Open Focus Stacking Rail (OpenBuilds & RatRig)

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

mawyatt wrote:
mjkzz wrote:JohnyM, I do not know if my controllers will work for you. Mike, good luck with your adventure.

This C-Beam thing is making focus stacking equipment a commodity product, it is stable, backlash free, wobble free. People can start doing PRECISION, high magnification focus stacking for under 350USD even with a LIFE time license of Zerene. This is something to be reckoned with.

Now I just found out that there are arduino base solutions for focus stacking, already, though not sure if they are suitable for this rail, it would not be hard to make it happen, so I might not have to make one :D
Peter,

It's been a fun but trying adventure developing this Precision Stack and Stitch system so far :D

Working quite well now, and the orchestrated movement of the motors and rails, with all the flashing LEDS (3 USB, 3 motor activation, 3 command, camera trigger, & strobe trigger) and strobes is fascinating to watch :)

I've tried to mess up the system by changing motor direction commands while the motors are moving during initial parameter setup defining start, end and step, and the system keeps precise track of all the motor positions in spite of my attempts to mess things up!! I've even pulled the motor power plug during sessions in the middle of all the motors moving, or course everything stopped!! When I restarted the session the motors returned to the staring positions with only a small error!! This is because the motor supply voltage is monitored on all 3 axises, and when it falls below a limit the step commands are halted, thus not accumulating a large position error. I'm still working on some of the error routines and may expand these even more.

As my programming skills improve I'm able to accomplish more which is encouraging to me at least.

Best,
Good luck, as I told you in PM long before, I think they are good as replacement for computer, I am not sure if it is suitable for direct hardware control. Anyways, this is not the thread to discuss that, so I will stop here.

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

elf wrote:You can find the TR8 leadscrews on Aliexpress in 1,2,4, and 8mm pitches. I have a 500mm long TR8-1mm for another project. The quality appears to be excellent. You will need to be able to machine the ends in order to make it fit the rail.
I did one for a customer.

JohnyM
Posts: 463
Joined: Tue Dec 24, 2013 7:02 am

Post by JohnyM »

@mawyatt
Thanks, i'll keep that in mind when i get to "power it all up" stage.
ray_parkhurst wrote:
elf wrote:You can find the TR8 leadscrews on Aliexpress in 1,2,4, and 8mm pitches. I have a 500mm long TR8-1mm for another project. The quality appears to be excellent. You will need to be able to machine the ends in order to make it fit the rail.
The leadscrews are no problem to find, but I'm wondering if the C-Beam nuts are available in these tight pitches...

Found one seller on aliexpress selling 2mm pitch c-beams:

https://www.aliexpress.com/item/2pcs-TR ... 6512CWiotB

These are brilliant looking! Single pieces of machined delrin with play-reduction tightening mechanism. When did these become available, and who invented them. Pure genius. With a good quality acme screw this is a super solution for low-load applications.
Im no expert on this, but it's been around for years.
From what i've read, it's not as brilliant as it seems. Backlash is there for a reason:
-to allow manufacuring tolerances
-for greasing
-to extend system life expectancy

Those nuts either dont elimiate it OR increse system wear overtime with backlash returning.

Best way (from what i've read) are spring loaded mechanisms.
This is simple marvel to me:
https://www.researchgate.net/figure/Str ... _280255199

Example of spring loaded nut:
https://www.helixlinear.com/media/57403/117984.png

Spring loaded nuts are supposed to "eliminate" backlash completly when rail is operating while still allowing for greasing and manufacturing errors via spring giving up.

In microscopes it's simpler as gravity does the job. But in some focusing blocks there is space (or they are already instlled) for springs, if that bliock is working horizontally (Nikon).

Hence my decision:
- Stick to proven rails
- Spring load all rails
- Enclose them all in V-mount with wheels serving as idler rail

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

One needs to try it before dismisses it, This rail has been there for long time, I now remember it was on Kickstarter when I had my water drop project there, I think it was 2013. Yet, no one has thought about using it, it is a shame that this community has missed it for so long.

I searched this site, there was a discussion long time ago, yet, it was dismissed NOT based on actual usage, but based on some "analysis". I am a firm believer of actual facts even if something goes against my "theory" or understanding of things -- sometimes, my understanding is limited, particularly in a field I do not have a lot of knowledge.

soldevilla
Posts: 684
Joined: Thu Dec 16, 2010 2:49 pm
Location: Barcelona, more or less

Post by soldevilla »

I have not faced the design of a lane without a game because of my limitations in electronics and programming, but I think that after this thread and the limitations of my old StackShot (which I must update its firmware) I'm going to start designing a lane without game and high precision. First simply by playing and then I will see if I get to build it. So many years working as a mechanical engineer must serve me for something, I hope ... :D

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

soldevilla wrote:I have not faced the design of a lane without a game because of my limitations in electronics and programming, but I think that after this thread and the limitations of my old StackShot (which I must update its firmware) I'm going to start designing a lane without game and high precision. First simply by playing and then I will see if I get to build it. So many years working as a mechanical engineer must serve me for something, I hope ... :D
Good we got a mechanical engineer here :-). Maybe you can start with one Arduino solution that was done LONG time ago, possibly 10 years ago, by a member here, I do not remember that link, but that was pretty impressive.

If that link can not be found, I think I might do one after all just for this rail. But it will not be as advanced as my controller.

By the way, in the spirit of openness, it is not that I do not want to disclose how I designed my controllers, it is just that there are a lot of people who just pick it up and claim it as public domain and copy the design because I DISCLOSED it on my own. I got someone who are even more devious than others -- they pretend to be your friend and ask a lot of detailed questions, and a month later, they push out their own product (in water drop equipment area).

On the other hand, if they, or their agents, get one of my controller in their hands and reverse engineer it and thus committing that "act", though, not legally binding, it is just something intentional, and a normal professional would not do.

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

JohnyM wrote: Best way (from what i've read) are spring loaded mechanisms.

Example of spring loaded nut:
https://www.helixlinear.com/media/57403/117984.png

Spring loaded nuts are supposed to "eliminate" backlash completly when rail is operating while still allowing for greasing and manufacturing errors via spring giving up.
This spring-loaded nut is what I have on my shelf, waiting for time to experiment. I agree that it seems the better way to go, and is constantly adjusting for variation in the screw characteristics, wear, etc.

But the C-Beam nut does show promise. The PTFE-loaded Delrin will last a long time under light loading conditions, and if it does start to wear, it can be tightened to compensate. It is also a "complete" solution with integrated mount, and is very compact, taking up very small space. As mjkzz says, I still think it's worth trying. Personally, I will stick with the industrial rails (THK), but it's fun to experiment.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

soldevilla wrote:I have not faced the design of a lane without a game because of my limitations in electronics and programming, but I think that after this thread and the limitations of my old StackShot (which I must update its firmware) I'm going to start designing a lane without game and high precision. First simply by playing and then I will see if I get to build it. So many years working as a mechanical engineer must serve me for something, I hope ... :D
I'm and EE but not a programmer, so this DIY effort was rather daunting for me without any real prior programming skills/experience. The mechanical stuff went OK, but I don't have the precision tools available, so maybe a little harder than someone with proper tools and knowledge like yourself.

The electronics are very easy now, almost anyone that can solder a few wires, can build their own controller if they wish. Of course the Stackshot controller you have is superb, and the Wemacro and MJKZZ controllers are quite good also, but if you want to "roll your own", then a Raspberry Pi 3B ($35), Pololu Tic-500 ($19) and a couple LEDs, resistors and connectors and you are in business!!

The software was/is the daunting part for me, especially since I didn't want to use any prior controlling motor code that might be for other macro rails, CNC or 3D printers, or adapted from the Adurino. I want the end result to behave the way I want, and not a version of someone else's work, thus the decision to write all my own code. The Pololu folks have produced an amazingly sophisticated stepper motor controller series that are relatively easy to use, and hide all the complex details behind simple commands. I have and use the Stackshot controller, which is my favorite (especially when operated from Zerene). I use it with all sorts of motors and rails, also have and use the Wemacro and MJKZZ systems, they are also very good.

I would think the use of the "Open Source" rail with any of the programmable controllers should work well.

As one engineer to another, I'll be happy to help if you decide on the route of DIY with the Raspberry and Pololu, not sure I would be much help with Adurino though!!

Best,
Last edited by mawyatt on Wed Nov 14, 2018 9:40 am, edited 1 time in total.
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

ray_parkhurst wrote:
JohnyM wrote: Best way (from what i've read) are spring loaded mechanisms.

Example of spring loaded nut:
https://www.helixlinear.com/media/57403/117984.png

Spring loaded nuts are supposed to "eliminate" backlash completly when rail is operating while still allowing for greasing and manufacturing errors via spring giving up.
This spring-loaded nut is what I have on my shelf, waiting for time to experiment. I agree that it seems the better way to go, and is constantly adjusting for variation in the screw characteristics, wear, etc.

But the C-Beam nut does show promise. The PTFE-loaded Delrin will last a long time under light loading conditions, and if it does start to wear, it can be tightened to compensate. It is also a "complete" solution with integrated mount, and is very compact, taking up very small space. As mjkzz says, I still think it's worth trying. Personally, I will stick with the industrial rails (THK), but it's fun to experiment.
Ray,

I doubt anything you can do mechanically to these Open Source rails will make as good an overall rail as the THK. Every THK I have has no observable backlash nor wobble.

Agree, it might be fun to play with these though, especially if you have a DIY controller :D

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Post by elf »

JohnyM wrote:Those nuts either dont elimiate it OR increse system wear overtime with backlash returning.
Actually, Acetal (Delrin) has a small amount of flex in it which allows nuts to ride firmly on both flanks of the thread. The split nut design can compensate for wear by applying more pressure with the set screw. A spring loaded nut system will also wear and the pressure of the spring compensates for it.

I have an Acetal nut on my macro stitching and stacking setup which has no measurable backlash. My rose engine uses TR8-8mm leadscrews with spring loaded anti-backlash nuts and it also has no measurable backlash.

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

mawyatt wrote: I doubt anything you can do mechanically to these Open Source rails will make as good an overall rail as the THK. Every THK I have has no observable backlash nor wobble.

Agree, it might be fun to play with these though, especially if you have a DIY controller :D
For sure the best mechanism for these rails is the pre-loaded ball screw with pre-loaded crossed ball bearing slider...eg THK, Misumi, HiWin, etc. but a simple linear rail based system with acme screw could be precise enough for XY application. I'm especially fond of very compact arrangements, and I tend to do a lot of DIY just putting together my THK systems (adding idlers and base plates, etc) so it will be interesting to see if I can make something small and effective with these DIY components.

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

elf wrote:
JohnyM wrote:Those nuts either dont elimiate it OR increse system wear overtime with backlash returning.
Actually, Acetal (Delrin) has a small amount of flex in it which allows nuts to ride firmly on both flanks of the thread. The split nut design can compensate for wear by applying more pressure with the set screw. A spring loaded nut system will also wear and the pressure of the spring compensates for it.

I have an Acetal nut on my macro stitching and stacking setup which has no measurable backlash. My rose engine uses TR8-8mm leadscrews with spring loaded anti-backlash nuts and it also has no measurable backlash.
That is how I feel, though I am not an mechanical engineer at all. I do have some concern with "elasticity" of delrin at very fine step (say 0.3125um), ie, the delrin probably just deform than moving the gantry till deformation is no longer possible and the gantry will jerk and move. Of course, this is my "theory", not fact. I will probably do a more serious check using Zerene, borrowing Rik's idea he outlined looong time ago :D

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

ray_parkhurst wrote:
mawyatt wrote: I doubt anything you can do mechanically to these Open Source rails will make as good an overall rail as the THK. Every THK I have has no observable backlash nor wobble.

Agree, it might be fun to play with these though, especially if you have a DIY controller :D
For sure the best mechanism for these rails is the pre-loaded ball screw with pre-loaded crossed ball bearing slider...eg THK, Misumi, HiWin, etc. but a simple linear rail based system with acme screw could be precise enough for XY application. I'm especially fond of very compact arrangements, and I tend to do a lot of DIY just putting together my THK systems (adding idlers and base plates, etc) so it will be interesting to see if I can make something small and effective with these DIY components.
Definitely, it is hard to beat ball screw and all those industrial strength rails, though I do not see how, there gotta be a reason (or reasons) the C-Beam is not used in industry automation. :D

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

One reason C beams and other aluminum base rails might not be popular with precision use is the difference in thermal expansion between Al (~22um/mC) for the base and stainless steel (10~17um/mC) for the screw. A few degrees C range in temperature could create a bending moment on the screw, in addition to the obvious length changes.

This becomes an interesting topic when one starts talking about sub-micron stepping and positioning. On a 200mm rail the screw length will change by 2~3 microns for just a 1 degree C change in temperature!!

I've noticed the THK KR type rails seem to use an alloy for the base, this might be to thermally match the TC of the screw. Since we have some mechanical engineers here, maybe they could elaborate on this more. I haven't actually done any tests, so this is just an observation on the THK rails.

BTW plastics have extremely large TCs, from 40~120um/mC, this range includes PVC, Nylon, Polycarbonate and so on. Long ago we used a material called Invar for a precision moving mirror interferometer, this was because of it's low TC of just 1.5um/mC.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

mjkzz wrote:
mawyatt wrote:
ray_parkhurst wrote:The mjkzz XYZR/SNS works well with the THK KR15 with USA and MachMotion motors. Only issue I had was OS compatibility, but I am using the XYZR/SNS on Win10 and it works fine. Motor current (both idle and peak) can be controlled so no overheating.

The C-beam is very interesting. I assume it has two delrin nuts, on in the main part and one in the narrow part, and that the screw adjustment is used to remove the "play" in the system? I'd worry that small variations in the acme screw would cause "tightness" of the system to vary quite a bit over the length of the screw. Would be interesting to test these out though. Do they come in tighter pitches, like a more standard 2mm pitch / 8mm diameter type? I suppose the tighter pitch may give the system more problems. Wider pitch is probably a lot more forgiving.
Ray,

The MJKZZ Stand Alone Controller doesn't have a current limit set point, at least it's not available on the display, so the one you mention must be the USB version, which Peter indicated does have current limit control. Do you know what stepper driver chip is used in the USB version?

Best,
Mike, all of my controllers have current limiting capabilities. For the stand alone one, it is on the second page, press PM to see i.
Peter,

Yes it does!! I didn't know (or just forgot), that your Stand Alone Controller has the current limit setting on the 2nd page when using the optical remote key fob PM button.

I also just confirmed that with the current limit set to 625ma the USA KR-15 rail works fine with this controller.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Post Reply Previous topicNext topic