Manipulating the Future

This project is a home-brew of a parallel manipulator robot known as a Delta-Robot. This robot has 3 degrees of freedom and allows for very fast positioning of it moving platform. The project started as an idea for learning\teaching kinematics. With too much on the go with other activities progress was idle for some time. Although the mechanical work is near done there is a load of software work now to make the robot useful.

The interesting feature of this robot is its design symmetry. There are relativley few different parts. The XYZ movement of the triangular platform (end effector) is accomplished via three angular joint mechanisms that are placed 120 degrees apart on the base.

Note the implementation shown is still missing the mount to hold it upside down from what is pictured.

The joint mechanisms are each controlled by a separate model airplane servo that pivots the fore arm. This arm is attached to a parallelogram seen as the parallel white rods in the photos. The parallelograms pivot on the fore-arm and the end effector. Again, three assemblies of the servo, fore-arm and parallelogram are placed at 120 degree angular spacing.

From the two photos shown next one can see the rather large range of motion the end effector can have with little change in the fore-arm's angular position.

The project is built with black lexan, while delrin and white fiber glass rods. Although it is not as light as it could be the implemention is sturdy and allows for a good deal of force.

The servos are controlled by a serial servo controller. (The small circuit board shown).

The end-effector platform can be seen below as the three-legged black lexan piece. The joints for the parallelograms are also made from black lexan. Some 8-32 nylon screws hold the pieces together.

More on all this later......

Boxing Day

Having newly made tools just lie around on a bench or in a drawer simply isn't right. Here is a weekend fun project of resawing, box-jointing and making a home for a tool from a previous post.

That't right the words are cut on the CNC. I use DeskEngrave to make the G-Code for the working. It is a free-bee PC tool. Try it if yo want to make some quick signs or lettering.

Pinion Jigs and fly cutting

I think I said it earlier that model engineering and in particular clock making is the art of making tools to make tools that make parts. Although there has not been much shop time lately I did take some time at the end of the summer (2010) to build a pinion milling jig. The plans for this came from Model Engineer's Workshop #164. Not having large mild steel on hand I made it from aluminum.

The lesson here came from proper grinding of a fly cutting bit. Have a look at this great site for details on grinding a fly cutter bit. http://start-model-engineering.co.uk/begin-with-bogs/fly-cutting/
This site is an excellent resource and a good list of model engine projects.

Tesla Turbine

The Motivation to Build

Sometime multiple things need to come together to make things happen. Sometimes people need to come together to create the drive for things happen as well. The motivation for this project came about as a father-son lets-build-something-cool in the shop and spend less time in the 'virtual world' of video games. argh..sigh... Dad.... Perhaps motivated by my observation that my son's interests are different than mine, as mine were different from my father's, but realizing all are linked by a common drive for understanding. Or maybe it was just a nice distraction to build something in the shop while talking between generations.

The Project

This project is a high-speed turbine, based Nikola Telsa's idea, as implemented with plans found here:
www.instructables.com/id/Build-a-15%2c000-rpm-Tesla-Turbine-using-hard-drive-/


and also here:
www.blogger.com/www.phys.washington.edu/users/sbtroy/Tesla_Turbine/Tesla_Turbine.html

This project came about after a few things came together
(1) watching a PBS show on Nikola Tesla www.blogger.com/www.pbs.org/tesla/,
(2) a chance surfing-stumble across a related link on the DIY website www.instructables.com/, and
(3) a sudden thought on what to do with a bunch of old hard-drives piling up in the junk-bin.
The first cool part here was taking apart old hard drives. It is amazing the converstation you can have with teenagers when you are busy on a fun project. Needed was the aluminum platters contained in each drive. These serve as the smooth serviced disks that Tesla used in his turbine.


Cutting the Ventalation Slots in Platters
The Hard Drive platters serve as the turbine disks. However they require slots cut in them to allow for the air (fluid) to exit dear the center of rotation. Three arcs were cut with a 0.250" end mill. Rather than use a rotary table, the CNC milling was machine was given some quick GCode.

















This is always fun to watch when the machine is cutting on its own. A simple clamping jig allowed for disks to the aligned on the table repeatedly. These disks are made of aluminum and cut fairly easily. The picture above shows the end result along with the needed spacer.


















Cutting Ventalation slots in the housing
Ow this posting is skipping a lot of steps. Below is a video of the ventalation slots being CNC'd into the side of the housing. Cutting in plexi is real easy. Can't say I know the actual cutting RPM, trial and error gave a nice result. Going too fast melts everything.





Stacking the Disks on the arbor
Below you can see the arbor carrying the disk stack. This build used 8 disks separated by about 0.030"





















More to come....

Round Table Affairs

This was a couple weekend project - ok - maybe a few evenings too once you include finish. I needed a table to put my coffee on when I sat in the sofa and this looked like a great design. This is a nice design that really calls for an Oak or Walnut construction but I had some Douglas Fir in the shop that was begging for existance as furniture.

The original plan is from Finewoodworking. Although not built completely to the drawings, it features the same slot mortises and tenons with a round top and shelf. You can find the plans and instuctions here. http://www.finewoodworking.com/projectsanddesign/projectsanddesignarticle.aspx?id=29314

Tools used include the table saw, jointer, planer, tenon jig, power mortising tool (oh you could hand cut those?), band saw and sanders of course. Sorry lathe - not this time.

I used one coat of sanding sealer and then three of shellac. The idea here was to get a nice but simple table into service as quick as possible.

Lots of fun to build in the workshop.

A light diversion

As a break from clock building and a chance to get at backlogged projects, I took a few weekends of shop hours to complete these sconces.

The backyard needed lights on the patio. In total the patio will need four, but this blog entry shows two under construction. This was a change from previous large scale furniture projects and thus required much less material. That's a welcome change on the pocket book.

This project is a modification of an Arts & Crafts Sconce found on Wood Store . net

http://www.woodstore.net/arwalscon.html

I modified the back for a simpler project. The light socket was changed to use a low-voltage screw base lamp. The light diffuser is a frosted Mylar. The plan calls for Mica sheets, but to date I am all out of that material...The roof material is copper sheet. I pounded the copper against a hard board to give it the hand-hammered look.

The wood used is padauk. This is an interesting species to use due to its reddish/orange color (and dust!). I had two short planks of this lying around for some time. As they say all good things come to pass (through the table saw sooner or later!) http://www.woodmagazine.com/materials-guide/lumber/wood-species-3/padauk/

The design of the lantern is simple enough with a solid back and three identical sides. All are mitred to 45 degrees. The resulting four pieces are then glued together, some copper sheet metal work, add a frosted Mylar insert and voila - you have (most of) a sconce.

Hard to see in this daylight picture, but at night they look just grand!

Getting a Handle on Things

For many steps in making clock parts there is a good reason to turn the lathe by hand. Any type of threading is such a reason. Well after turning the chuck by hand for all the screw and stand-off threads needed for this project, I got tired. So I guess it is time for another tool.

This time I needed a hand crank so I could turn the lathe manually. An upcoming step in the clock barrel manufacturing will require screwcutting a 12 tpi "groove" in which the cable will wind. having worn out my hand turning the chuck manual for other screwcutting tasks, it seemed reasonable to build a hand-crank.

For a plan I use yet another from the following website. http://www.toolsandmods.com/ralph-patterson.html The design is Hand crank for 7x lathe (version 3).

The shank is made from 1.125 tool steel turned down on one end to 0.795 so as to fit in the lathe headstock spindle. It is drilled through for a 5 1/2 inch 3/8-16 bolt. The end of the shank is tapped to accept the 3/8-16 threads. The shank is then cut on a 30 degree diagonal. As the bolt is tightened the shank pieces slide diagonally to wedge themselves in the lathe spindle.

The arm is made from 0.250 aluminum cut to a key-hole shape. A square hole is cut in the larger end and fits over a matching square cut in the shank. Bending the s-shape took some prying in a vice.

The handle grip is a piece of oak that was turned to a comfortable fit in the hand. I used a 5/16 bolt to fasten it to the arm. I rubbed it with linseed oil as a preemptive strike of getting machine oil on it. Smells nice - looks nice.

The whole assembly fits into the lathe spindle as shown. The crank wedges in the spindle as the wedge-bolt is tightened.

The lathe can be easily turned by hand, both forwards and backwards, and makes a simpler job of manual screw cutting. Why didn't I build this first?

The humble task of making cheese head screws

Yes you can buy screws - but if you are building a clock and are trying to learn the craft of home shop machining, what better activity than to make Cheese Head screws - from scratch. These screws are used to hold the clock plates to the stand-offs.

The making of the screws has a few steps that are somewhat time consuming but the results are worth the time. The experience of doing, making mistakes and learning is all found here. It is humbling to get it right despite being a simple item like a scew.

This task required a few steps and excersises a few skills; turning, threading, parting, and slitting.

(1) Parting and Turning The screws are made of 3/8 drill rod. The material is held in the three-jaw chuck and then faced. A parting tool is used to cut what will be the bottom of the head. This is about 11 mm in from the faced end and done to a depth that leaves 3 mm diameter on the stock. The faced end to the part is then turned down to 4 mm in diameter.

(2) Threading. The stock was turned to 4 mm in diameter and then threaded with 4 mm x 0.7 die (used in the tail-stock die holder). The intial parting in step 1 done to a 3 mm dia leaves a small area unthreaded next to the bottom of the head. Threading is all done with the lathe OFF and the chuck rotated by hand.




(3) Parting to size. The stock is again parted. This time the part forms the top of the head. The end result is a threaded screw with a nice flat (cheese) head. But - oops - it has no slot to make it useful.

(4) Slitting. The screw needs a slot. So over to the milling machine to use a 1/16th slitting saw and a jig to do this. The jig is simply a piece of alumimum tapped for the 4 mm x 0.7 thread. The screw blank is hand threaded on and the saw is centered on the screw head diameter.

One lesson learned here by trial and big error is the cutting speed of the tool needs to be slow. The saw is 2 1/2 inches in diameter. My first attempt was at a real highspeed which lead to disaster. The HSS slitting tool instanly went dull. The corrected process to cut the slot (under CNC control) was to set the head RPM low - maybe 300-500 (sorry no tach yet) and a feed of 2 inches per minute. A first cut of 0.01 inches was done to establish the slot. Then two more cuts at 0.035 and then 0.070 depth were done. It sounds horrible when cutting but the results are great.

With CNC control on the mill this was the time to sip Saturday morning coffee while the machine cuts away - at least after the sequence of G-Codes was entered. Using EMC2 in MDI mode (manual command entry) it was simple work to enter a few lines of G-Code to repeat the slitting operation at the depth passes mentioned. Now if I only had the lathe CNC'd then the enter screw making process would be much much quicker.....

8 Day Weight Driven Clock Movement

I've been working on a brass clock movement for about a month. This is an 8 day movement with a one second pendulum. The plans are from a book by John Wilding called "How to make a weight driven 8 day wall clock".

So far I have the front and rear plates made. The pillars are cut to size and the ends are turned to dimension. I have just started into the barrel and great wheel. Shown are the blanks for these pieces.

This is my first venture into making an actual movement. There are many lessons learned along the way.

One lesson is that Metal working is all about using your tools to make tools to make your parts. The other lesson is it takes me longer to produce results in metal compared to wood. Must be the experience. Go figure.

Not shown here is the effort in getting to this stage. The front and rear plates which measure 6.5 by 4.5 inches were rough cut on my (old) table saw. Cutting brass this way is very loud - make sure you have hearing protection!

The rough sized plates were pinned together with small split-pins. The combined plates where then milled to dimension on my milling machine. I used a fly cutter to make a clean edge. The plates were separated after this.

The corner holes where drilled on the plates on the milling machine. A jig was used to hold the corners in position while I drilled a 7/32 hole. The plates where pinned back together and then the holes where reamed to 1/4 inch. I later enlarged the back plate holes slightly as I tapped them for 5/16-24.

The pillars where done on the lathe and took a new tool to complete. The pillars are made of 1/2 inch brass round. One end is turned to 1/4 inch to fit the front plate holes. The other end is turned to 5/16 and threaded 5/16-24 to fit the back plate. The 1/4 inch ends are to be tapped to accept a screw. The screws need to be made as well. The plans call-out threads in English BA sizes so I have been converting to my closest Imperial or Metric size on hand.

In order to thread accurately with a die I built a tail-stock die holder. Yes this was an excuse to make another tool, but it gave me a chance to mill a hex hole to fit the dies, knurl an edge for better grip and make a huge amount of aluminum swarf as I turned some stock to size.

The tool is a modification of a design by Ralph Patterson. Look here for this and many other mini-lathe mods. http://www.toolsandmods.com/ralph-patterson.html

I use a Jacob chuck in the lathe tail-stock to hold the 1/2 inch drill rod. The die holder is reamed to 1/2 inch on center so that it floats on the drill rod. A piece of 3/8 drill rod is threaded into the die holder to stop it from rotating while you thread.

The die is held in the holder via a set screw. The setup holds the die co-centric with whatever is in the lathe chuck. The threading work is done with the lathe OFF and rotated by hand. I start threading with the die's "start" side presented to the stock. Then I turn the die around in the holder to get threads closer to the shoulder on the pillar.

This has become a fun project and an exercise in working to precision. The book is 64 pages long and I have barely made it to page 15. At this rate I am looking at a completion some time in the fall of 2009.

Spherical Positioner

This is the finished spherical positioner deployed in an anechoic chamber. The system allows a wireless product to be viewed at any orientation of the device with respect to a test antenna. This system works in spherical co-ordinates which tend to numb the brain compared to rectangular co-ordinates when you think about it too long. The positioner allows for a Theta axis rotation of 360 degrees although we limit to 180 degrees and a Phi axis rotation of 360 degrees.



The axises uses steppers and a simple drive system controlled by EMC2 to give computer control over Theta and Phi. A home sensor allows each axis to find zero again when need be.