Z shift? any proven solution??

This is the place to show off your felix printer hacks and improvements.
Post Reply
stratogavster
Posts: 120
Joined: Tue Jul 17, 2012 9:04 pm
Location: United Kingdom

Z shift? any proven solution??

Post by stratogavster » Mon Nov 26, 2012 10:12 pm

Hi again all....

Have spotted z-shift / distortion obviously more noticeable in parts with vertical walls....

Wondered if anybody had thoughts on the real source of this.....?

I've read people experiencing a bent z-screw etc, though i'd like to add that perhaps the repeatable positional accuracy and the slack in the z-slide are also significant factors? (wobble enabled by said 'slack' and bent z-screw).

Hoping to solve it but hoping others have run a few iterations through the mill before I commit to anything....as currently I can only see expensive solutions ahead (trapezoidal z-screw and better slide with no slack).

Is there any mileage in the backlash setting is SFACT to help mitigate this to come extent?

:D
Attachments
z_shift.jpg

Kjetilei
Posts: 334
Joined: Fri Mar 16, 2012 8:26 pm
Location: Stavanger, Norway

Post by Kjetilei » Mon Nov 26, 2012 11:34 pm

A wild guess would be that it is caused by imprecise positioning of the extruder due to x-/y-axis backlash.

You can experience a similar phenomenon if you continue to print a part with imperfect 1st layer. The print will start out wrong but will steadily fall in line with the intended path.

To see if it is a z-axis problem it would be good to print a taller part like a square ore something and see if you get a repeating pattern.

Although annoying, is it critical for the kind of application you intend to use the parts? Maybe it is ok to live with the imperfection (if the only solution is expensive upgrades). But it could also be a trivial/cheap reason for the imperfection :)

stratogavster
Posts: 120
Joined: Tue Jul 17, 2012 9:04 pm
Location: United Kingdom

Post by stratogavster » Mon Nov 26, 2012 11:43 pm

Hi,

Where this is an annoyance is that some of my parts have subtle draft angles....2 or 3 degrees sometimes and this Z shift or drift as it were causes issues there....add to that my lack of knowledge with SFACT / SKEINFORGE thus the standard profiles are giving insufficient and generally damaging support material.

Most of my parts are generally not very tall and are extremely complex is shape, so i'm hoping for high accuracy and repeatability....

Longer term had aimed for some self-upgrades though any tips in the meantime are of course superb too :)

Kjetilei
Posts: 334
Joined: Fri Mar 16, 2012 8:26 pm
Location: Stavanger, Norway

Post by Kjetilei » Mon Nov 26, 2012 11:47 pm

Maybe you could share one of the troublesome parts with Guillaume (or us) and he/we could try to print it on other printers to see how much difference there is?

I love my 1.5 upgrade but still haven't had the time to fine tune it but based on the results Guillaume has achieved with the old 1.0 setup, high accuracy and repeatability should be possible with a fine tuned 1.0 as well.

stratogavster
Posts: 120
Joined: Tue Jul 17, 2012 9:04 pm
Location: United Kingdom

Post by stratogavster » Mon Nov 26, 2012 11:57 pm

Will certainly plan to....had already ran a challenging part past Guillaume before I bought my machine, so will try and share a more recent one...

V1.5 hopefully on it's way very soon!! never was fond of the IGUS linear stuff :?

In an effort to combat the z-shift I have got hold of a super-straight piece of M8 stainless threaded bar which I'll cut and fit....and I may add a second guide bearing just about the z-coupling. Incidentally my z-coupling has never been the stock printed one but an expensive SOT type (sorry bad pic from web-store)
1276_SOTsmall.jpg
1276_SOTsmall.jpg (6.13 KiB) Viewed 1000 times

Trnquill
Posts: 66
Joined: Fri Oct 05, 2012 10:01 am

Post by Trnquill » Tue Nov 27, 2012 11:17 am

I had quite visible waving along Z-axis (vertical walls) when I did my first prints. The "wave length" of the waves (ie. repeat interval of disformation) was quite accurately 1,25mm, which just happens to be the raise of Z-axis screw for one full turn. My Z-screw was visibly bent and I couldn't make it straight no matter what I tried.

Things I did to remedy the waviness:
1) Replaced the Z-screw with a straight one. 1m of quality threaded rod cost only few euros.
2) Replaced the Z-screw coupling with similar (but all aluminum) to what stratogavster showcased in the previous post
3) Got rid of Z-slider play by adding more tape inside the slider, between the aluminum body and nylon (?) sliding element.

Now my prints have perfect walls, no waving or wobbling any more.

Kjetilei
Posts: 334
Joined: Fri Mar 16, 2012 8:26 pm
Location: Stavanger, Norway

Post by Kjetilei » Tue Nov 27, 2012 5:25 pm

I'm not certain that the problem shown in the photo is due to a wobbly z-axis. Although I can see something that looks like a repeating pattern every 5th(ish) layer on the left photo, it is the right close-up photo that shows the main problem. My bet for the main problem would be backlash but I might very well be wrong.

User avatar
gfeliksdal
Site Admin
Posts: 402
Joined: Sat Feb 25, 2012 10:40 pm
Location: Netherlands
Contact:

Post by gfeliksdal » Tue Nov 27, 2012 5:54 pm

Nice couplings. Where do you get them? And are they reasonably priced.
Hard to see what's going on in the picture. It could also be that the heated bed is too hot. That can give some deformations too. Or that the first layer is squashed too much.
To optimally determine the backlash values for your x and y-axis. You could print two little cubes at a distance of approx 100mm from eachother, Try to vary the lash parameters in skeinforge and I'm sure you will see a big difference in the results, the difference is very well noticable with a Felix 1.0.see the attachment. It is also a nice test to see if you have the hot-end temp set correctly or the retraction settings right.

I'm currently looking to replace the acme rod, which has a bigger diameter (10mm) for the z-rod. To get a perfectly straight 8mm rod with a length of 330 mm is almost impossible, especially after cutting it to length and afterwards mounting it. Also the small pitch makes the wobble better noticable if it is not tuned correct. If you know any places were to get low costs rods including nuts, I'm all ears :)
Attachments
leakage_test_v2.STL
(1.25 KiB) Downloaded 48 times

satoer
Posts: 436
Joined: Mon Oct 01, 2012 8:00 pm
Location: Netherlands

Post by satoer » Tue Nov 27, 2012 7:47 pm

Have you considered upgrading the Felix with Ball screws Guillaume? There probably way more expensive than the cheap nut and threaded rod solution. But I think it’s the best solution to get rid of all the wobbles.
Image
My Felix 1.5 printed designs on Thingiverse
E3D V6 Hot end upgrade for your Felix
Tweezer bracket for your Felix
Led bar upgrade for your Felix

stratogavster
Posts: 120
Joined: Tue Jul 17, 2012 9:04 pm
Location: United Kingdom

Post by stratogavster » Wed Nov 28, 2012 12:46 pm

Nice couplings. Where do you get them? And are they reasonably priced
My particular coupling can be found at http://www.worldofcnc.com/couplings/cla ... 1_382.html

Really great guys and very helpful. Not cheap but awesome quality! (around £17).

Also they have all manner of lead screws and nuts too for the future Z upgrade! :)

Post Reply