Z wobble / banding / backlash fix that worked for me

Hey everyone!   I've been thinking about my fix,  and why it works...

I'm no scientist..  But correct me if I'm wrong, the basic issue with a bent rod or even the minutiae of backlash is the print bed moves left, right,  forward and back during the z axis change.  This is due to all sorts of physics that have been explored in depth by other people. 

Solidoodle tried to combat this issue with two smooth z axis rods,  which in a perfect world would have eliminated the wobble.  (I'm using the term wobble generically)

But due to the play in the bed bushings relating to these rods,  the wobble remains. 

The wobble is there because of the play in the bed, you never know where the bed will settle after a z axis change.

I played with the print bed while I printed,  and noticed the wobble was reduced significantly when I pressed the front of the bed to the extreme up,  left, right or down position. 

This is due to when I pressed on the front of the bed,  it maxed out the play in the bushings.

I figured I could get the same effect using springs. 

I decided to mount the springs like I did,  simply because of the availability of the bed leveling screws and the front belt screws. 

I'm sure you would get the same effect if you were to drill holes in the bed support and the bottom of the frame. 

The springs simply remove the play in the bed bushings.  Therefore removing the wobble. 

I'm thinking of removing the springs and using two adjustable braces that press on the smooth rods and force the bed to one of the z extremes. 

I think this will work because the rods and the bushings are circular and if you push hard enough,  the bed will settle in a predictable location

*follow*

Congrats on exploratory work!

It seems to me that this is very similar to some of the spring loaded anti-backlash devices, except by loading the bed you not only reduce play in the main rod but do the same for lateral movement in the bushings at the same time.  I personally hadn't noticed a lot of banding that I would expect from horizontal movement (shifted, like a wobbly stack of coins instead concentric rings of differing size) but anything that removes looseness from the system sounds like a good idea.

My only concern is that using springs makes the loading force dependent on the bed position, increasing the strain on the bed as it gets lower.  Using something that provided a constant force seems like it would be safer.  Perhaps the spring could be attached to an arm that was threaded onto the rod at one end and pulled the bed from below.  Alternatively, someone in the past has suggest just weighing down the bed.  It seems like hanging some lead weights off the near end would have much the same effect but would provide a constant force downward.

Well everyone,  I attached the springs to the front case screws (bottom)  and the wobble is back.  I'm going to reattach them to the top,  as I had perfect printing with them in that position.

That throws all physics out the window.  I can't think of a reason that when they are attached to the top it is perfect,  but when they are on the bottom it goes to hell. 

Any thoughts?

@pesce Interesting fix, do you see platform banding when moving up and down z?