Getting pictures of the inside of these actuators was a bit tricky. But then I realized these companies like to brag with their patents:
So the wires are indeed moving in both directions. The kinematic is surprisingly simple. There is a 3 DOF and 6 DOF version. Each wire restricts only one DOF so 6 wires fully constrain the stage.
Overall a super interesting mechanism. For larger movement it will have some cosine error. For an automated system one could probably just correct for that in software. For manual actuation it might be a bit annoying when the stage moves out of focus on lateral movement. I wonder if there is a simple way to get truly linear motion with double parallelogram flexures for the z axis only and keep x and y with single flexures or flexure pairs.
First a XY stage that is suspended by two double parallelogram flexures. I connected the two intermediate stages to make this more stable. I am still not sure if it is needed or even helps at all.
I cut this pretty small: 7 x 8 x 8 cm. As a result the x travel of 15mm works still well but the y travel is only like 6mm (3mm in each direction). The blades with 3 by 1.5mm are 4 times stiffer in y direction than x.
The line is very slippery so I tried holding it in place with hot glue. This work reasonably well but is a mess - both when applying and from the look afterwards. The flexures are 2mm long and on the soft side. They would probably need a thicker line or being shortened. The stage has though very good range of motion but the flexures are not that stiff against sheering or collapsing under excessive load. But may be this is an acceptable price to pay for a large range of motion.
Then I tried to wrap my mind around the Thorlabs patent and whether double parallelogram flexures can be used there. It took me a while to figure out how everything fits in 3D space. But the result is surprisingly straight forward:
This is the first attempt and there are still a few minor clearance issues. E.g. the flexures are too long and canât move inwards. But these are easy to fix. The stage is still missing. It will be a square on top directly connected to the z flexure. There will be two triangle pieces connected underneath covering the corner of the cube we are looking onto that are connected to the other two flexures.
To actuate the axis a bar will be added across the cube. The simplest way is just to add a screw in the opposite wall. For z -and may be even y - Iâll add a linkage to divert the movement by 90 degrees to not need the crew on the bottom. This should be pretty easy to do (and similar to what Thorlabs does with their stage).
This thing is pretty big. After the xy stage I wanted to give the flexure the best chance to work. They can easily deal with the 20mm travel in each direction, although fixing them to the stage will alter the way they need to bend. The cube is 15 cm nominally but will end up more like 17 or 18 cm in each direction. As such it is big enough to probably house all the actuators and electronics inside.
I designed a double-parallelogram flexure stage for 3D printing, a while back, and definitely recommend it as a design. Iâm not sure wood is the best material for flexures though.
Yeah, this idea is to use wood for prototyping and then with to another material - probably POM. But it turns out Wood doesnât perform that badly - especially for these long blade flexures. The Issue is more that is does not provide the needed width to height ratio for the blades.
I have fallen into an optics, illumination, Köhler design rabbit hole lately. This is the reason why there have not been any news on the mechanics. To steer myself back to that I thought about doing a simple, one axis stage for the illumination that can go on top of the regular, inverse OFM.
It would be kinda cool to be able to use that for both the illumination and the optics module as an alternative to the upright microscope. So an illumination module with the mechanical interface of the optics module would be handy⊠Wait a secondâŠ!
Has anyone tried to use the upright components to assemble a mechanised illumination stage for the regular OFM? That should just work IMHO. Just put the upright Z axis and upright illumination module on top.
I might still design a laser cut version but itâs kinda cool to discover that this should work already.
Wood has a similar problem to 3D printed parts, thereâs a preferred orientation. In order to be any good, flexures have to be printed with their cross section perpendicular to the Z axis so theyâre isotropic in the plane of motion. Itâs harder to guarantee that with a laser cut piece of wood.
Sure, one needs to be aware of the limitations of the materials and the manufacturing processes. But I donât see a fundamental difference between 3D printing or laser cut plywood and EDM or water jet cutting. None of these can create flexures in arbitrary orientations. Iâd argue that 3D printing is probably the most versatile here as upright flexures of one or two wall thicknesses should also work. Everything else needs to assemble the mechanism from separate parts to have the right orientation in the flexures. Your XYZ stage does seem to be any different in that regard.
High Resolution Optics Module with the cheapest 35mm (aka 185mm) objectives on Aliexpress
Upright Illumination Module without a diffusor but with a 5mm LED
No steppers or Sanga board
Lasercut z axis for both modules even if the one for illumination doesnât quite work right
No X or Y axis at all
Cheapest pre-prepared slides on the internet
Still this seems to work kinda fine. The thing I am a bit confused about is that there doesnât seem to be an easy way to adjust the illumination/camera gain to the specimen. Metal microscopes have 3 different knobs to adjust the illumination and to noob like me it looks like they are actually used quite a bit. Yes, I understand that we (well, I) donât have Köhler Ilumination. I still had expected an easy way to adjust either the LED brightness or the camera gain/exposure time or both. Am I just missing something? Is everyone else just using the auto detected values? Or adjusting them in the Settings if necessary?
In the camera settings there is an auto gain and exposure button that will adjust the brightness without altering the colour correction/ lens shading table.
It would be useful to have brightness adjust in the image panes of the web app as well.
