Router placement.

[dstranz]

The resistance of the workpiece to being cut by the router bit results in torque about the point where the router head is clamped to the Z-stage. The direction of the torque depends on the direction the bit is traveling at any particular time, but will be exactly 180 degrees from the point of contact of the bit with the work. If the router head is clamped as far down as it will go, this torque will be greater than if it is clamped higher up.

The physics is described by a lever arm - if you balance a beam on a fulcrum (pivot point) such that an equal length of beam is on either side (like a see-saw), then you can make it balance with an equal weight on each end. If you put the fulcrum point much closer to one end, then it takes a much heavier weight on the short side to achieve balance; conversely, it takes a much lighter weight on the long side to cause the short side to deflect. In this case, the router head is the beam, and the clamp is the fulcrum.

This is exactly what is happening when you mount the router head far down in the clamp - the lever action of the long shaft of the router causes much greater torque on the clamp. Since the clamp can't move, the torque is transmitted to the screws and rods of the Z-stage and gantry. It doesn't take a lot of cutting resistance to cause deflection.

If instead you make the business end of the router head the short side of the lever by raising the router up in the clamp, then it takes a far greater cutting resistance to result in significant deflection of the stage and gantry.

Even when the router is not cutting, placement still will have an effect. The router mount is cantilevered off of the gantry, so there is another fulcrum where the Z-stage is attached to the X rods. If the router head is low in the clamp, there is a much longer lever arm because the center of mass of the whole Z assembly is farther away from the pivot point. Moving the router up as far as possible raises the center of mass and decreases this lever arm, thus reducing the torque applied to the Z-stage attachment point.

[Wolffie1]

I have mounted my router with the bottom edge as close to the clamp as possible.
I do not have nor do I want a dust boot, the air from the router blows the dust away from where the router is cutting.
I am in the process of making another clamp to be positioned higher up in order to keep the router level at all times.
Just to add:
I am using a big 1/2" router, hence the need for a support clamp.
Cheers
Wolffie

[rungemach]

A lot of good discussion on the router issue so far in this thread, all good points.

Please allow me to add that IMHO, plastic is not the ideal material for the router mount. It is by nature slippery, and also loses its grip over time as the plastic slowly deforms under pressure. (creep) If you use the stock clamp, you are best off mounting the router as high up as possible to minimize the twisting forces from the bit moving through the material.

The next step up is to make a clamp from a more suitable material. I use aluminum. The stock size clamp in aluminum is a big improvement.

Better yet is the billet clamp that Sam at Dixie Billet makes. I believe he charges around 50 dollars for a clamp that is over 1 inch thick billet aluminum. Sam can be reached at business@dixiebillet.com

With a strong clamp, the router position becomes less important.

Just changing from plastic to aluminum will make a big difference

Ultimately, upper and lower mounts (with the top mount fastened to the Z axis plate) are the best option for a smooth cut and a less flexible z axis plate.

some discussion of HD clamps can be found around page 16 of this thread as well.
http://www.cncsharktalk.com/viewtopic.p ... &start=150

[dstranz]

I like the idea of a separate top clamp fixed to the Z-axis plate. That should eliminate most of the torsional forces around the lower clamp because the top of the router is now locked in position.

How do you attach the top clamp to the Z-axis plate?

I guess if you are using aluminum for this you can drill and tap, but given the propensity of plastic to creep I'm not sure whether that would be viable if I used the original clamp instead. I suppose I could mill another plywood clamp and epoxy a nut into the side of it or mill a slot for a captured nut.

[rungemach]

You could use a piece of aluminum angle and have it attach to the clamp rear bolts and then bolt it to the back plate.
This picture shows a prototype I made of a stacked clamp arrangement.