width to length ratio when folded in half. 

In this example I have a wood panel to use that is 24" x 36" x 3/4" thick.  If I use the 24" side as the width of my table top and multiply it by 1.41... I get 33.94....  So with 36" of potential length I have room for the kerf of a table saw or CNC 1/4" profile cut. 

The Whole Top

One half of this 24 x 33.94" top is 16.95 x 24. When rotate 90 degrees and centered over the base it is the same ratio as the starting dimension.

Half the Top

Soss hinges between the table halves are hidden when unfolded, but will be visible on the inner edge when folded in half.

Rotated and Centered

One challenge was to find a single point of rotation for the top that would keep the top centered over the base after being folded over and rotated 90 degrees. 

Project a line from the center of 1/2

To find a common point that will be in the same position between unfolded and folded/rotated draw a line 45 degree from the center of the side that will be on the bottom and rotated.  

Project the same line, but from the rotated/centered outline. 

Draw the same 45 degree line from the rotated view of the bottom half.  Where those lines intersect is the only point the top can be rotated about to remain centered over the base in both positions.

The inner outline shown in the 2 views above is the outline of a base that would be inset 20% from the large perimeter. You can see it is still inset a small amount under the folded top. 

Here is one option for a base.  The legs extend to the inner edge of the 20% inset outline. A bracket attaches  to one stretcher to provide a pivot position for the top to rotate about. 

Under the folded top.

Under the open top.

Need a larger table?  What is the largest rectangle of that same sq.rt. of 2 ratio that can be made from a 24" x 60" panel? Account for a 1/8" wide kerf on any cuts.  How much scrap will be left after cutting the two table halves?

As the table base is a little smaller than 1/2  of the table top, you could also make the folded half a drop leaf rather than just double the smaller top thickness. This side would hang down and block access to one side of the base, but if that is not important the table top will benefit from being the same height above the floor in both unfolded and dropped side positions.  

Leave a comment if you need any more information.

4D

Dividing Any Line into 3rds or 5ths Using a Triangle.

 I discovered this method to divide any line into 3 or 5 equal sections by just playing around with the drafting features of my CNC software.  It has come in handy many times while creating joinery for furniture connections.  Sizing a tenon for the end of a stretcher there is a 1/3 rule which suggests the tenon thickness should be 1/3 the thickness of the stretcher.  When I'm connecting a narrow board into a wider board I like to increase the relative size of the tenon to 3/5ths of the stretcher thickness. One 5th for the perimeter shoulder and three 5ths for the tenon.  

I haven't seen this method used anywhere on the web or in print.  It may exist out there some where.  When I look for methods to divide any line by 3 or 5 equal sections I find most use an adjacent connected line with marked off known segments, projected parallel across to the starting vector to divide it.  I used that trick when drafting was done with a parallel bar, triangles, compass and straight edges. There is another easy trick to dividing the width of a board using a ruler/scale held diagonally across the board at marks equal to the number of divisions you want. 

This triangle method starts by adding two lines to your starting vector to make a triangle. Now find the center points of all three sides of the triangle.  This works with any size or shape of triangle, although the density of lines created is easier to draw and see on a large triangle closer to equilateral. Be accurate.  Mistakes here propagate through following steps. 

Find Centers of Each Side.

Next connect the centers to each other.

Connect Centers

Now draw in the medians by connecting the corners to the opposing center points.

Corners to Centers

You can see where each center line intersects the center triangle lines.  If you draw a line from a corner through the opposing intersections, it will cross the opposing face at 1/3 over from a corner. 

Thirds, in Red

Do this for all 3 sides.

3rds on all 3 sides.

The new vectors intersect the previous vectors, and provide useful nodes we can use to divide each side into 5 equal segments. 

Project to find 5ths, in Green.

It is easy to get confused about which intersections to use.  Basically for each face there is a red/red, a red/black. a red/black. and a red/red intersection to pass through. 

5 Equal Segments.

I usually only need to go far enough to find 5ths on one side, but it is easy to project the other two sides to divide them both into 5 segments.

5ths on all 3 Sides.

There are further divisions that can be found with the new intersections created. I rarely need 7ths or 9ths though and have a much more automated way of dividing line segments in my CNC software.  

For example:

7ths. In Orange. 

9ths. In Purple

I suspect this method will work for odd numbers up to infinity so long as you have infinite patience and a very very fine pencil. I extended it to 11ths with no trouble but the density of vectors is making it harder with each step up.

11ths

Of course since 13 is a lucky number I went ahead and proved that 13 segments of a vector can also be found. 

13ths.  Just Because.

I dedicate this post to my Sister Vickie's husband Chris Z. A useful application of geometry. 3rds and 5ths (and 7 and 9 and 11 and 13), with no parallel lines needed.  If no one else before me has discovered this method, I name it the D.A.Brown Method. Published here first on 7/9/2023. 

As a final teaser, you can also divide the triangle sides by even numbered segments using a variation of projected lines through intersections of previous vectors used.   Although this works with any shape of triangles, you can also do the same with 4 sided polygons. With rectangles, trapezoids, or parallelograms this is easiest to see. How about 5 sided polygons? 

4D

Zero Clearance Fence for my Miter Saw

 The normal gap between fence sides of most miter saws is wide enough to allow rotating the blade angle or tilting the blade for angled or compound miter cuts.  This gap width is a bit dangerous for making small cutoffs.  The small cutoff part can get kicked back and thrown by the blade. 

I made this zero-clearance fence add-on to add safety to my Bosch Glide miter saw. 

Zero Clearance

The saw already had holes in the fence side for bolts.  A quick measure of height off the bed and location side-to-side of all the holes and I drew of patterns for these parts in my CNC software.   What the CNC thought it would be cutting out:

Cut from 1/2" plywood

The slots allow me to slide the plywood back one inch.  The top half of the fence itself can be loosened and slid back to allow miter cuts with the blade tipped down. I can open up the gap a small amount for angled miter cuts with the blade remaining vertical. 

1/4-20 T-bolts are used with wing nuts on the back side to hold the add-ons in place. 

I've posted a .CRV3D (Aspire) and a .CRV (VCarve) file on Vectric's forum for anyone using their software. 

Comments welcomed

4D

A Board Stretcher Joint. Just for Fun.

These are renderings from my CNC software of a zipper joint I came up with. A simple way to connect the ends of two boards. Easy to cut using a 3/16" end mill with the boards clamped flat.  This joint could even be used across the center of a table top. For furniture buyers with trouble deciding which woods they want their furniture to made from.  

Mix wood types

Both sides use the same toolpath

Are there schisms between members of your family?  Sport fans who don't agree on the teams they support?  Hardwoods are available in many colors, and two different colors of wood could be used to make a zippered table with both team colors.  

Just for fun. Very possible. 

4D