Eliminating Hadron z wobble

Topics Related to the ORD Bot Printer

Re: Eliminating Hadron z wobble

Postby orcinus » Fri Dec 21, 2012 1:57 pm

gyrogearloose wrote:Of course you do. Flexible couplers are not the answer.


Not true. They *are* the answer.
Allow me to explain...

This:
The 2 grub screws, once done up, will push the rod slightly sideways or even at a slight angle. If the grub screws were spaced at 120deg, it would probably help.


... does not happen with flex couplers. Because they clamp the rods instead of pressing on them laterally with grub screws.
Of course, if you have a bad flex coupler with the hole(s) drilled off-center, you're back at the beginning. *

But yes, flexible couplers most certainly are a solution for the very problem you've described.
I've initially solved it with the original black couplers by shimming the rod inside the hole with a folded piece of heat shrink tubing. Clamp-type couplers were the permanent solution.


* Coincidentally, after i've switched from 8mm rods to 6.35mm ACME rods and corresponding adapters, i got some wobble back on one side only, due to a badly machined flex adapter (the 5mm hole and the 6.35mm hole have not been drilled concentrically), so it's back to shimming :)
orcinus
 
Posts: 720
Joined: Tue Feb 14, 2012 4:03 am

Re: Eliminating Hadron z wobble

Postby cvoinescu » Fri Dec 21, 2012 2:01 pm

That looks amazingly better, although there is still some wobble (you can see it best in the shadowed area).
cvoinescu
 
Posts: 501
Joined: Thu Aug 09, 2012 9:12 am
Location: Camberley, Surrey, UK

Re: Eliminating Hadron z wobble

Postby cvoinescu » Fri Dec 21, 2012 2:09 pm

orcinus wrote:... does not happen with flex couplers. Because they clamp the rods instead of pressing on them laterally with grub screws.


It's true that the flexible couplers clamp the rod, but not true that it solves the problem. If you examine a coupler carefully, you will notice that one half of that clamp is attached much more rigidly to the rest of the coupler than the other half. So when you tighten that screw, one jaw closes more than the other, and the hole moves off-center. However, there are flexible couplers that have both the clamping action and a grub screw, and (I think by sheer dumb luck and not any smart engineering from our Chinese eBay friends) the grub screw is on the jaw that moves less. Using both the clamp screw and the grub screw, it is possible to get the rod very close to being centered.
cvoinescu
 
Posts: 501
Joined: Thu Aug 09, 2012 9:12 am
Location: Camberley, Surrey, UK

Re: Eliminating Hadron z wobble

Postby orcinus » Fri Dec 21, 2012 2:12 pm

You're right, mine had both (although in my case, clamping them shut put the rods dead on center, while using the clamp and the set screws made it go off center - might have been sheer luck).
orcinus
 
Posts: 720
Joined: Tue Feb 14, 2012 4:03 am

Re: Eliminating Hadron z wobble

Postby gyrogearloose » Fri Dec 21, 2012 10:43 pm

cvoinescu wrote:That looks amazingly better, although there is still some wobble (you can see it best in the shadowed area).


I have the part right in front of me, and I don't see any "wobble". If the z axis wobbles the layers become uneven. On the part, you can count the layers (4/mm) and they are all the same.
gyrogearloose
 
Posts: 65
Joined: Tue Dec 18, 2012 10:57 am

Re: Eliminating Hadron z wobble

Postby gyrogearloose » Fri Dec 21, 2012 10:58 pm

orcinus wrote:
gyrogearloose wrote:Of course you do. Flexible couplers are not the answer.


Not true. They *are* the answer.
Allow me to explain...

This:
The 2 grub screws, once done up, will push the rod slightly sideways or even at a slight angle. If the grub screws were spaced at 120deg, it would probably help.


... does not happen with flex couplers. Because they clamp the rods instead of pressing on them laterally with grub screws.
Of course, if you have a bad flex coupler with the hole(s) drilled off-center, you're back at the beginning. *

But yes, flexible couplers most certainly are a solution for the very problem you've described.
I've initially solved it with the original black couplers by shimming the rod inside the hole with a folded piece of heat shrink tubing. Clamp-type couplers were the permanent solution.


Thank you, you just proved my point. the "flexible couplers" don't work because they are flexible, but because the way the clamp the rod helps keeping the rod centered. With my solution, the rod is centered to the coupling by default because the coupling acts like a nut that is screwed onto the end of the rod. What we are building is a precision machine with an expected accuracy of a fraction of a mm (A resolution of 0.15mm on the part requires a mechanical accuracy of 0.015mm on the machine). Any engineer worth a pinch will tell you that. For that to happen, everything has to run true! You cannot expect a flexible component to compensate for the inaccuracy of another component. This will never work, believe me. If you can see even the slightest up and down movement on the motors when jogging the z axis up and down, you got to go back to the drawing board. The rods have to be straight to start with, they have to be adjusted exactly parallel in reference to the frame and they have to be centered to the coupling. That' all, no need to build extra sliders to try and keep it straight.
gyrogearloose
 
Posts: 65
Joined: Tue Dec 18, 2012 10:57 am

Re: Eliminating Hadron z wobble

Postby cvoinescu » Sat Dec 22, 2012 10:42 am

gyrogearloose wrote:
cvoinescu wrote:That looks amazingly better, although there is still some wobble (you can see it best in the shadowed area).


I have the part right in front of me, and I don't see any "wobble". If the z axis wobbles the layers become uneven. On the part, you can count the layers (4/mm) and they are all the same.


slight-wobble-detail.jpg


That's where it's most visible. It's also very subtly noticeable in other parts of the image. Don't get me wrong, it's a good print! I was just stating the obvious (to me), that the wobble has been greatly reduced but not eliminated.
cvoinescu
 
Posts: 501
Joined: Thu Aug 09, 2012 9:12 am
Location: Camberley, Surrey, UK

Re: Eliminating Hadron z wobble

Postby fma » Sat Dec 22, 2012 5:55 pm

Almost perfect!

I checked the Z movement: with increments of 0.1mm, for 1 complete turn (Z=1.25mm), I got several 0.09mm, then a few 0.11mm. Do you think this is enough to see something? My results are not as bad as your first ones, but not as good as your last ones...
Frédéric
fma
 
Posts: 213
Joined: Mon Aug 20, 2012 7:53 am

Re: Eliminating Hadron z wobble

Postby cvoinescu » Sat Dec 22, 2012 8:52 pm

That depends on how the 0.09s and 0.11s are grouped. If you have 0.09, 0.09, 0.09, 0.09, 0.11, 0.11, 0.11, 0.11, 0.11, then it's going to look worse than if they alternate or are mixed randomly.
cvoinescu
 
Posts: 501
Joined: Thu Aug 09, 2012 9:12 am
Location: Camberley, Surrey, UK

Re: Eliminating Hadron z wobble

Postby orcinus » Sat Dec 22, 2012 10:31 pm

gyrogearloose wrote:Thank you, you just proved my point. the "flexible couplers" don't work because they are flexible, but because the way the clamp the rod helps keeping the rod centered.


Exactly - flexible couplers such as those were made to cancel out issues caused by non-parallel shafts, and to some smaller extent, non-concentric shafts. Think cardan (universal) joint. A bent rod is something completely different, flexibility of the joint won't help you with that, some motion is bound to get lost to deformation no matter what.

But that's okay, because most of the wobble people initially had was caused by the shafts not being concentric, because of the set screws in the original couplers, and only a small portion of it is due to bent rods.
orcinus
 
Posts: 720
Joined: Tue Feb 14, 2012 4:03 am

PreviousNext

Return to ORD Bot

Who is online

Users browsing this forum: No registered users and 44 guests

cron