Centering collars (3-point-steady) for smooth rods alignment?
Posted by realthor
|
Centering collars (3-point-steady) for smooth rods alignment? March 31, 2016 02:42PM |
Hi I am exploring an idea of having the smooth rods ends mount designed with a larger hole than the diameter of the rod itself and then have a 3-point centering system similar to how lathes' do the centering of the part on the chuck.
I can see that I would be introducing complexity in the system and that alignment calibration would be hell albeit having the possibility to do it to perfection and, depending on how much larger the receiving hole is, can take a quite out-of-square frame and still get a square printing system.
What do you think.. is this worth any further thought? How would one go about squaring the whole thing? It's hard enough to do it with fixed ends so I am not that confident in the idea myself but before discarding it I wanted to ask around.
Edit: I've just learned that this is called a "3-point-steady" in lathe machining lingo.
Edited 8 time(s). Last edit at 03/31/2016 05:18PM by realthor.
RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff
I can see that I would be introducing complexity in the system and that alignment calibration would be hell albeit having the possibility to do it to perfection and, depending on how much larger the receiving hole is, can take a quite out-of-square frame and still get a square printing system.
What do you think.. is this worth any further thought? How would one go about squaring the whole thing? It's hard enough to do it with fixed ends so I am not that confident in the idea myself but before discarding it I wanted to ask around.
Edit: I've just learned that this is called a "3-point-steady" in lathe machining lingo.
Edited 8 time(s). Last edit at 03/31/2016 05:18PM by realthor.
RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 02, 2016 07:58AM |
Quote
the_digital_dentist
It is harder to align the Y axis rails when the XY mechanism isn't built on a flat plate. For the Y axis in your machine it might indeed be a good idea to use one of the three point adjusters on each rail.
Just as I thought. Although I can leave one fixed, as I said before your suggestion might me applicable for this case.
I'm trying to devise a method to use these spacial adjusters so that one ends up with a perfectly square Y even if the frame leaves some to be desired in terms of squareness. The X rail/rods can be used to square the Y as it slides on it you can notice where there is more friction or where is less and adjust the screws on the collars accordingly. But here's the gotcha: adjusting 3 screws at 120 degrees is very unintuitive because as you retract one you can affect Z also in the way I have designed them. So I am really thinking of changing the arrangement into a more manageable 0-90-180 adjusting system like the sketch below:
In this way, after you get Z right for 3 corners (one will be the reference), you can play left-right to get Y rods/rails parallel. The top is just the clamping system.
RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 02, 2016 10:49PM |
I am confused.
What is Orange piece for? ease in removing rod?
Quick rod replacement every day?
Is this, two guide rods that will define an axis?
By match drilling two plates -- the distance between guide rods would be set.
Mounting the plates parallel to each other would assure rods are not twisted,
and define / set plane.
A simple way to align plates
--- place laser pointer on one plate or across rods
and shoot it across a room , -- mark spot
move laser to other plate
align to mark on wall
A little mis-align would show up as vertical inches in laser spot difference
(depending how far wall is away from rods being aligned)
The mark on wall could be reference for all other alignments.
A level line all around your work room could be used to align all axes?
BWDIK?
What is Orange piece for? ease in removing rod?
Quick rod replacement every day?
Is this, two guide rods that will define an axis?
By match drilling two plates -- the distance between guide rods would be set.
Mounting the plates parallel to each other would assure rods are not twisted,
and define / set plane.
A simple way to align plates
--- place laser pointer on one plate or across rods
and shoot it across a room , -- mark spot
move laser to other plate
align to mark on wall
A little mis-align would show up as vertical inches in laser spot difference
(depending how far wall is away from rods being aligned)
The mark on wall could be reference for all other alignments.
A level line all around your work room could be used to align all axes?
BWDIK?
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 03, 2016 04:37AM |
Quote
the_digital_dentist
3 screws will move it anywhere in space that it needs to go. To me it seems like it's overconstrained if you use four screws. I think you could position the three screws to make the alignment easier. The other thing you could do is position the rail horizontally by mounting the rail holder on a slot in the frame for side to side motion, and then shim it for vertical. You should only have to set it up once and never touch it again, so while screws to turn are nice, they aren't completely necessary.
Yes I know d_d but it is very unintuitive. 4 screws seem a lot and unnecessary just like 4-screws bed plane adjusting but it's not the same. With 4 screws you first define Z by dialing in first the lower screw and then just fasten it by turning the upper vertical screw. That's it for Z, you now have it dialed in, all you have to be concerned about is left-right micro-adjustment so that the rods are square to each other, which is again very simple.
With 3 screws at 120 deg it is simply too difficult to adjust for a certain alignment. It's easy on the lathe where you only have to adjust one 3-point-steady but here you just have too many of them and it's hell when you turn one screw you have to un-turn another and you modify with this the position in all 3 axes... then you have to realign all others in consequence. It reminds me of calibrating Delta, where you dial in for one tower, then re-do all towers to match, then check again, etc
BTW: did you ever had a Christmas tree base that has this 3-point-steady system to make it vertical? It's incredibly difficult to dial in the right amount of turns per each screw to obtain the vertical.
Quote
cozmicray
I am confused.
What is Orange piece for? ease in removing rod?
Quick rod replacement every day?
Yes the initial intention was to first install the plastic holders then add the rods. It's not necessary at this point, when I can take the plastic holder off without disassembling the frame.
However, plastic gives and the amount of disassembling I've been doing/witnessing being done on Solidoodle just makes a way to easily take out the Y rods so you can replace parts on Y or on X a welcomed option.
Quote
cozmicray
A simple way to align plates
--- place laser pointer on one plate or across rods
and shoot it across a room , -- mark spot
move laser to other plate
align to mark on wall
A little mis-align would show up as vertical inches in laser spot difference
(depending how far wall is away from rods being aligned)
The mark on wall could be reference for all other alignments.
A level line all around your work room could be used to align all axes?
This is intriguing and I am interested in this method even if I don't have a laser level. Can you make a quick napkin sketch as I can imagine what you wanted to say but am not sure i got it right.
Edited 2 time(s). Last edit at 04/03/2016 04:47AM by realthor.
RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 03, 2016 04:13PM |
Keep it simple?
You really only need a $3 laser pointer with a nice straight round / square barrel
that gives nice small spot
Length of measurement points here exaggerates the mis-alignment
Example:
My mill is in middle of room
lay laser pointer on bed pointing to right wall mark on wall
swing laser around and point toward left wall mark on wall
Put laser on mill head which is 40cm above bed
mark on L R walls 40cm above other mark
adjust mill head so laser spot is on mark
check it on other wall
Mill head is aligned with some good precision to bed
Assume Mill shaft orthogonal to head (well spec sez it's 90+-0.0001 degree)
mill cutter should be orthogonal to bed ( good enough for government work)
============
(Water level) Hey piece of clear tubing filled with water can show you level
even if tubing is 100 feet long
A weight on a string will show you plumb
Example: Laser level on eBay (nice flat bottom) $5.50
[www.ebay.com]
You really only need a $3 laser pointer with a nice straight round / square barrel
that gives nice small spot
Length of measurement points here exaggerates the mis-alignment
Example:
My mill is in middle of room
lay laser pointer on bed pointing to right wall mark on wall
swing laser around and point toward left wall mark on wall
Put laser on mill head which is 40cm above bed
mark on L R walls 40cm above other mark
adjust mill head so laser spot is on mark
check it on other wall
Mill head is aligned with some good precision to bed
Assume Mill shaft orthogonal to head (well spec sez it's 90+-0.0001 degree)
mill cutter should be orthogonal to bed ( good enough for government work)
============
(Water level) Hey piece of clear tubing filled with water can show you level
even if tubing is 100 feet long
A weight on a string will show you plumb
Example: Laser level on eBay (nice flat bottom) $5.50
[www.ebay.com]
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 04, 2016 04:30PM |
Yes your right about laser being parallel to case.
If round --- roll it around and see if spot stays centered.
It's a relative thing and good procedure
results in good alignment.
Some good measurement and some Trigonometry
you may set angles, do alignment
even measure the diameter of the earth.
I have a inexpensive laser with flat magnetic bottom
and a lens can be a line rotated any way
or if lens is removed a nice bright spot.
Beware even low power lasers can do eye damage!
If round --- roll it around and see if spot stays centered.
It's a relative thing and good procedure
results in good alignment.
Some good measurement and some Trigonometry
you may set angles, do alignment
even measure the diameter of the earth.
I have a inexpensive laser with flat magnetic bottom
and a lens can be a line rotated any way
or if lens is removed a nice bright spot.
Beware even low power lasers can do eye damage!
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 07, 2016 10:57PM |
At first glance it seemed like a good way to test rail alignment, but after thinking about it, I realized it isn't.
Let's say the rails are parallel but one end of the pair is higher than the other. When you mark the spots on the wall they will not match. Then you have to decide if the rails are not parallel vertically or if the rails are parallel but not level. If the ends are at the same level and the rails are not parallel in the horizontal plane the marks on the wall will line up as if the rails are parallel. So the laser pointer trick doesn't distinguish between vertically nonparallel rails and vertically parallel but not level rails, and it can't tell you anything about whether the rails are parallel in the horizontal plane.
The technique requires further development to be useful.
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Let's say the rails are parallel but one end of the pair is higher than the other. When you mark the spots on the wall they will not match. Then you have to decide if the rails are not parallel vertically or if the rails are parallel but not level. If the ends are at the same level and the rails are not parallel in the horizontal plane the marks on the wall will line up as if the rails are parallel. So the laser pointer trick doesn't distinguish between vertically nonparallel rails and vertically parallel but not level rails, and it can't tell you anything about whether the rails are parallel in the horizontal plane.
The technique requires further development to be useful.
Quote
cozmicray
Keep it simple?
You really only need a $3 laser pointer with a nice straight round / square barrel
that gives nice small spot
[attachment 75869 Laser_align.jpg]
Length of measurement points here exaggerates the mis-alignment
Example:
My mill is in middle of room
lay laser pointer on bed pointing to right wall mark on wall
swing laser around and point toward left wall mark on wall
Put laser on mill head which is 40cm above bed
mark on L R walls 40cm above other mark
adjust mill head so laser spot is on mark
check it on other wall
Mill head is aligned with some good precision to bed
Assume Mill shaft orthogonal to head (well spec sez it's 90+-0.0001 degree)
mill cutter should be orthogonal to bed ( good enough for government work)
============
(Water level) Hey piece of clear tubing filled with water can show you level
even if tubing is 100 feet long
A weight on a string will show you plumb
Example: Laser level on eBay (nice flat bottom) $5.50
[www.ebay.com]
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
|
Re: Centering collars (3-point-steady) for smooth rods alignment? April 08, 2016 05:19PM |
Level --- meaningless?
?? a horizontal position or condition
?? an imaginary line or surface everywhere at right angles to the plumb line.
?? the horizontal line or plane in which anything is situated, with regard to its elevation.
Parallel lines or parallel planes better
X axis --- to --- build plate parallel
By your own measurements with digital dial gauge
The rails are parallel because the mounting plates are matched drilled
The laser is facilitating making the two separated mounting plates parallel
so no twist in rail system
Then make the rail system parallel to reference ( build platform )
or whatever is the chosen "reference" plane
Shooting both sides of the surface of your granite build block
could provide a level line, or parallel line on the wall?
or
by procedure you align your printer reference to the line on the wall
every time you do any alignments
The wall somewhat far away from unit being aligned
provides a lot of displacement for angular mis-alignment
Maybe that wall orthogonal can be put into play
or
breakout the optical bench and 16 precision mirrors / prisms
and the smoke and glasses to see lines
hummmm
new house plans, needs
--- granite block in basement floor, precision aligned walls, alignment cubes
?? a horizontal position or condition
?? an imaginary line or surface everywhere at right angles to the plumb line.
?? the horizontal line or plane in which anything is situated, with regard to its elevation.
Parallel lines or parallel planes better
X axis --- to --- build plate parallel
By your own measurements with digital dial gauge
The rails are parallel because the mounting plates are matched drilled
The laser is facilitating making the two separated mounting plates parallel
so no twist in rail system
Then make the rail system parallel to reference ( build platform )
or whatever is the chosen "reference" plane
Shooting both sides of the surface of your granite build block
could provide a level line, or parallel line on the wall?
or
by procedure you align your printer reference to the line on the wall
every time you do any alignments
The wall somewhat far away from unit being aligned
provides a lot of displacement for angular mis-alignment
Maybe that wall orthogonal can be put into play
or
breakout the optical bench and 16 precision mirrors / prisms
and the smoke and glasses to see lines
hummmm
new house plans, needs
--- granite block in basement floor, precision aligned walls, alignment cubes
Quote
the_digital_dentist
At first glance it seemed like a good way to test rail alignment, but after thinking about it, I realized it isn't.
Let's say the rails are parallel but one end of the pair is higher than the other. When you mark the spots on the wall they will not match. Then you have to decide if the rails are not parallel vertically or if the rails are parallel but not level. If the ends are at the same level and the rails are not parallel in the horizontal plane the marks on the wall will line up as if the rails are parallel. So the laser pointer trick doesn't distinguish between vertically nonparallel rails and vertically parallel but not level rails, and it can't tell you anything about whether the rails are parallel in the horizontal plane.
The technique requires further development to be useful.
Sorry, only registered users may post in this forum.