Drawing curves

[Mark Wybierala]

Lets say for example, you are wanting to draw full-size plans for a guitar neck and you want to incorporate a single action truss rod in a curved bed. What do you folks use to draw the curve? This would be a curve with something like a ten foot radius. The actual measured radius doesn't really matter but you just want a consistent curve that is going to be appropriate for the space you have. Is there are handy tool that you use?

I recently used a 36" long 1/4" dowel with a string tightening it into a bow but there was a slight concern that the curve might not be constant.

[Chuck Tweedy]

this site is close to helpful, but it looks like the circle radius is limited to something like 36 inch
https://www.blocklayer.com/circle-dividereng.aspx

[Chuck Tweedy]

i would use QCAD and print it out - but that is not an option for everyone

[Bob Gramann]

There’s always the old fallback method: get something stiff, one radius long. Peg one end. Attach a pencil to the other end. Draw. I’ve done it up to 25 feet.

[Barry Daniels]

I've got a french curve drafting template that is very useful for all sorts of curves.

[Steve Sawyer]

I had to draw very large-radius curves for the stretchers on an arts & crafts sofa and love seat I built. Basically I used a roll of that wire used for hanging pictures to rig up a long-radius compass. It is very light and easy to work with and does not stretch (critical). I used a couple of saw horses, but you can fasten the stock or your template directly to the floor with double-sided tape, and ditto to a piece of plywood or something for the pivot (a nail or screw). Use a long tape measure to get the pivot point equi-distant from the ends of the template or stock, put a loop in the wire to go over the pivot point, and a loop to go around your pencil or scribing knife and strike off your curve.

[Peter Wilcox]

In my experience a single action truss rod in a curved channel doesn't use a constant radius. As the neck gets thinner and narrower toward the head stock, a tighter radius is needed to exert more pressure to counteract the decreased stiffness of the neck, to keep the fret board at a constant relief radius. I remember one of the first necks I made, using a constant radius channel, had more curvature as it approached the head stock, making it impossible to get adequate action all along the neck.

I doubt there's a good way to calculate the change in radius without knowing the change in cross sectional area along the neck, and stiffness of the wood. Fender used 2 different radii for the channel, meeting somewhere near the midpoint of the neck length, similar to the truss rod plan in the link (third neck down).

https://www.electricherald.com/diagrams ... lan-01.pdf

[Brian Evans]

I put a tension bend in a rod or a long ruler, I use the string + peg = radius method, and I use double acting truss rods...

[Mark Wybierala]

I thought there might be some supercool way that ya’ll got this done — one of those impressive geezer techniques that I missed along the way. But it’s nice to know that I’m in the same boat as everybody else.

[Brian Evans]

Drawing a precise elipse or oval requires some pretty impressive geezering...

[Jim Hepler]

Mark, I think the geezer math you are looking for is something called a long compass. I remember some discussions about it in some threads about radius dishes and also perhaps sanding blocks for putting the cross section curve on fingerboards. In any event, search the archives for "long compass" and it should turn them up. I think it involves using a shallow "v" shaped apparatus with a pencil at the point which is slid across two pivot points. As I recall, the math tells you the angle of the v for a given radius.

[Bob Orr]

Use a two way straight rod?

[Steve Sawyer]

Mark, I think the geezer math you are looking for is something called a long compass. I remember some discussions about it in some threads about radius dishes and also perhaps sanding blocks for putting the cross section curve on fingerboards. In any event, search the archives for "long compass" and it should turn them up. I think it involves using a shallow "v" shaped apparatus with a pencil at the point which is slid across two pivot points. As I recall, the math tells you the angle of the v for a given radius.

"A neat property of this method is that the angle covered by the arc is exactly equal to twice the change in angle between the two strips of wood. To make an arc that covers 90 degrees for example the angle between the strips should be 135 degrees (180 - 90/2)."

https://woodgears.ca/shop-tricks/large_arc.html

[Mark Wybierala]

I’m simply making a neck with a conventional one-way curved compression truss rod. I have the side view on the neck drawn and I want to draw the curve as deep as I can without getting into trouble. Then I’ll copy the curve onto a lexan template. I’ve taken a 3/16” 24” long piece of rod and installed a tuning machine on one end so I can tighten a guitar string to adjust the curve. It works.

[Mark Wybierala]

I’m making the necks using a three piece sandwich method where the curved bed for the truss rod is precut to shape — it’s just a 3/16” thick center section of contrasting wood. The price of truss rods has me making my own. I’ve done this years ago and it works but you need to use templates to get a consistent result. In this case, I’m building four electric mandolins and there isn’t a lot of room.

[David King]

Mark,
Using a string to pull a straight rod into a curve will not yield a constant curve but something closer to a parabola as the string exerts it's greatest effect at the center of the rod and progressively less towards the ends. You can use this effect to your advantage by starting with a rod and string that are twice the length of your intended truss rod and tracing the apex of that curve at the nut and either end of the rod at the heel so that you end up with a progressively tighter curve from the heel to the nut.
There are Fender tele blueprints from 1987 out in the wild that show two different radii from the 20th fret to the 9th and from the 9th to the nut. They would be 255" and 115" respectively. Supposedly this was for an 80's "Biflex" rod but I've seen the plans and I believe the numbers go back to plans from the 1950s. Other plans show 148" constant curve. Some Strat plans show a constant radius of 164". Other plans specify an entry angle of 3º at the headstock and 2º at the heel. You want at least .188" of total deflection, preferably .200" if you have the room. Keep 3mm of meat at the back of the neck under the 9th fret and you will be OK.
Remember that the strings are moving in an elliptical pattern and you want your fretplane to follow that elliptical shape.

[David King]

For mando necks you probably don't need that much deflection on such a short neck even if you had the room for it. Gibson used straight truss rods that were parallel to the back of the neck and they worked. Eventually they went to a curved rod like on a Fender. Their earliest truss rod blueprint shows a reverse curve that really looks totally wrong and may have been an attempt to throw off their competition.