Using StackShot to drive a linear stage
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Jotafoto has been modifying some micrometric rails to be used with stackshot; last week I recieved my copy.
The Rail is a 25mm travel melles griot and the actuator is a 1/4-80 Newport actuator with 1" travel
So there is 0,3175mm travel per revolution; so with the 200 step/rev motor gives me 0,00158 full steps or up to 0,0001mm substeps in stackshot's high precision mode (16 substeps per step, Thanks Rik!!)
The motor is placed in a 90º aluminium plate and this one is hold on a flat aluminium polate that also holds the rail.
The motor is linked to the newport actuator via a double universal joint
Regards
Javier
The Rail is a 25mm travel melles griot and the actuator is a 1/4-80 Newport actuator with 1" travel
So there is 0,3175mm travel per revolution; so with the 200 step/rev motor gives me 0,00158 full steps or up to 0,0001mm substeps in stackshot's high precision mode (16 substeps per step, Thanks Rik!!)
The motor is placed in a 90º aluminium plate and this one is hold on a flat aluminium polate that also holds the rail.
The motor is linked to the newport actuator via a double universal joint
Regards
Javier
Last edited by seta666 on Tue Jun 05, 2012 5:16 am, edited 1 time in total.
- rjlittlefield
- Site Admin
- Posts: 23561
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
To put things in perspective, this is exactly 5 times finer than the standard StackShot, which has a 1/4-16 thread. It is somewhat less fine than a typical Olympus focus block, which has 0.200mm travel per rotation. Some common Nikon blocks are twice that fine, at 0.100mm travel per rotation.seta666 wrote:the actuator is a 1/4-80 Newport actuator with 1" travel. So there is 0,3175mm travel per revolution
A minor correction here: the StackShot controller uses 16 microsteps per full step, not 20. Thus there are 3200 microsteps per rotation, giving about 1/10 micron per microstep on average. If you measure carefully, the largest microsteps will probably turn out to be about twice that big, due to nonlinearities in the motor drive currents.or up to 0,00008mm substeps in stackshot's high precision mode (20 substeps per step)
The StackShot is usable but marginal at 40X NA 0.50. In testing, my unit showed a largest single microstep around 1.4 microns (see HERE). With at least 5 times that accuracy, Javier's rail should be robust and controllable with all objectives.pierre wrote:Is it not a bit too much of precision ?
I thought stackshot was ok for lense mag 100x ?
--Rik
- Craig Gerard
- Posts: 2877
- Joined: Sat May 01, 2010 1:51 am
- Location: Australia
Also note:
In Javier's example a Newport High Precision Adjustment Screw (very fine) has been used in place of a Vernier Micrometer.
http://www.newport.com/AJS-Series-High- ... ifications
Craig
In Javier's example a Newport High Precision Adjustment Screw (very fine) has been used in place of a Vernier Micrometer.
http://www.newport.com/AJS-Series-High- ... ifications
Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"
Well, I have used the stackshot up to 100X, but anything above 20x is to much for it really. Any linear stage (newport, melles griot, etc..) is more robust and has less wooble than the stackshot. This is easy to tell by looking at the dust trails the stack leaves. With a linear stage it looks like a straightpierre wrote:Hello Javier,
Is it not a bit too much of precision ?
I thought stackshot was ok for lense mag 100x ?
HAve you a special need?
line (or almost) With stackshot it looks more like a coil or snake
Also now with this rail a will be able to work at high magnification 20-40X without needing substeps and with substeps I can do 60-100X no problem. 25mm travel is more than enough for my needs and I still have the stackshot rail for low magnification work
Thank you Rik, I wrote it by memory without looking at the specification of stackshot; the way I use it is by telling the stackshot in custom configuration that that the rail gives 317.5mm per revolution so when the stackshot shows 99um(minimum value) it really means 0.000099mm and 1mm on stackshot is 1um in reality, 10mm is 10um and so onrjlittlefield wrote: A minor correction here: the StackShot controller uses 16 microsteps per full step, not 20. Thus there are 3200 microsteps per rotation, giving about 1/10 micron per microstep on average. If you measure carefully, the largest microsteps will probably turn out to be about twice that big, due to nonlinearities in the motor drive currents.
EDIT At the end I left real values in custom configuration; 0,317mm per revolution which allows me a minimum step of 1um; the problem is that there is a bug in stackshot controller.
When in auto-Dist mode I try to put 10mm/step value it jumps to maximum value and when going from maximum value to minimum it will also jumps when reaching 10mm mark
You can see this bug here
http://www.youtube.com/watch?v=AEuvZwBd ... detailpage
Again, this is not only because of the finer pitch of the actuator (5 times more precisse) but also because of amore solid rail with less wooble; Zerene stacker will have to make less corrections between shotsrjlittlefield wrote: The StackShot is usable but marginal at 40X NA 0.50. In testing, my unit showed a largest single microstep around 1.4 microns (see HERE). With at least 5 times that accuracy, Javier's rail should be robust and controllable with all objectives.
--Rik
Yes, those are more precisse than the one I have but I fell the one I have is more than enough for my needs ;-)Craig Gerard wrote:In Javier's example a Newport High Precision Adjustment Screw (very fine) has been used in place of a Vernier Micrometer.
http://www.newport.com/AJS-Series-High- ... ifications
Here is one more picture of a provisional setup with a 6 AXIS micrometric set up for framing; under the wooden base there are 4 sorbothane semispheres to absorb vibrations
Regards
Javier
- Craig Gerard
- Posts: 2877
- Joined: Sat May 01, 2010 1:51 am
- Location: Australia
Unfortunally yesterday I sent my gear to Portugal, incluiding the stackshot as I have some work to do there for 3 weeks starting next Monday. I won't be able to do it until I cam back as this linear stage stays home as I have not tested it yet.Craig Gerard wrote:Javier,
When you get some time could you take some video footage of the motorised translation in operation? Nothing elaborate, just some forward and backward motion.
Craig
However I did some quick tests by moving it forwards and backwards and the movement was very smooth. This kind of high precission actuators may be slow when moving by hand but as you can change the sped of the motor this is not a problem when motorized
I have asked Jotafoto to make a video of his own motorized rail, he will do it
Regards
Javier
-
- Posts: 674
- Joined: Sat May 01, 2010 6:12 am
- Location: Nice, France (I'm British)
Thanks Craig, I was puzzled by that part and thought it was a stripped-down micrometer or something.Craig Gerard wrote: In Javier's example a Newport High Precision Adjustment Screw (very fine) has been used in place of a Vernier Micrometer.
http://www.newport.com/AJS-Series-High- ... ifications
- Craig Gerard
- Posts: 2877
- Joined: Sat May 01, 2010 1:51 am
- Location: Australia
seta666 and jotafoto,
Thanks. Looking forward to the demonstration
Chris,
Those Newport fine adjustment screws are very...'fine'. I'm playing around with a couple of similar items at the moment. I'm using a translation stage with 2" of travel. Javier's idea to use the double Universal Joint should provide adequate coupling to the stepper motor....this is an interesting project.
Here is an example of the New Focus (Newport) screws (1/4-80).
50.8mm travel version:
http://search.newport.com/?x2=sku&q2=9303-K
Craig
*edit: amended hyperlink. Included an additional hyperlink. Corrected content.
Thanks. Looking forward to the demonstration
Chris,
Those Newport fine adjustment screws are very...'fine'. I'm playing around with a couple of similar items at the moment. I'm using a translation stage with 2" of travel. Javier's idea to use the double Universal Joint should provide adequate coupling to the stepper motor....this is an interesting project.
Here is an example of the New Focus (Newport) screws (1/4-80).
50.8mm travel version:
http://search.newport.com/?x2=sku&q2=9303-K
Craig
*edit: amended hyperlink. Included an additional hyperlink. Corrected content.
Last edited by Craig Gerard on Thu Jun 07, 2012 5:26 pm, edited 3 times in total.
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"
d'oh. Why didn't I think about this, it's so obvious.seta666 wrote:This is easy to tell by looking at the dust trails the stack leaves. With a linear stage it looks like a straight
line (or almost) With stackshot it looks more like a coil or snake
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.
Jotafoto is lazy (and busy) man, I do not know when he will do itCraig Gerard wrote:seta666 and jotafoto,
Thanks. Looking forward to the demonstration
Newport do not appear to have a fine screw (without knob) for 2" travel.
Here you have one 2" actuator
http://search.newport.com/?q=*&x2=sku&q2=AJS100-2
The one I am using is the old 1" version (black knob), the knob has just been removed
Javier's idea to use the double Universal Joint should provide adequate coupling to the stepper motor....this is an interesting project.
At the begining he was not using the double universal joint ant there were some oscillation/vibration problems as it is almost imposible to have motor and actuator aligned perfectly as some lather work is involved; the double universal joint can cope with this sligth missaligment but this is something Jotafoto can explain better than I
Regards
Javier
Hi, long time without being here. Let's see if google translator lets us communicate.
The double universal joint is effective, if the axes of motor and actuator screw are perfectly aligned or at least parallel.
If they are perfectly parallel, double universal joint transmits engine power, without jumps; if the axes are not parallel (the double universal joint allows the transmission angle to 90 degrees) then the position of the engine has to be perfect to avoid jumps and swings.
In my project, if it put the motor perfectly at 90 ° in relation to the axis of movement of the rail, the position of the motor in the housing is not as critical because the double universal joint is responsible for transmitting smooth movement. The problem is if you have not perfectly matched those planes that I have named, at 90 º.
Tomorrow if I have time I will upload a video to youtube explaining it (in Spanish, sorry), but I think will be enough to understand this.
Another method would be with a flexible joint but I think it would be worse; anyway, I'll try it too.
I equipped my set up with a newport 1/4-100 actuator and the accuracy I have with the 200 steps, is 0.00005 inch, or 0.00127 mm per step. In the custom config of the old version stackshot must be set to 254 mm distance per revolution and divided by 1000 , the updated version set to 2.54 mm and divided by 10. With Seta666's actuator, in the old version 1/4-80 317.5 mm and 3.175 mm the new version In any case, the backlash is set to 0.
regards
Javier Replinger
The double universal joint is effective, if the axes of motor and actuator screw are perfectly aligned or at least parallel.
If they are perfectly parallel, double universal joint transmits engine power, without jumps; if the axes are not parallel (the double universal joint allows the transmission angle to 90 degrees) then the position of the engine has to be perfect to avoid jumps and swings.
In my project, if it put the motor perfectly at 90 ° in relation to the axis of movement of the rail, the position of the motor in the housing is not as critical because the double universal joint is responsible for transmitting smooth movement. The problem is if you have not perfectly matched those planes that I have named, at 90 º.
Tomorrow if I have time I will upload a video to youtube explaining it (in Spanish, sorry), but I think will be enough to understand this.
Another method would be with a flexible joint but I think it would be worse; anyway, I'll try it too.
I equipped my set up with a newport 1/4-100 actuator and the accuracy I have with the 200 steps, is 0.00005 inch, or 0.00127 mm per step. In the custom config of the old version stackshot must be set to 254 mm distance per revolution and divided by 1000 , the updated version set to 2.54 mm and divided by 10. With Seta666's actuator, in the old version 1/4-80 317.5 mm and 3.175 mm the new version In any case, the backlash is set to 0.
regards
Javier Replinger
Last edited by jotafoto on Fri Jun 08, 2012 12:37 pm, edited 1 time in total.
- Craig Gerard
- Posts: 2877
- Joined: Sat May 01, 2010 1:51 am
- Location: Australia
Javier,jotafoto wrote:Hi, long time without being here. Let's see if google translator lets us communicate.
The double universal joint, is effective, if the axes of motor and actuator screw are, or perfectly aligned, or at least parallel. If they are perfectly parallel, double universal joint transmit engine power, without jumping. And if the axes are not parallel (the double universal joint allows the transmission angle to 90 degrees) then the position of the engine has to be perfect to avoid jumps and swings. In my project, if it put perfectly at 90 ° plane holding the motor in relation to the axis of movement of the rail, and the position of the motor in the housing is not as critical because the double universal joint is responsible for transmitting smooth movement. The problem is if you are not perfectly matched those planes that I have named, at 90 º. Tomorrow if I have time I will go to you tube a video explaining (in Spanish, sorry), but I think will be enough to understand the video. Another method would be with a flexible joint. But I think it would be worse. Although I'll prove it too. I equipped my team with a screw newport 1/4-100 and accuracy I have with the 200 steps, is 0.00005 inch, or 0.00127 mm. In the custom config. the stackshot must be put with the old version, 254 mm distance per revolution and divide by 1000 and the updated version 2.54 mm and divide by 10. Seta666 Lane, in the old version 1/4-80 317.5 mm and 3.175 mm the new version In any case, the back lash is set to 0.
regards
Javier Replinger
Understood, language translation okay. Thanks! Looking forward to the video
Craig
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"
This type of connector also works well for small mis-alignments: http://heli-cal.com/cm/Products/Flexibl ... /Home.html
And another: http://onlinecatalog.tpa-us.com/categor ... -couplings
And another: http://onlinecatalog.tpa-us.com/categor ... -couplings
- Craig Gerard
- Posts: 2877
- Joined: Sat May 01, 2010 1:51 am
- Location: Australia