Has anybody tried to straighten the screw?
Just put it on a sheet of glass,  find the high spots and tap down then down on something soft like wood using a plastic hammer.

In many cases part you'd want alter is the threads of the screw.

Got before and afters?

Here's a photo attached. 

There are lots of photos from other folks with more extreme examples on the z-wobble-compensation thread, but I had an unusual case of a nonlinearity that no not only had a sinusoidal component, but also had a discontinuity that caused regular groove in the output of .1mm prints.  I got great results measuring the bed position with a laser pointer setup and using it to define a custom compensation curve:

http://www.soliforum.com/topic/1165/fir … pensation/

ccorwin wrote:

Has anybody ever made a third rail at the door side of the table? I think this might help with z axis wobble. I might print some parts out to see if it will work. I think in the front left or right corner with an arm atatchment to the underside of the table.

While your case may of course be different, I believe most of the banding people experience is not due to actual lateral movement in the bed, but instead due to z-axis moving in a non-linear manner when the z-rod presses sideways against the nut, which causes layers with unequal heights.  Many people (myself included) have now effectively eliminated banding using Rincewind's modified firmware that lets you calibrate z-axis movement to compensate for the non-linearity.

Got before and afters?

Has anybody ever made a third rail at the door side of the table? I think this might help with z axis wobble. I might print some parts out to see if it will work. I think in the front left or right corner with an arm atatchment to the underside of the table.

While your case may of course be different, I believe most of the banding people experience is not due to actual lateral movement in the bed, but instead due to z-axis moving in a non-linear manner when the z-rod presses sideways against the nut, which causes layers with unequal heights.  Many people (myself included) have now effectively eliminated banding using Rincewind's modified firmware that lets you calibrate z-axis movement to compensate for the non-linearity.

Currently the third rail is removed, as I am replacing that with a proper rod and linear bearing.