Thread: Parabolic curve for pivoting router fence

Originally Posted by Dusty Dave

Not sure if this kind of thing has been discussed here much, I couldn't find anything in a quick search although I may be using the wrong terms.

Here is an old youtube video that shows the concept and the advantages, but gives nothing away as to how it is achieved - unless you follow some of the comments which point to a special curved slot at the pivot end of the fence. Some folks on other forums have tried to figure out the maths and exploded their heads in the process.

Dusty
what I saw in that youTube video is that at around the 2:00 mark, the demonstrator rotates the work piece through 180 degrees BEFORE completing the router cut. I stopped watching a few seconds after that point. If the demonstrator is rotating the work piece to demonstrate the fence's ability to "automatically" center the cut the demonstrator is using poorly concealed slight of hand.

It would work equally as well if the moving clamp is on a straight T channel with the clamp "F" able to move along the fence.

To work really accurately the pivot point at "A" would have to be centred on the face of the fence but since in most cases the size of the timber being routed is small compared to the distance "A to C" you may find that the error introduced by having the "A" under the body of the fence is so minute that it won't be of any concequence.

The distance "C to F" will in fact be close enough to twice the distance from the centre of the cutter at "B" with a straight clamping channel. Provided that the edge of the pin at point "C" is lined up the centre of the cutter.

I think that the curved clamping path is a complication that in fact gives no real benefit over a straight one.

Originally Posted by ian

Dusty
what I saw in that youTube video is that at around the 2:00 mark, the demonstrator rotates the work piece through 180 degrees BEFORE completing the router cut. I stopped watching a few seconds after that point. If the demonstrator is rotating the work piece to demonstrate the fence's ability to "automatically" center the cut the demonstrator is using poorly concealed slight of hand.

Ian, I think he was routing two opposite ends of the groove to show that they met perfectly in the middle - ie that they were perfectly centred. If they weren't, you'd notice the step where the two grooves met in the middle. You're suggesting he routed the same groove twice to make it look like a centred single cut. Maybe watch again and see what you think? I reckon he's only going halfway in from each end with each cut.

Ian,
By swapping the timber over at the 2 min. mark, wasn’t he just proving that when routing from both ends to the centre, the groove sides lined up in the middle proving it was central?
i noticed he didn’t rout the full distance, from one end to the other, when he turned it over which if done, may have covered up that it wasn’t cutting in the centre.

If he had just routed the full length, without rotating, how would we know it was in the middle?

Here's the patent, along with pics. Should make it easier to reconstruct: US5988242A - Pivoted fence with locating post
- Google Patents

Originally Posted by Bernmc

Ian, I think he was routing two opposite ends of the groove to show that they met perfectly in the middle - ie that they were perfectly centred. If they weren't, you'd notice the step where the two grooves met in the middle. You're suggesting he routed the same groove twice to make it look like a centred single cut. Maybe watch again and see what you think? I reckon he's only going halfway in from each end with each cut.

Originally Posted by Lappa

Ian,
By swapping the timber over at the 2 min. mark, wasn’t he just proving that when routing from both ends to the centre, the groove sides lined up in the middle proving it was central?
i noticed he didn’t rout the full distance, from one end to the other, when he turned it over which if done, may have covered up that it wasn’t cutting in the centre.

If he had just routed the full length, without rotating, how would we know it was in the middle?

Yes, I was wrong.
you are both correct, I stopped watching far too soon.

Originally Posted by Bohdan

It would work equally as well if the moving clamp is on a straight T channel with the clamp "F" able to move along the fence.

I don't think so, as there would then be nothing stopping the fence sliding freely along the line between points A and F. The curve is the secret sauce that deliberately moves the fence to cancel out the arc that would normally occur when 'swinging' a clamped workpiece over the cutter, allowing you to make a straight crosscut in a piece clamped to the fence.

I am inclined to agree that the inaccuracy from pivot A being under the fence rather than on the face of it would probably be small, and it shouldn't affect the crosscut path as the construction method would guarantee a straight path over the cutter regardless of the location of the pivot. Having said that, having a track on the face of a fence is no bad thing and it shouldn't be too hard to engineer a way to have it pivot around the face.