Inaccurate cuts, with a twist! [RESOLVED - DULL BIT]

I’ve been trying to do some fairly exact cuts lately (guitar necks and fretboards), and I am not satisfied with the accuracy I’m getting. Things come out wider by a millimeter or two every time.

I have checked that the X-Carve (the 1000 x 1000 w/ Dewalt router) is square and that when not making a cut things run well. I get exact distances traveled to as close as I can eyeball with a metric ruler over distances of ~600mm on both X and Y.

To better zero in on where the problem is, I designed a simple object to carve. It’s a 50mm x 50mm square (19mm tall, but not a factor). What I get as a result though comes out repeatedly as 50.2mm x 51.17mm, within a few hundredths of mm.

So, Y is a little worse relative to X, which is itself a little off.

This was with a soft pine for stock, and is actually less severe than what I am seeing when using maple and rosewood, so I think the material is a factor. But why differing amounts of inaccuracy on each axis?

Now, the twist!

When I ran the test square the first couple of times, it was with the wood grain running along the X axis. So, using an off-cut from the same piece of pine, I oriented the stock so that the grain ran with the Y axis.

And I got a different result.

Now it’s 51.72mm x 50.5mm. It would have been nice if the dimensions flipped without each getting slightly worse, but the effect still seems apparent. This is also incidentally the grain orientation i use for my necks and fretboards, which have the corresponding inaccuracies on X too…

The thing I don’t know is why. My uninformed theory is that when turning a corner so as to become perpendicular to the wood grain, the material is forcing the bit slightly further out from its intended path. But the cuts are being done in multiple shallow passes of only 1mm each, so it seems kind of nuts.

Any thoughts as to what I can check next? I am going to cut some larger test squares, and I am assuming the amount that it is off by is constant, and not proportionate to the object I’m carving. Either way, I’m, not sure what to adjust.

By the sounds of it I think you are on the right track, the sideway force when the bit is carving is causing the path deflection.

If you place a 2,5 and 5kg sideway force in Z and then Y direction, what kind of deviation do you read?

I would see if there is anything that can be improved on the Z and Y mechanically improving overall rigidity. The stock Z axis is decent but not rigid by any means.

I do not have a way to exert exact amounts of force the way you mentioned, but I did notice a little too much wiggle when pushing against the bit. I had not tightened everything up as much as I should have around the router holder when I adjusted it to be properly perpendicular to the wasteboard recently, so I fixed that.

Unfortunately, while I was at it, I tightened a few other things too, like some rollers on X and Y. Things are not better now, just different. I probably should have just tightened around the holder and measured again.

What is the typical way to perform a test like you’d mentioned?

Dial indicator and weight+wire hanging over the edge (with a pulley if possible if you want to minimize friction)
But I wouldnt go too technical about it, its just a simple test to du by hand to realize that there is movement.
Having some data will provide feed back on any changes made later, if so inclined :slight_smile:

Well, this is a little embarrassing, but I’ll put it up here just in case anyone has the same “issue”.

Ultimately, it was a dull bit.


Yup, it was struggling through its cuts, more so when working across wood grain, and while there may have been some tweaking that helped too (tightening up some wheels, etc…), the big change was when I put on a fresh bit.

Now my cuts are very close (tenth of a mm and less), and that’s good enough for me.

Unfortunately the Z axis on stock machines is fairly flimsy. Because of this things like sharpness of bit, grain of wood, material hardness, tramming, etc. all can cause inaccuracy. I ran into these issues several times. Even with a very rigid modified machine and linear Z I still get pockets out of spec due to deflection and tool stick out. Although, most of what I do is aluminum. It’s a never ending battle.