pirvan wrote:

I suggest you read this wiki page on banding:

http://www.soliwiki.com/Banding

Pay particular attention to this part:

"As a primary cause with readily available free solutions, backlash should be the first problem to be addressed when attempting to reduce both random and periodic banding."

For my part, I had pretty noticeable banding, which was reduced substantially when I installed a well nut on top of the assembly (http://www.soliforum.com/topic/2341/poo … -required/).  I eventually installed the ballscrew, which reduced banding down to nothing.

It was then, that I noticed the periodic ripple.  The surface is nice and smooth, but when you look at the piece at an angle, or when you photograph the piece from and angle with a flash, it's then that you notice the slight ripple.  Looking straight on it's not really noticeable.  That is the wobble banding.

Very interesting fix.  I will give this a shot before continuing on.  Thanks.

After applying this fix, the 18 lines per inch (which matches the threads per inch of the Z drive screw) actually increased.  It surprised me, but thinking about it made sense in that I am now more tightly bound to the drive nut, and the backlash was previously "absorbing" some of the lateral wobble.  So.. while still and excellent fix, it is back to anti wobble work.

I suggest you read this wiki page on banding:

http://www.soliwiki.com/Banding

Pay particular attention to this part:

"As a primary cause with readily available free solutions, backlash should be the first problem to be addressed when attempting to reduce both random and periodic banding."

For my part, I had pretty noticeable banding, which was reduced substantially when I installed a well nut on top of the assembly (http://www.soliforum.com/topic/2341/poo … -required/).  I eventually installed the ballscrew, which reduced banding down to nothing.

It was then, that I noticed the periodic ripple.  The surface is nice and smooth, but when you look at the piece at an angle, or when you photograph the piece from and angle with a flash, it's then that you notice the slight ripple.  Looking straight on it's not really noticeable.  That is the wobble banding.

Very interesting fix.  I will give this a shot before continuing on.  Thanks.

http://www.soliwiki.com/Banding

Pay particular attention to this part:

"As a primary cause with readily available free solutions, backlash should be the first problem to be addressed when attempting to reduce both random and periodic banding."

For my part, I had pretty noticeable banding, which was reduced substantially when I installed a well nut on top of the assembly (http://www.soliforum.com/topic/2341/poo … -required/).  I eventually installed the ballscrew, which reduced banding down to nothing.

It was then, that I noticed the periodic ripple.  The surface is nice and smooth, but when you look at the piece at an angle, or when you photograph the piece from and angle with a flash, it's then that you notice the slight ripple.  Looking straight on it's not really noticeable.  That is the wobble banding.

So I'm guessing that the idea is to allow the nut to oscillate or "wobble" in the X Y axis while the whole assembly is fixed to the print bed.  How are you going to fix the top of the assembly to prevent torquing.  For this to work right, the top and bottom of the assembly need to be perfectly rigid with respect to the surface they're bolted to (bed platform).

Another thing is backlash.  Your device doesn't account for that, and most of the visible vertical banding is a result of backlash.  Z wobble is much less pronounced and is something you usually become aware of after you fixed the backlash, simply because you now have a very smooth surface, with a periodic surface wave that occurs like clockwork, every so many layers.

Are you sure it's the wobble you want to fix, or is it backlash?

Viewing the efforts and comments of others on this forum, it appears that backlash is best handled by firmware.  The efforts to counteract the wobble have run a gambit of retrofits.  I believe there are many ills to attack, but it seems that the first one most folks here are trying to solve is the wobble first, others later.  This design, while not addressing backlash, should not introduce any due to the tight fit between the nut holders and the plates, especially the bottom plate.  The top plate should be held in rigid relationship once proper sized spacers are in place and all four screws are tight.

Thanks for the comments, this type of feedback and discussion is what I hoped for to improve the design as much as possible before committing to making any changes.

Another thing is backlash.  Your device doesn't account for that, and most of the visible vertical banding is a result of backlash.  Z wobble is much less pronounced and is something you usually become aware of after you fixed the backlash, simply because you now have a very smooth surface, with a periodic surface wave that occurs like clockwork, every so many layers.

Are you sure it's the wobble you want to fix, or is it backlash?

Can you explain how this would go together and the principle behind its functionality.   I may be a bit thick, but as is, without an explanation, I can't really figure it out.

Certainly.  Here is an image which may help.  The nut is capured between end caps which are seated into the matfhing hex cutouts in the plates.  This prevents the nut from turning, but should alllow it to move laterally.

Four spacers (not shown) would be placed between the plates at the bold holes.