Drylin Linear Guides
Posted by uGen
|
Re: Drylin Linear Guides February 06, 2013 03:02PM |
Hi uGen,
with "common PZT's" I'm adressing the single-actuator-piezo's, that only bend or expand in one axis - here a sample:
The green block embedded in the tweezer body is a monoblock-PZT that is driven with the black box, that converts 5VDC to 300VDC with analog modulating/moving one of the tweezer-tips over (only!) 0.3mm at the tip with up to 300kHz frequenzy.
***
The second type were 'propagating wave motors' with two combined monoblock-piezos from EDO, that were then (around 2001) something around 60 USD per motor and maybe 80 USD per driver:
Here the two brown piezo-blocks can perform a phase-shifted combined oszillation, so when pressed against a smooth and hard surface, the white tip will start to 'hop' and pushing the surface in one direction (or the other, depending on phase).
The black box with two of the motors facing each other and a aluminiumoxide slider between them is a demonstrator, the red part is a 'single-drive' for endless linear or circular (with a ring) moving of a slider set between the motor tip and the ball bearing.
The big metal block with the three motors attached is a 'car' lying upside down - when set on a glass-plate, the three 'motors' can move the body linear (the two parallel) and turn it (the single at one end or in combination with the other two) ... so it can drive freely in a vacuum ambient without outgassing or parts sensible to vacuum or strong magnetic fields (like common motors).
***
And here are some of the PiezoLegs and some of my prototypes from developing linear stages (the black one on the right):
Here the active element is a comb-shaped part with 4 legs, consiting of some ten stacked PZT-sheets, two stacks per leg.
With clever activating this sheets every leg can shorten or elongate and bend forward or backward (in the direction of travel).
The PIC-controllers in the driver board are programmed for an activating sequence, that perform a sort of 'galopping' with the four legs, what moves an attached slider. The speed of this movement can be varied - slow traveling is made with shorter steps (maybe 100 nanometers), fast moving rises this length up to 300 nm per step. Between this steps a pure 'bending' of all legs synchrone will move the slider with nanometer or sub-nanometer resolution ...
They sold then (2002/2003) single demonstrator kits (one motor, the driver in the black box with two buttons and a 12V-PS) for 395 Euros ... we got them as OEM something cheaper with special programmed PIC's, but don't know the actual price. I hope the target price per motor dropped significantly since then
Edited 2 time(s). Last edit at 02/06/2013 03:17PM by VDX.
Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]
Call for the project "garbage-free seas" - [reprap.org]
with "common PZT's" I'm adressing the single-actuator-piezo's, that only bend or expand in one axis - here a sample:
The green block embedded in the tweezer body is a monoblock-PZT that is driven with the black box, that converts 5VDC to 300VDC with analog modulating/moving one of the tweezer-tips over (only!) 0.3mm at the tip with up to 300kHz frequenzy.
***
The second type were 'propagating wave motors' with two combined monoblock-piezos from EDO, that were then (around 2001) something around 60 USD per motor and maybe 80 USD per driver:
Here the two brown piezo-blocks can perform a phase-shifted combined oszillation, so when pressed against a smooth and hard surface, the white tip will start to 'hop' and pushing the surface in one direction (or the other, depending on phase).
The black box with two of the motors facing each other and a aluminiumoxide slider between them is a demonstrator, the red part is a 'single-drive' for endless linear or circular (with a ring) moving of a slider set between the motor tip and the ball bearing.
The big metal block with the three motors attached is a 'car' lying upside down - when set on a glass-plate, the three 'motors' can move the body linear (the two parallel) and turn it (the single at one end or in combination with the other two) ... so it can drive freely in a vacuum ambient without outgassing or parts sensible to vacuum or strong magnetic fields (like common motors).
***
And here are some of the PiezoLegs and some of my prototypes from developing linear stages (the black one on the right):
Here the active element is a comb-shaped part with 4 legs, consiting of some ten stacked PZT-sheets, two stacks per leg.
With clever activating this sheets every leg can shorten or elongate and bend forward or backward (in the direction of travel).
The PIC-controllers in the driver board are programmed for an activating sequence, that perform a sort of 'galopping' with the four legs, what moves an attached slider. The speed of this movement can be varied - slow traveling is made with shorter steps (maybe 100 nanometers), fast moving rises this length up to 300 nm per step. Between this steps a pure 'bending' of all legs synchrone will move the slider with nanometer or sub-nanometer resolution ...
They sold then (2002/2003) single demonstrator kits (one motor, the driver in the black box with two buttons and a 12V-PS) for 395 Euros ... we got them as OEM something cheaper with special programmed PIC's, but don't know the actual price. I hope the target price per motor dropped significantly since then
Edited 2 time(s). Last edit at 02/06/2013 03:17PM by VDX.
Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]
Call for the project "garbage-free seas" - [reprap.org]
|
Re: Drylin Linear Guides February 12, 2013 10:03AM |
Pacca Wrote:
-------------------------------------------------------
> Nanometric precision is really an interesting
> argument! But to use this evolved linear movement
> technology, how to remove the quality-bottle-neck
> of the extruder?
... I've designed my Tripod not for printing, but as demonstrator for magnet-ball joints and microassembly, paste-dispensing and micro-laserwelding - here a resolution of some microns or submicrons is essential for assembling/fabbing bulk microsensors.
The mechanics/joints can be scaled from macro- (some meters) to micro-dimensions (some mm's) ... I've operated it with linear piezo-motors with 22mm travelling range and micro-grippers for fine-placing/assembling 0.2mm thick rod-lenses on fiberscopes
Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]
Call for the project "garbage-free seas" - [reprap.org]
-------------------------------------------------------
> Nanometric precision is really an interesting
> argument! But to use this evolved linear movement
> technology, how to remove the quality-bottle-neck
> of the extruder?
... I've designed my Tripod not for printing, but as demonstrator for magnet-ball joints and microassembly, paste-dispensing and micro-laserwelding - here a resolution of some microns or submicrons is essential for assembling/fabbing bulk microsensors.
The mechanics/joints can be scaled from macro- (some meters) to micro-dimensions (some mm's) ... I've operated it with linear piezo-motors with 22mm travelling range and micro-grippers for fine-placing/assembling 0.2mm thick rod-lenses on fiberscopes
Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]
Call for the project "garbage-free seas" - [reprap.org]
|
Re: Drylin Linear Guides February 14, 2013 07:48PM |
So, yesterday I received the second sample from Igus. This time, I ordered a complete 80mm wide linear rail with 4 WJUME-01-10 adjustable bearings mounted as a carriage.
The main concern I had was play in the bearings, so in order to properly test that out, I had to have a system similar to what would be used on a 3D printer (or any Cartesian robot).
Here goes:
Appearance:
As you can see below, the whole thing looks massive. The rail itself and the carriage top plate are made of aluminium (I wrote about the bearings before, so I won't go into detail here).
What struck me as odd first, but proved to be no problem at all is that the carriage is attached to each bearing by only one screw (it's tight enough not to shift at all). You use the other threaded hole to attach whatever you want to mount onto the plate.
Findings:
This carriage glides really well! It's zippy, smooth and silent. Nice for a high-quality printer. But alas, I could tilt the carriage into yawing motion rather easily. The whole carriage can be lifted some tenths of a mm in Z direction while it remains play free perpendicular to the line of motion. This might be good enough for the Y axis, though, because the forces trying to lift / tilt the carriage are small compared to what I exerted with my hand.
That was fresh out of the box, though. As soon as I tightened the adjustment screws, you could easily see play getting less and less, until it was quite acceptable (better than our 8mm rods affixed to plastic parts, anyway). I needed quite the force to lift the carriage, so it should all be well, right?
Unfortunately, here comes the trade-off of reduced play: At this stage, the bearings were so tight that the carriage was rather difficult to move. Not as in "it's stuck, let me get a hammer" difficult, but tight enough that I am having doubts whether our NEMA17 motors can reliably pull the carriage around quickly for longer periods of time without overheating the drivers. Given that the carriage won't slide down by it's own when subjected to gravitational pull, this might be a nice design for a belt-driven Z for example.
The problem about the bearings is that the adjustment screw only presses against the glide pad from the side. So you can easily eliminate play in the sideways direction. As you can see below, the screw was got a flat head that sits against the pad. The screws' axis is actually below the center of the 10mm supported rail, but because the screw's interface head is rigid, only a small corner is pressing against the plastic, and that a little bit up, so you have nothing that holds the underside of the bearing pad in place. You can see the gap between the screw and the glide pad that causes play in Z direction quite clearly.
There are actually ways to improve this design: either make the screw's interface head flexible (by means of a ball joint for example), so that it can tilt a little bit to also clamp down onto the pad from below or change the position and angle of the screw so that it points diagonally upward towards the center of the rail. By tightening the screw, pressure would be exerted from below and the side, eliminating play in both directions.
The first solution is expensive and the second rather impractical as it would be difficult to reach the screw in that position...
Oh, in case you wonder about these cracks on the rail: I don't exactly know what caused them. but you can't feel them by running your finger over the rail. However, they become ever so slightly noticeable by scratching over the surface with your fingernails. The previous example also had these defects. Maybe it has to do with the anodized layer?
Well, I am tired now, will continue the discussion about piezo motors tomorrow...
Edited 2 time(s). Last edit at 02/14/2013 07:52PM by uGen.
The main concern I had was play in the bearings, so in order to properly test that out, I had to have a system similar to what would be used on a 3D printer (or any Cartesian robot).
Here goes:
Appearance:
As you can see below, the whole thing looks massive. The rail itself and the carriage top plate are made of aluminium (I wrote about the bearings before, so I won't go into detail here).
What struck me as odd first, but proved to be no problem at all is that the carriage is attached to each bearing by only one screw (it's tight enough not to shift at all). You use the other threaded hole to attach whatever you want to mount onto the plate.
Findings:
This carriage glides really well! It's zippy, smooth and silent. Nice for a high-quality printer. But alas, I could tilt the carriage into yawing motion rather easily. The whole carriage can be lifted some tenths of a mm in Z direction while it remains play free perpendicular to the line of motion. This might be good enough for the Y axis, though, because the forces trying to lift / tilt the carriage are small compared to what I exerted with my hand.
That was fresh out of the box, though. As soon as I tightened the adjustment screws, you could easily see play getting less and less, until it was quite acceptable (better than our 8mm rods affixed to plastic parts, anyway). I needed quite the force to lift the carriage, so it should all be well, right?
Unfortunately, here comes the trade-off of reduced play: At this stage, the bearings were so tight that the carriage was rather difficult to move. Not as in "it's stuck, let me get a hammer" difficult, but tight enough that I am having doubts whether our NEMA17 motors can reliably pull the carriage around quickly for longer periods of time without overheating the drivers. Given that the carriage won't slide down by it's own when subjected to gravitational pull, this might be a nice design for a belt-driven Z for example.
The problem about the bearings is that the adjustment screw only presses against the glide pad from the side. So you can easily eliminate play in the sideways direction. As you can see below, the screw was got a flat head that sits against the pad. The screws' axis is actually below the center of the 10mm supported rail, but because the screw's interface head is rigid, only a small corner is pressing against the plastic, and that a little bit up, so you have nothing that holds the underside of the bearing pad in place. You can see the gap between the screw and the glide pad that causes play in Z direction quite clearly.
There are actually ways to improve this design: either make the screw's interface head flexible (by means of a ball joint for example), so that it can tilt a little bit to also clamp down onto the pad from below or change the position and angle of the screw so that it points diagonally upward towards the center of the rail. By tightening the screw, pressure would be exerted from below and the side, eliminating play in both directions.
The first solution is expensive and the second rather impractical as it would be difficult to reach the screw in that position...
Oh, in case you wonder about these cracks on the rail: I don't exactly know what caused them. but you can't feel them by running your finger over the rail. However, they become ever so slightly noticeable by scratching over the surface with your fingernails. The previous example also had these defects. Maybe it has to do with the anodized layer?
Well, I am tired now, will continue the discussion about piezo motors tomorrow...
Edited 2 time(s). Last edit at 02/14/2013 07:52PM by uGen.
|
Re: Drylin Linear Guides February 16, 2013 02:13PM |
Can you read in my mind?Quote
uGen
Oh, in case you wonder about these cracks on the rail
So, if I understand correctly, this linear guide is not suitable for a printer / light mill with nema17 (because of the vibration caused by milling, which can rotare the carriage around the yaw angle)...
|
Re: Drylin Linear Guides February 16, 2013 05:27PM |
I would have doubts about driving this thing with timing belts, yes. If you were to use leadscrews like on the Aluminatus (by Trinitylabs), the higher friction shouldn't be problematic.
To use this specific rail for milling wouldn't be wise. As you have stated, the forces applied by the mill would rotate the carriage. However, the bigger models (16mm and up) allow for the bearing blocks to be axially rotated 90° on the rail, so that you can also adjust preload and play from above. This theoretically should eliminate all play. The disadvantage is that you would need 6 bearing blocks, further increasing friction (but this isn't really a problem if you equip the mill with leadscrews or ballscrews) and complexity because Igus designed the bearing blocks in a way that allows for easy setup in either configuration, but not combined if used with the double rail.
You can use the single rails like this, however:
One rail would allow for vertical adjustment and the other for horizontal.
Here is a drawing to illustrate the two possible bearing block configurations:
To use this specific rail for milling wouldn't be wise. As you have stated, the forces applied by the mill would rotate the carriage. However, the bigger models (16mm and up) allow for the bearing blocks to be axially rotated 90° on the rail, so that you can also adjust preload and play from above. This theoretically should eliminate all play. The disadvantage is that you would need 6 bearing blocks, further increasing friction (but this isn't really a problem if you equip the mill with leadscrews or ballscrews) and complexity because Igus designed the bearing blocks in a way that allows for easy setup in either configuration, but not combined if used with the double rail.
You can use the single rails like this, however:
One rail would allow for vertical adjustment and the other for horizontal.
Here is a drawing to illustrate the two possible bearing block configurations:
|
Re: Drylin Linear Guides February 26, 2013 01:53PM |
Yesterday I recieved my samples from Igus!
I was oriented towards a compact system, so I ordered a T miniature guide rail with a TW-04-15 carriage and a N27 compact guide rail with three differents carriage, i.e. a long NW-11-27-80, a preloaded NW-01-27P and a longer preloaded NW-21-27-60P.
T system
This is not the adjustable version (not in stock when I ordered this sample) but however I wanted to try this solution.
The only problem of this carriage is a bit of play around the yaw and the roll axis (Yaw, pitch and roll angles with the x axis along the rail) but I think that this shouldn't be noticeable with a dual rail assembly.
@uGen: How is the play with the adjustable TWE carriage, have you tried it?
N system
The upper right carriage is the NW-11-27-80. It is twice the length of the standard one and it has the same play, unfortunately too much for a precise move I think (more than the TW carriage).
The lower carriage is the NW-01-27P preloaded version. It has a small spring element on the left that reduces very much the play in the movement (but I don't know how this evolves with wear).
The carriage in the middle is the NW-21-27-60P preloaded longer version (50% longer than the standard one). It reduces even more the play (virtually zero), but the downside is an increase of the force necessary for the sliding.
I was oriented towards a compact system, so I ordered a T miniature guide rail with a TW-04-15 carriage and a N27 compact guide rail with three differents carriage, i.e. a long NW-11-27-80, a preloaded NW-01-27P and a longer preloaded NW-21-27-60P.
T system
This is not the adjustable version (not in stock when I ordered this sample) but however I wanted to try this solution.
The only problem of this carriage is a bit of play around the yaw and the roll axis (Yaw, pitch and roll angles with the x axis along the rail) but I think that this shouldn't be noticeable with a dual rail assembly.
@uGen: How is the play with the adjustable TWE carriage, have you tried it?
N system
The upper right carriage is the NW-11-27-80. It is twice the length of the standard one and it has the same play, unfortunately too much for a precise move I think (more than the TW carriage).
The lower carriage is the NW-01-27P preloaded version. It has a small spring element on the left that reduces very much the play in the movement (but I don't know how this evolves with wear).
The carriage in the middle is the NW-21-27-60P preloaded longer version (50% longer than the standard one). It reduces even more the play (virtually zero), but the downside is an increase of the force necessary for the sliding.
|
Re: Drylin Linear Guides March 24, 2013 04:01PM |
Hi uGen!
Any news in your research for a "perfect" linear sliding system?
I'm doing some research in my spare time...and I think that I will try (as soon as I can) two options
Standard skate bearings assembly, like this or this
V groove bearings sliding on angles, like this (standard tracks for this type of bearing are expensive...and I cannot understand if could be possible to slide on the angle of standard 20x20 t-slot extrusion)
What do you think about these solutions?
Edited 1 time(s). Last edit at 03/24/2013 04:14PM by Pacca.
Any news in your research for a "perfect" linear sliding system?
I'm doing some research in my spare time...and I think that I will try (as soon as I can) two options
Standard skate bearings assembly, like this or this
V groove bearings sliding on angles, like this (standard tracks for this type of bearing are expensive...and I cannot understand if could be possible to slide on the angle of standard 20x20 t-slot extrusion)
What do you think about these solutions?
Edited 1 time(s). Last edit at 03/24/2013 04:14PM by Pacca.
Sorry, only registered users may post in this forum.