Z-Banding reduction using helical beam coupling (Page 2 of 2)

A number of posts and sites have recommended going to a skinnier rod to reduce banding, but has anyone tried taking this to the far extreme?  Stainless M3 rods are readily available, and a 5mm-to-3mm flexible coupling is only $5.70 shipped:

http://www.ebay.com/itm/Flexible-Coupli … 33734eb3c0

For that price, I figure I'd give it a shot just to learn something.  Is it too thin that is would be too easily bent?  I'll find out, but I'm betting that even if there were still some banding the fine thread pitch would make it barely noticeable.

Lawsy, the problem I saw when looking at ballscrews was the pitch. I was concerned that you would need to reduce the steps so much that layer height wold start to be become an issue.
Didn't sit down and do the math though as it all got a bit expensive for the gain.

I'm still trying to get a decent solution with oldham couplings at the print bed, it should be a relatively simple swap if I can get everything to fit correctly.

i try to do the same.. can you give the link where you find your ballscrew ?

do you take the 200 or 300 mm?

200mm

It will need shortening very slightly I think.

What's the nut and washer trick?  It looks like the nut would lock the rod to the coupler; does that effectively force the rod and coupler into better alignment because the nut and the coupler are acting on different "sides" of the threads on the rod?  (Sorry; I'm not familiar with some of the terminology.)

Well, I replaced the stock threaded rod with another 5/16" rod, though I have a few 8 mm rods lying around that I might use if I want to re-do it.  However, I am pleased with the results.  In order to get it to fit, I had to cut 1/2" off of the z axis motor shaft, which was a bit nerve-wracking; using my Dremel to cut lengths of five dollar threaded rods is one thing, but the motor is a beautiful little piece of equipment and I hated to cut it. 

Unfortunately, my extruder had previously been getting worse and worse, and I ended up taking that apart and unclogging it.  I'll just summarize that as "Mike's Fiery Kitchen Adventure." Here's something useful I learned: you can just heat up the barrel while pushing a piece of ABS through the feed, and as soon as the barrel gets hot enough, the entire plug of solidified crap will just shoot out of it, speeding things up immensely.

In any case, calibrating the z axis wobble resulted in a much lower variance.  For the delta between successive 0.1 mm turns, the distance in mm between the points (at a distance of about 12 feet) was:
7    6    6    7    7    6    7    7    6.5    7    7    6.5    8
... whereas with the stock rod, my previous data from about a month ago was:
12    14    11    9    8    6    5    4.5    5.5    4.5    5.5    8    10    11    12
The differences in magnitude between the different sets is meaningless, but the variance within each set is immediately apparent.

Of course, after unclogging the extruder, there are a lot of things that need calibration, but after a long while, here's a photo of the results (a 25mm cube). 
[I'm not allowed to post links, so I'm trying to attach the picture.]
[The preview doesn't show that the attached picture is attached, so here's link without the prefix:
h t t p colon slash slash img547.imageshack.us/img547/9418/7wsu.jpg ]

Also know as PEM nut. See distributors list.
http://www.pemnet.com/fastening_product … r_new.html

You are looking for SP (standard profile) or CLA (specific to aluminum mounting)
This data sheet has pretty much everything you need to know and also a bunch of stuff you don't.
http://www.pemnet.com/fastening_products/pdf/cldata.pdf

I'm pretty happy with the 5/16" replacement, though that may change with time.**  Now I've actually been going around replacing prints I'd given to people when I still had the horrible banding problem with the much improved prints I'm getting now.  The replacement rod is actually somewhat bent as well, but with the coupler absorbing much of that, the two vertical guide rods keep the bed in a straight line.  I still have some wobble, but as I'd mentioned previously, the variance in successive 0.1mm layers was much smaller than with the stock rod.

I am now a big believer in $65 stainless steel helical couplings.

** I assume it will never end; there will always be things to improve.  I still have some banding, but the quality is improved enough that people ask if I've used some sort of acetone vapor treatment when touching my new prints.  I'm no longer obsessing about the banding; just getting the added structural integrity from having equal-height layers makes things come out very solid.  I still may go for an 8mm rod, but I have a vague feeling I'll be obsessing about adding a second extruder first...