Drafting a spiral?

Parametrically, you want r(theta) =3D pitch * theta / 2pi, with theta =3D (0, N * 2pi), N turns, for theta in radians. The centerline of the trace follows this path, so you need to offset it (either side, perpendicular to the path) by half the width to get the actual profile. Then add caps on the endpoints to make a contiguous perimeter.

If all you have are circles and splines, you might get away with increasing radius semicircles joined end-to-end. It will look wrong, but the overall effect will still be right.

Tim

If this is for hobby or non-profit use: Cadsoft Eagle has a free version that allows small circuit board sizes (but I guess you could zoom that up with some other SW for the antenna if needed). Install that. Then download "print-inductor.zip" from here:

I have not used this particular download file myself but it should unpack a *.ulp file (User Language Program) which creates spiral PCB inductors. This ULP file needs to be copied into the same directory where all the ULPs are that come standard with the software. File should be safe and tested quite well, Mr.Zaffran is one of the engineers who works at Cadsoft.

Try the schematic editor as well, maybe you like it.

You could probably do it using Postscript, which is actually a powerful drawing language in its own right. An online book probably has everything you need:

You design a form on the computer that you can make in manufacturing. The form is a cardboard (paperboard) helix that you print out, and roll up onto a tube, so you can paste it onto a cardboard tube. The print job prints the line of your antenna onto the paper. The finished tube has lines on it now that you use a form for the positioning of the antenna elements. Once everything is assembled, you remove the tube from within the antenna, and all that is left is the antenna itself.

So the CAD design is for the printed helix before it gets rolled up onto the tube form that is the diameter of your helix.

Done right, it will have your layout lines in all the right places so that the antenna builders can build the antenna. Automated process can be derived from similar fixtures and forms methods, but this antenna style sounds difficult to automate too much.

old_x = 0 old_y = 0

for turns = 1 to 32 for theta = 0 to 359 r = (360 * turns + theta) / (150/32) new_x = r * cos(theta) new_y = r * sin(theta) drawline (old_x, old_y, new_x, new_y) old_x = new_x old_y = new_y end for end for

or something like that...

It's a cheap but excellent PCB layout program that has spirals and polygons built in. Set turns, track width, start radius,distance

Others have suggested good ways to do the figuring of the spiral. Making it into Gerber plots is another step.

The GEDA PCB program stores its PCB layout in an easy to understand format. You can make your program produce the spiral in a file that you can then copy and paste into the ASCII file that is the layout. Then you can make Gerber plots. This way, you can do other parts too.

I think you want: (turns + theta/360) * (150/32)

For gerbers:

MOVE(0,0); // D02 for turns = 1 to 32 for theta = 0 to 359 r = (turns + theta/360) * (150/32) new_x = r * cos(theta) new_y = r * sin(theta) DRAW(new_x, new_y); end for end for

With a compass, I've always done the "Fibonacci Spiral"...

...Jim Thompson