*
Heres some pictures of a device to draw large ellipses, such as for laying out full scale elliptical stairways. I made this for use in my shop, and is a close copy of smaller versions I have seen. It consists of a trammel bar that has pivoting dovetail blocks attached to it. These attachment points are the length of the semi-minor and semi-major axis respectively, from the drawing end. These dovetail blocks slide in two dovetailed tracks that are perpendicular to each other. These perpendicular tracks only neeed to be as long as the difference between the two axis lenghts. The trammel bar can be any lenght. If one needs to use a router to route an ellipse, the trammel bar would be heavier.
The blocks need to be a nice fit, and then varnished. Then a little spray of WD40 is all to make it worl real smoothly.
Replies
*
Stan;
You do nice work, stair fairys are a seperate breed, but I couldn't keep from noticing the saw burn marks on a member of your jig, save the nice touches for the clients?
*Brisketbean: I did not make any attempt to remove the saw marks or hide them from my post. I let myself get away more than my clients do. Ha.
*Stan,I've never seen such a jig, although I've read about them. Apparently, it describes ellipses very accurately. There would probably be a lot more of these devices around today if we all had a nice big workshop to store and use them in.It will be interesting to hear from others who may use such a device.
*Ken: It does draw real nice ellipses. The trick is having nice fitting dovetails, varnished, then lubricated so they both slide smoothly as the trammel bar end traces out the ellipse. Like I said, I did not invent this thing, just made it larger than one I saw one day for cutting elliptical picture frames. Its neat to watch it draw out an ellipse. By the way, if the track were a full cross, then you could draw a complete ellipse, instead of a 1/4 ellipse that this one does.
*
View Image © 1999-2001"The first step towards vice is to shroud innocent actions in mystery, and whoever likes to conceal something sooner or later has reason to conceal it." Aristotle
*Stan I have a simular jig ( cruder) where the bottom rail extends to the other side so i can complete the other side of the curve in one pass. I use this for window and door tops and openings all the time and it works quite nicely, but I have been told that my very nice curves are not "true"elipses. The jig makes curves close enough for me, but find the right architect and you will have a complaint.
*Jay: Thats a good idea to have your jig able to draw the other quarter. Somebody correct me if I am wrong, but I would say that this jig does trace out a true ellipse.Ken? Ted? Joe? any agreement?
*Jay, finding complainants easily wouldn't surprise me. I know of at least three ways of drawing ellipses that I've used, and a non ellipse.* two marks on a trammel set at 1/2 major, and 1/2 minor axes length, * string and three nails using the foci method,* then there is constructing them using the rectangle method * and an approximation using a rhombus and four arcs where you want to avoid too elongated a geometrical shape.Ellipses are an interesting topic. I don't think I know enough to contribute anything else useful. Sliante.
*Joe: Nice work. Very desciptive. I have to get to work, but I will definately study your site further tonight.
*Jay Zog,If the trammel is constructed properly, like the one that Stan shows, it should draw a true ellipse every time.Ken
*To someone who is not quite familiar with the workings of this jig, I will briefly describe what is happening: When the block furthest from the drawing end is at the channel intersection, the trammel bar pencil is tracing the ellipse at its longest, or semi-major axis. As this furthest block slides away from the intersection, the other block is moving towards the intersection. While this is happening, the trammel pencil is tracing out a decreasing radius. This radius then is at its smallest, or semi-minor axis, when the closest block is in the intersection. Should my jig have had a full cross channel made, this ellipse would now continue with its 2nd quarter being drawn,then 3rd, then a full ellipse when it has gone a full rotation.
*Stan,I once had a "toy" which was essentially your jig with a knob on the handle, and made so that the handle would cycle the full 360 degrees. I think it was called a "do-nothing" toy. You just sit and crank it round and round. Anyhow, it's not terribly hard to prove that you get an ellipse (high school math and some insight), and I have seen the problem presented in various places.
*Ted: I remember at a wood show, they had a similar device called a BS grinder. You would crank it like you said.In Fine Homebuilding they have the same similar jig for sale to make elliptical picture frames.
*Stan, Joe, BB, ...anyone?I don't wan't to sidetrack your thread but I have a somewhat related question. I've seen stairs, usually straight or hour-glass, where the first four steps are bow front treads and risers. The amount of curvature diminishes as you go up until the fifth tread is straight. Are these curves elliptical? If not, how do you lay these out?Thanks Much,Jerry
*Jerry: I usually like you said, start the bowed risers 4 or 5 treads up. I draw an arc for the highest riser that gives a curve pleasing to the eye, but not a radical amount. 2 to 3 inches of bow makes a nice amount. This arc is not elliptic.If you want a formulae that quickly gives this, I will be more than happy to, but I am headed out the door after this post. I would like to post a drawing also if you should like.Whatever this radius is, then you must make your tread nosing radius 1.25 inches larger(or whatever you prefer for your nosing overhang}.Now for the next riser below, take the radius of the top curved riser and add to this the run of the tread. This gives you the radius of the next curved riser below. Again, add your nosing overhang to this to find your next tread nosing radius. Then just proceed on down the stairs increasing each curved riser radius by the run of the tread, and adding the nosing overhang to find the treads nosing radius. Basically, keep the run at the mitered risers the same as all the other treads. The only run change will be on the top bowed riser, and that will be just in the middle where you initially started the bowed risers/treads. Now, I usually also flare the stringers so as these radiused risers are perpendicularly mitered. A flared and bowed stairfront is nice. I will post a picture of one tonight in this thread if I can find it.Hope this helps. I will be glad to answer in more detail this evening. Good luck Stan
*Jerry,
View Image © 1999-2001"The first step towards vice is to shroud innocent actions in mystery, and whoever likes to conceal something sooner or later has reason to conceal it." Aristotle
*Jerry: Heres a little set of bowed stairs that has basically what I was talking about including flared skirtboards. The top bowed riser is followed by risers that increase in radius by the run of the tread.
*Stan,Thank you for the detailed explaination and for the picture. Your method produces a beautiful and symetrical stair. If I understand you correctly the first three steps would have a consistent run, ie. 10", then the fourth step would have a run of 12"-13" (depending on the radius) and the remainder of the steps would run at 10". Have you looked at Joe's drawing? It may be what I'm looking for. The radius increases as you go up the stair, so that the curved treads seem to slowly fade to straight treads.Thank Again for Your Time,Jerry
*Joe,Thanks for the drawing. This is the type of curved treads I'm interested in. I like very much how the curved treads seem to gradually become straight treads. How did you arrive at the radii for the three steps? It seems that you add a certain amount of bow to the run of the third step (say 1.5") then keep the run consistent from then on, resulting in an ever increasing radius. Am I close?Thanks,Jerry
*Jerry: The run remains the same up the stairs. Only the top bowed tread would have a longer run, and that would in the middle of the tread. The run at the edges would be the same.I could not get Joes drawing to come up.
*Jerry,
View Image © 1999-2001"The first step towards vice is to shroud innocent actions in mystery, and whoever likes to conceal something sooner or later has reason to conceal it." Aristotle
*Stan,
View Image © 1999-2001"The first step towards vice is to shroud innocent actions in mystery, and whoever likes to conceal something sooner or later has reason to conceal it." Aristotle
*Page 15 of Circular Work in Carpentry and Joinery (George Collings, originally written in late 1800s, but available with annotations by a new author) shows how to lay out a full elliptical head with nothing but a straightedge. For large-scale work such as door openings a long straigt ripping from tempered hardboard works. Make it about 2.5 to 3.0 inches wide. The method works great to make a full size pattern out of 1/4" or 1/2" plywood.Very easy.... it works.....and simple. Several pages are devoted to different layout methods of ellipses. If I had a scanner, I would scan the page in.
*Joe,I am using Netscape 4.7. It said I downloading an unknown file type.When I clicked on "More Info" Netscape came back with "0 Plug Ins Found".What do I do now?Frank
*Frank,
View Image © 1999-2001"The first step towards vice is to shroud innocent actions in mystery, and whoever likes to conceal something sooner or later has reason to conceal it." Aristotle
*CStanford: Heres a scan of page 15.
*CStanford: Thanks for reminding me of that other method. I just bought a scanner last week so I was able to post it for you.I have tried this way, with nothing but two nails riding the edge of an l-frame. It works,, but this dovetailed jig is really smooth. Also I can route elliptical returns for my treads with this same jig.
*Joe,"Simply "projecting" the arc of the curve a set distance from the riser face" seems to work fine for what I want. I drew out a few variations and they look nice. I checked out the link and realized that I had read that part of your site some time ago. Thanks for the refresher; that method will be useful. I did have a problem with your site though. The pictures overlap some text and in some places the text is overlapping other text making it difficult to read. I'm running IE5.5/NT4.0(sp6) and my res is 1024X768. Also, why couldn't I view your source? If you need to know anything else let me know.Thanks,Jerry
*Jerry: I was digging in my scrapbooks, and I found another picture of the bowed step I showed a frontal shot of. This is a side shot where you can see the bowed treads and risers better.
*Joe: I finally downloaded the whip program that you supplied a link to. I have had static on my phone line messing up my computer staying connected for any length of time. Its fixed, I downloaded, it works,Thanks
*Stan,
View Image © 1999-2001"The first step towards vice is to shroud innocent actions in mystery, and whoever likes to conceal something sooner or later has reason to conceal it." Aristotle
*Thanks Stan! {NM}
*
Heres some pictures of a device to draw large ellipses, such as for laying out full scale elliptical stairways. I made this for use in my shop, and is a close copy of smaller versions I have seen. It consists of a trammel bar that has pivoting dovetail blocks attached to it. These attachment points are the length of the semi-minor and semi-major axis respectively, from the drawing end. These dovetail blocks slide in two dovetailed tracks that are perpendicular to each other. These perpendicular tracks only neeed to be as long as the difference between the two axis lenghts. The trammel bar can be any lenght. If one needs to use a router to route an ellipse, the trammel bar would be heavier.
The blocks need to be a nice fit, and then varnished. Then a little spray of WD40 is all to make it worl real smoothly.