I've built a trammel (Less than 4ft square) to try to make an 8ft by
4ft oval table.It works as a trammel but is too small using my
current geometry to make an 8ft table. When I've looked at pictures of
commercial jigs they are smaller than mine but make bigger tables. My
slide arm is 4ft long with a central pin 2ft from the router. Is there
a different formulae to get a scaling effect.
Chris
"Chip Buchholtz" wrote:
> It might be easier to use the string and nails method rather than
the
> tramel method. The hard part is finding string that will not
stretch.
Forget the string, layout the nails with a tape measure and use a
batten.
Trim proud, then clean up fith a fairing board.
Formula for an elipse:
X^2/A^2 + Y^2/B^2 = 1
This will give you an elipse for 1/4 of the table.
Make a 1/4" hardboard template and use it to duplicate the other 3
sides.
Lew
RE: Subject
As I remember, this is a project to construct a 48" x 96" eliptical table.
Using the formula given, divide the 96" dimension ("X" Axis) into 3"
intervals, then calculate "Y" Axis values which locates the nail location
for all four quadrants.
After that, it is batten and fairing board time for a full size, 1/4"
hardboard template for use with a router and a pattern bit.
++++++++++++++++++++++++++++++++
> Trim proud, then clean up fith a fairing board.
>
> Formula for an elipse:
> X^2/A^2 + Y^2/B^2 = 1
Lew
Like the OP, I too prefer the trammel method. It's simple and direct.
I've built the elliptical mirror designed by David Marks on Woodworks.
I didn't do the string method and cut to the line as he described.
Instead, I did what the OP says he wants to do and built a trammel to
use my router and a 1/2" spiral carbide end mill to cut the ellipse
directly.
While I was figuring out how to do that, I used Autocad to draw my
planned trammel. For any ellipse, there is a long axis & a short
axis. Because an ellipse has two centers, the trammel you're building
will move on two points.
R--------------------------------------------------P1----------------------P2
Formula:
The distance between those two points (P1 and P2) is (long axis -
short axis)/2. This is the distance from P1 to P2.
Since you'll be cutting it directly from your material, you'll likely
be useing a router with a bit of some diameter. To locate the center
of the router R from P1, use the answer from the formula above and
subtract it from 1/2 of the long axis of the ellipse. Add to this 1/2
of the cutter diameter. That gives you the distance from P1 to R.
Using those answers, I knew I would build an arm containing P1 & P2
and would attach that arm to a router base to get point R.
The base I built is an MDF square with two dados at 90 degrees to each
other in the manner of a big X. Into those dadoes, I put aluminum
T-track. The T-track crosses at the center of the X. To allow the
two points to move past this, I mitered the T-track together at the
center of the X. Each point, P1 & P2 are affixed to a T shaped slide.
What I mean is two separate T shaped pieces. I chose to use a steel
dowel pin in each of those sliders to locate my arm.
I know that if I were to size the trammel base to contain the point P2
so it remained inside the T-track, that when I would use the trammel
to guide my router, that the cutter may strike said base at the narrow
part of the ellipse. I got around this problem by sizing the base to
allow point P2 to go outside the base. You can do this by making the
sliding part long enough that you have several inches of it remaining
inside the T-track. Remember, as you move the arm around, and the
slides are sliding, all they have to do is miss at the center.
That's pretty much it. To get a better grasp of what I've attempted
to describe with words, I suggest that you make full scale drawing of
what I've tried to describe.
I have an Autocad drawing of all this from my own trammel design.
On Thu, 29 Nov 2007 14:36:30 -0800, [email protected] wrote:
>So by using the math I said before and reversing it. You can only
>have a 32 inch difference between the long axis and the short axis.
>
>I can't see a way to make the whole table with the undersized trammel.
>I can only suggest that you look at rebuilding it. Sadly I think the
>rebuilt trammel would be almost as big as the table top at about 50
>inches square
>
>R--------------------------------------------------P1----------------------P2
>
>For the home built trammel to work the distance between P1 and P2
>needs to be 24 inches. Where R is the router and P1 and P2 are the
>
>
>On Thu, 29 Nov 2007 18:47:12 +0000, Chris
><[email protected]> wrote:
>
>>I'm sorry I did not make it clear. I have made a board 32 inches
>>square with crossed dados ( not corner to corner ) in which bearings
>>run. however the centre where the dados cross is only 16 inches from
>>the edge of the square which would put the centre point of the trammel
>>outside the square when the end point was in the cross track.
>>
>>
>>Clear as Mud ?
>>
>>
>>Chris
>>
>>
>>
So by using the math I said before and reversing it. You can only
have a 32 inch difference between the long axis and the short axis.
I can't see a way to make the whole table with the undersized trammel.
I can only suggest that you look at rebuilding it. Sadly I think the
rebuilt trammel would be almost as big as the table top at about 50
inches square
R--------------------------------------------------P1----------------------P2
For the home built trammel to work the distance between P1 and P2
needs to be 24 inches. Where R is the router and P1 and P2 are the
On Thu, 29 Nov 2007 18:47:12 +0000, Chris
<[email protected]> wrote:
>I'm sorry I did not make it clear. I have made a board 32 inches
>square with crossed dados ( not corner to corner ) in which bearings
>run. however the centre where the dados cross is only 16 inches from
>the edge of the square which would put the centre point of the trammel
>outside the square when the end point was in the cross track.
>
>
>Clear as Mud ?
>
>
>Chris
>
>
>
On Nov 29, 12:50 pm, [email protected] wrote:
> Your close. You need to make one more part a t-track or dovetail
> trenched square sheet of wood. Add some standoffs under the trammel
> to account for the height difference.
I've used a framing square instead of a cross-slotted base
board. Clamp it down, draw one quarter of the ellipse, flip
it over, draw the next quarter, and so on. You can do the same
with two lengths of 1 x 4 butt jointed together into a sufficiently
large "L" if your framing square isn't large enough.
Your close. You need to make one more part a t-track or dovetail
trenched square sheet of wood. Add some standoffs under the trammel
to account for the height difference.
The only formulae I can find is to subtract the minor axis from the
major axis and divide it by two. That give you the distance you need
between the center sliding
This is because you need 2 centers to make an elipse. These will
slide in the slots cut in the wood.
In your case you would subtract 4 from 8, leaving 4 or 48 inches
divide tha by 2 is 24 inches. So your three points on your 48 inch
trammel would be your center point, one at 24 inches and one at 48.
Back to the center board. it will need to be a bit over 24 inches
square. (you don't want the guide points to fall out of it.) cut
your guide tracks diagonally across the guide board so it's a big X.
Secure the guide board to the center of the board to be cut into an
elipse. (using sticky tape or screwes if you are cutting from the
back or bottom.) Insert the guide points then connect them to the
trammel. Cut your elipse.
On Thu, 29 Nov 2007 17:06:38 +0000, Chris
<[email protected]> wrote:
>I've built a trammel (Less than 4ft square) to try to make an 8ft by
>4ft oval table.It works as a trammel but is too small using my
>current geometry to make an 8ft table. When I've looked at pictures of
>commercial jigs they are smaller than mine but make bigger tables. My
>slide arm is 4ft long with a central pin 2ft from the router. Is there
>a different formulae to get a scaling effect.
>
>
>Chris
Chris <[email protected]> wrote:
: I've built a trammel (Less than 4ft square) to try to make an 8ft by
: 4ft oval table.
It might be easier to use the string and nails method rather than the
tramel method. The hard part is finding string that will not stretch.
http://benchnotes.com/Laying%20out%20an%20oval/laving_out_an_oval.htm
--- Chip
On Thu, 29 Nov 2007 09:50:44 -0800, [email protected] wrote:
I'm sorry I did not make it clear. I have made a board 32 inches
square with crossed dados ( not corner to corner ) in which bearings
run. however the centre where the dados cross is only 16 inches from
the edge of the square which would put the centre point of the trammel
outside the square when the end point was in the cross track.
Clear as Mud ?
Chris
>Your close. You need to make one more part a t-track or dovetail
>trenched square sheet of wood. Add some standoffs under the trammel
>to account for the height difference.
>
>The only formulae I can find is to subtract the minor axis from the
>major axis and divide it by two. That give you the distance you need
>between the center sliding
>
>This is because you need 2 centers to make an elipse. These will
>slide in the slots cut in the wood.
>
>In your case you would subtract 4 from 8, leaving 4 or 48 inches
>divide tha by 2 is 24 inches. So your three points on your 48 inch
>trammel would be your center point, one at 24 inches and one at 48.
>
>Back to the center board. it will need to be a bit over 24 inches
>square. (you don't want the guide points to fall out of it.) cut
>your guide tracks diagonally across the guide board so it's a big X.
>
>Secure the guide board to the center of the board to be cut into an
>elipse. (using sticky tape or screwes if you are cutting from the
>back or bottom.) Insert the guide points then connect them to the
>trammel. Cut your elipse.
>
>
>On Thu, 29 Nov 2007 17:06:38 +0000, Chris
><[email protected]> wrote:
>
>>I've built a trammel (Less than 4ft square) to try to make an 8ft by
>>4ft oval table.It works as a trammel but is too small using my
>>current geometry to make an 8ft table. When I've looked at pictures of
>>commercial jigs they are smaller than mine but make bigger tables. My
>>slide arm is 4ft long with a central pin 2ft from the router. Is there
>>a different formulae to get a scaling effect.
>>
>>
>>Chris