Ellipse machines

[Jack Curtis]

George, thought about your ellipse machine today when I read in The Toolchest of Benjamin Seaton 2nd ed about his trammel ellipse machines, p. 76. It's a 3 trammel rod, 2 of which are inserted into a metal cross with two grooves. Still very cool.

[Kenneth Speed]

I'm not trying to one up anybody but there is a video at Popular Woodworking in which Steve Shanesy. one of the editors, demonstrates how to generate an ellipse with paper, pencil, a straight edge and a compass. The method is elegant, not mind bogglingly difficult ( Even for me!) and applicable to ellipses of virtually any size whatsoever. I used the method to generate an ellipse for small paterae and will use it soon to make a larger ellipse for a drawer front inlay.

[Jim Koepke]

For my last big ellipse the trammel and framing square method was employed.

Works fine for me.

Explained here:

http://zo-d.com/stuff/squares-and-measuring/how-to-use-a-framing-square-to-draw-an-ellipse.html

jtk

[John Coloccia]

You can also just use a string and two push pins, but George's ellipse making machine is very artful. It's quite a bit of craftsmanship to make something like that work without binding, believe me! :)

[george wilson]

I have an old book about 3/4" thick that shows a myriad of things you can do with a framing square. Its uses are amazing.

I guess the ellipse machine is still in the Anthony Hay Cabinet Shop,hopefully. Someone else said that the inlaid tool chest had disappeared. I have no idea.

[Jack Curtis]

For my last big ellipse the trammel and framing square method was employed.

Works fine for me....

Thanks, Jim, looks pretty clear and easy. The cool thing about Seaton's set was it was clear how it worked, which wasn't true for me with George's photo (personal failing, I'm sure), maybe it's too pretty. :) It was also very interesting that a tool kit from 200 years ago included such a thing, I'm loving this book.

[Jack Curtis]

You can also just use a string and two push pins, but George's ellipse making machine is very artful. It's quite a bit of craftsmanship to make something like that work without binding, believe me! :)


Have you tried to make one, John?

[John Coloccia]

Have you tried to make one, John?

Something similar.

[Jack Curtis]

I have an old book about 3/4" thick that shows a myriad of things you can do with a framing square. Its uses are amazing....

Yeah, I've got a couple of booklets on framing squares. The square's are almost too flexible and useful, hard to grasp it all.

[Jack Curtis]

Something similar.

I see the main problem as making the channels and runners. The trammels and rod are pretty easy. I may try to make a wooden version, have an old boxwood plane that needs recycling.

[Maurice Ungaro]

George,
Do you think you might be able to reproduce your ellipse machine, if on paper only? Would be kind of fun to try to make one.

[John Coloccia]

FYI, this kind of device is called a Trammel of Archimedes. A practical method for construction such a thing, or at least how *I* would happen to approach it if I want to make it work without much fuss, is to build it out of some stable wood (some nice baltic birch ply works...maybe honduran mahogany), and simply route two criss crossing channels. It turns out that they don't even need to be 90 degrees to each other, I don't believe. I believe this because an ellipse is essentially sine vs cosine where they are scaled differently (as opposed to a circle where they are scaled the same), and you will get that if you have 2 pairs of right triangles (the 90 degree case) or generally any two pairs of triangles with complementary angles (needs to come in pairs or it won't be symmetrical anymore). So without analyzing this any deeper than that, combined with a handwave, I'm going to say that the angle isn't critical.

Then I would take a block of UHMW plastic sized to fit the channels (or make channels sized to fit your UHMW), and attach a small collar to it. Then I would stick a metal rod through the collars, affix pencil to the end, and viola....a working ellipsograph. The trick is getting it tight enough that you can make large ellipses without wobbling all over the place, but loose enough to still actually work. I think using UHMW for the sliding blocks gives you a fighting chance of doing that without too much difficulty.

[Jessica Pierce-LaRose]

With the one George made, isn't the precision required in the brass on top more than the wood on the bottom, since the brass is actual bearing surface, where as the wood only needs to allow the larger "button" that won't fit pop out of the channel in the brass room to slide? Sounds like you're describing a version, John, without the "t-track" thing going on?

[John Coloccia]

With the one George made, isn't the precision required in the brass on top more than the wood on the bottom, since the brass is actual bearing surface, where as the wood only needs to allow the larger "button" that won't fit pop out of the channel in the brass room to slide? Sounds like you're describing a version, John, without the "t-track" thing going on?

My version has the sliding blocks in a track. I was just noting that using UHMW blocks for the moving parts will make it easier to get working tightly and smoothly. I was also noting that I don't believe the two tracks need to be 90 degrees to each other, without formally analyzing it but just going off of intuition, so you don't have to kill yourself getting an exact 90 degree angle lest you end up with a cockeyed ellipse. I believe it will work for any two angles.

[george wilson]

There is nothing complicated about making an ellipsograph. Have you guys seen those little wooden toys that were called "smoke grinders"(or another unmentionable grinder!!) They usually are a pair of dovetail shaped grooves at 90º,with runners in them. The ellipsograph just has a bar attached to the 2 sliders,with a pencil on the end. 2 little pins stick out of the bottom to keep the device from moving while you use it.

Mine used T slots,because the original used them. As little as they actually get used,brass guides aren't really necessary. Metal sliders help,as wooden ones are more easily split by tightening screws on them to hold the bar. But,if you are careful,wooden ones would do.

You don't really need to make the little swivel knife attachment. It is nice,but a luxury.

[Jack Curtis]

...Then I would take a block of UHMW plastic sized to fit the channels (or make channels sized to fit your UHMW), and attach a small collar to it. Then I would stick a metal rod through the collars, affix pencil to the end, and viola....a working ellipsograph. The trick is getting it tight enough that you can make large ellipses without wobbling all over the place, but loose enough to still actually work. I think using UHMW for the sliding blocks gives you a fighting chance of doing that without too much difficulty.

Great idea on the UHMW, I've got some of that around here somewhere, save that boxwood. Don't know about the plywood, the glues are a bit hard on my edges. Thanks.

[Jack Curtis]

...

You don't really need to make the little swivel knife attachment. It is nice,but a luxury.

Thanks, George, beautiful as ever; but don't you need the brass guides to define the T/dovetail channels in this implementation? Pretty cool to have a knife attachment.

[george wilson]

I think you could do just fine using only hardwoods with no brass to make up the T slots-or dovetails. Since this is a very special purpose tool,it would not get worn out drawing a few ellipses.

Metal sliding surfaces are better for having close tolerances without sticking and binding,BUT,those brass tracks on mine are still attached to the mahogany,and the mahogany could still move a teeny bit,taking the brass tracks with them. I used some good,old HARD mahogany in this tool,and so far(3-4 years),the tool still works fine.

[Eric Brown]

George - quick question. How did you attach the brass to the wood?
I'm guessing brass pins. Thanks.

Eric

[george wilson]

I silver soldered flat head screws to the underside of the brass and drilled holes into the wood. Then,epoxy was put into the holes,and the screws and their plates were lowered onto the holes and clamped till dry. That way,I had no remnants of screw heads to show later on the brass,when something shifted a bit,as it always does. A little trick I learned when making presentation pieces for the Reagan presidency. Also came in handy on PGA trophies and other work that had to be extra neat.

The brass plates were mitered at 45º and silver soldered to make the right angle plates. If you use the lower grade silver solder that is used by plumbers on copper pipes,it has a more brassy color than the better grades which have more silver in them. These tricks are worth knowing.

No matter how careful you are when mounting metal to wood with visible screws that have been filed flush,the tiniest wood movement will show the outline of the heads.

[Jack Curtis]

I think you could do just fine using only hardwoods with no brass to make up the T slots-or dovetails. Since this is a very special purpose tool,it would not get worn out drawing a few ellipses....

It is special purpose, and there's hardly any reason to make one except to extend skills. However, if one were to engage in period inlays and the like, I can see a lot of potential use.

[Harlan Barnhart]

I'm not trying to one up anybody but there is a video at Popular Woodworking in which Steve Shanesy. one of the editors, demonstrates how to generate an ellipse with paper, pencil, a straight edge and a compass.
I saw that. Is that a true ellipse? It seems to me to be a close approximation made by combining a large and a small arc. Doesn't the curve of the ellipse vary continually?

[george wilson]

I haven't seen the video,Harlan,but you are correct: an ellipse is not part of a true circle anywhere. "Approximate ellipses" were used by draftsmen to make a quick ellipse many years ago. They drew these "ellipses" by using 4 sections of the approx. ellipse made up of arcs of 2 different radius circles.

We were taught to draw these approximate ellipses in college for use on isometric drawings.

[Carl Beckett]

There is nothing complicated about making an ellipsograph.

This reminds me of one of my college professors. An advanced class. We were all struggling with one of the
Problems so he finally decides to show us how

Starts out with the statement ' This problem, as if if was a problem, can be solved by.......'. And begins to write equations. Two huge chalkboards of equations later, we have the solution.

I always enjoy seeing your work George

[Leo Passant]

I made an ellipse plotter about 25 years ago for producing large elliptical table tops (up to 10 feet by 7 feet). I routed a dovetail cross in a large square of MDF and planed a couple of small dovetail mahogany sliders to suit the dovetail channels. I then attached a trammel with incremental markings to the sliders with round head screws and washers. A block was glued to the other end of the trammel which had a hole drilled in it to accept a pencil and a machine screw to secure the pencil. The sliders were waxed and their leading edges minutely chamfered to allow for easy crossing of the channels.

[george wilson]

Do make the sliders about 3 times longer than the gap they must traverse. If you make them too long,they may collide with each other passing the gaps. Experiments may be needed.

[Jack Curtis]

Do make the sliders about 3 times longer than the gap they must traverse. If you make them too long,they may collide with each other passing the gaps. Experiments may be needed.

What's a "slider" and what's the gap? This has got me going round and round. I understand the collision aspect; but how does making a slider longer help avoid collisions?

[Jack Curtis]

I made an ellipse plotter about 25 years ago for producing large elliptical table tops (up to 10 feet by 7 feet). I routed a dovetail cross in a large square of MDF and planed a couple of small dovetail mahogany sliders to suit the dovetail channels. I then attached a trammel with incremental markings to the sliders with round head screws and washers. A block was glued to the other end of the trammel which had a hole drilled in it to accept a pencil and a machine screw to secure the pencil. The sliders were waxed and their leading edges minutely chamfered to allow for easy crossing of the channels.

Fantastic, Leo. How large was this MDF square? How long the trammel rod?

[george wilson]

The sliders are the pieces that slide back and forth in the grooves. I didn't say that making them longer will prevent collision. If they are TOO long,they WILL collide. What you need to do is reach a happy medium where the sliders don't get cockeyed trying to bridge the gap in the tracks,but not TOO long,or they may intercept each other. I believe mine were about 3 times the width of the gap.

[Leo Passant]

Fantastic, Leo. How large was this MDF square? How long the trammel rod?
The table sizes I made varied and I actually had several different sizes of 3/4" thick MDF base boards. The dimensions of the base board need to be double the distance between the long axis and short axis pivots on the trammel – plus a couple of inches (to keep the sliders from coming out of the tracks when they align with their respective axes).

The trammel needs to be as long as half the long axis distance – plus a couple of inches, again, to allow for the long axis pivot, and sufficient room to mount a pencil-carrying block.

One improvement I made was to make the pencil block moveable with an in-built clamp. I also made an acrylic trammel with an in-built router base (stepped down to account for the thickness of the base board) which aided in forming the edge decoration on some tables.

This is one of the oak wake tables I used the ellipse plotter for...

234790

234791

[Jack Curtis]

The sliders are the pieces that slide back and forth in the grooves. I didn't say that making them longer will prevent collision. If they are TOO long,they WILL collide. What you need to do is reach a happy medium where the sliders don't get cockeyed trying to bridge the gap in the tracks,but not TOO long,or they may intercept each other. I believe mine were about 3 times the width of the gap.

OK, understood, thanks.

[Jack Curtis]

Thanks, Lee, gorgeous tables.

[Jim Koepke]

One way to think of an ellipse is as a circle viewed at an angle.

jtk

[george wilson]

Actually,a circle viewed at an angle is really a perspective circle. The radii of the arcs of the circle nearest to you will always be somewhat larger than the arcs which are farthest away. This will not so easily be noticed if the circle is drawn on a piece of paper and viewed near to the eye. But,look at a car tire or another larger circle that you can get some distance from carefully,and if your eye is trained,you will see this.

The perspective circle will resemble an egg,which has a large end and a smaller arc on the other end. Imagine this egg resting at an angle and tilted up on the larger end. That approximates a perspective circle.

If perspective can be eliminated,the circle would become an ellipse. But,in the real World,we always have perspective.

[Kenneth Speed]

Harlan wrote, "I saw that. Is that a true ellipse? It seems to me to be a close approximation made by combining a large and a small arc. Doesn't the curve of the ellipse vary continually?"

No, it isn't a true ellipse but what impresses me is that the method makes it possible to make an ellipse or false ellipse of virtually any size required. In my case, I was making satinwood paterae for a federal style work table I am building/ attempting and I needed tiny 1" x 2" ovals. No other method I knew would allow me to draw one so small. I have to confess I ended up using an architects plastic template to make the ovals I need.

I have to confess I have never heard a piece of furniture criticized because one of its components was not a true ellipse.

Ken