Eliminating Hadron z wobble

Topics Related to the ORD Bot Printer

Re: Eliminating Hadron z wobble

Postby fma » Wed Dec 19, 2012 10:08 pm

BTW, it could also be possible to capture a M8 nut directly in your design, instead of using the original Hadron part... This should lead to a smaller stuff.

I would also use a little piece of metal rod for the index, and maybe close the long hole, to prevent it from opening...

Anyway, this will save me money: I was about to order ACME screw/Delrin nuts...
Frédéric
fma
 
Posts: 213
Joined: Mon Aug 20, 2012 7:53 am

Re: Eliminating Hadron z wobble

Postby brnrd » Wed Dec 19, 2012 10:13 pm

Yes. I thought about using a captive nut. Since I had the nuts that came with the kit, I just used them. I suppose I could make a version that has a tight cavity for an M8 nut so there's no backlash instead of the cutout for the Ord Bot z nut.

But this won't really change the size since the nut still has to be held by something.
brnrd
 
Posts: 111
Joined: Mon Dec 03, 2012 10:25 pm

Re: Eliminating Hadron z wobble

Postby cozmicray » Wed Dec 19, 2012 10:26 pm

I don't know how your thing works?
If it had a M8 thread in it (or a M8 nut) then the white plastic may provide a preload
and the thing act as an anit-backlash nut.
It seems your peg in a slot constrains in Y-axis and allows slop in X-axis?
Is that where the slop is in the maker-slide?

:?:
Ray



brnrd wrote:I just uploaded a thing to eliminate z wobble due to bent threaded rods. I'm still in the process of building a Hadron so I haven't tested it yet, but I expect it to work.

Image
cozmicray
 
Posts: 168
Joined: Thu Sep 20, 2012 4:32 pm
Location: Blue Bell, PA

Re: Eliminating Hadron z wobble

Postby brnrd » Wed Dec 19, 2012 11:08 pm

cozmicray wrote:I don't know how your thing works?
If it had a M8 thread in it (or a M8 nut) then the white plastic may provide a preload
and the thing act as an anit-backlash nut.
It seems your peg in a slot constrains in Y-axis and allows slop in X-axis?
Is that where the slop is in the maker-slide?

:?:
Ray


The white plastic parts don't come into contact with the threaded rod. So, it doesn't act as an anti-backlash nut. What it does is to allow for movement of the threaded rod and the plastic with the captive nut (black) and the attached bottom white part in both the x and y direction without moving the x carriage. Note that if the rod is bent too much, then it can result in a change in z height which would also result in uneven walls that mirror that M8 threads. In that case, a plastic washer between the two white parts may help.
brnrd
 
Posts: 111
Joined: Mon Dec 03, 2012 10:25 pm

Re: Eliminating Hadron z wobble

Postby brnrd » Wed Dec 19, 2012 11:10 pm

fma wrote:BTW, it could also be possible to capture a M8 nut directly in your design, instead of using the original Hadron part... This should lead to a smaller stuff.

I would also use a little piece of metal rod for the index, and maybe close the long hole, to prevent it from opening...

Anyway, this will save me money: I was about to order ACME screw/Delrin nuts...


I uploaded a version using captive M8 nuts. I haven't printed it yet so I hope the nut cavity of 13 mm for the is tight enough.
brnrd
 
Posts: 111
Joined: Mon Dec 03, 2012 10:25 pm

Re: Eliminating Hadron z wobble

Postby orcinus » Thu Dec 20, 2012 12:56 am

It's not the X-Y movement of the gantry that causes the wobble.
It's the energy wasted on deformation (and the resultant deviation in Z step height) that does.
orcinus
 
Posts: 720
Joined: Tue Feb 14, 2012 4:03 am

Re: Eliminating Hadron z wobble

Postby brnrd » Thu Dec 20, 2012 1:14 am

There are two possible artifacts that result from bent lead screws.

1. Side to side movement of the gantry which moves the extruder in the same direction. This results in x-y shifts in the layers with z. If you observe the walls of a print, it would bulge out on one side and bulge in on one side with the same spacing as the thread pitch.

2. Up and down movement of the gantry. This changes the layer thickness as you mentioned and causes the extruded width to vary with height at the period of the thread pitch.

The Ord Bot has a more rigid z mechanism than the Mendel designs (makerslide vs 8 mm smooth rods), but it only takes fractions of a mm to result in uneven walls. It seems to me that a bent or slightly off-center 8mm threaded rod connected to the motor shaft with a rigid coupler would have enough force to move the gantry especially when near the motor.
brnrd
 
Posts: 111
Joined: Mon Dec 03, 2012 10:25 pm

Re: Eliminating Hadron z wobble

Postby orcinus » Thu Dec 20, 2012 2:00 am

I seriously doubt that #1 is possible on an ORD Bot in the way described. That's why you see the motor plates bending on ORD Bots with bent rods.
There's nowhere for the gantry to go, the Z makerslide profiles are rock solid (provided you've squared them, tightened all the nuts and have the gantry V-wheels tight).

What can happen to the gantry, however, is left and right side of it moving along Z out of parallel, which would cause one side of an object to pop-out and the other side to pop-in. That effect is much much less severe and much rarer than whole layers bulging and getting inset due to varying Z step heights.

(all of this is based on observations on my machine, so take with a grain of salt)
orcinus
 
Posts: 720
Joined: Tue Feb 14, 2012 4:03 am

Re: Eliminating Hadron z wobble

Postby brnrd » Thu Dec 20, 2012 2:18 am

I'll find out when I start printing with this configuration. :)

It seems to me that using the rigid z couplers while allowing the lead screws some freedom to move in the x-y plane would also fix the up and down z movement better than using the flexible z motor couplers with the z nuts rigidly fixed on the gantry.
brnrd
 
Posts: 111
Joined: Mon Dec 03, 2012 10:25 pm

Re: Eliminating Hadron z wobble

Postby cvoinescu » Thu Dec 20, 2012 2:03 pm

I think orcinus is nearly spot-on.

A bent rod with a rigid coupler will will try to rotate the motor shaft out of vertical, pointing toward the outside of the bend. If the motor can move due to a flexible plate, it will do so. The plate is anchored on one side, so it acts as a hinge: as the motor bends with the shaft pointing toward the machine, it also lifts; when it bends away from the machine, it lowers.

Now, if both shafts are bent "in phase", both ends of the carriage will raise and lower at the same time, and you'll see effect #2 above.

If the shafts are out of phase, one end will lower while the other will raise, rotating the carriage out of horizontal. The nozzle is below the center of this rotation, so it will move sideways, and you'll see something very much like effect #1, but also with effect #2 on the left side of the bed in opposite phase from the right side (and fairly smooth in the center).

In practice, I expect the phase to be somewhere in between, so there would be both a sideways motion of the nozzle, and a variation of layer height. In any case, the layer height effect would dominate, whether it's uniform along X or goes from thick on the left to thin on the right and vice-versa.
cvoinescu
 
Posts: 501
Joined: Thu Aug 09, 2012 9:12 am
Location: Camberley, Surrey, UK

PreviousNext

Return to ORD Bot

Who is online

Users browsing this forum: No registered users and 29 guests