What's new
What's new

Optical instrument for PCD insert's setting

I don't think he was claiming that was the pixel size, but rather the image of a 1 micron move on the sensor, after optical magnification. But maybe I misunderstood?

That's not what I gathered from this statement:

We measured the real resolution of our system directly on the machine. We moved the cutter relative to the microscope by 1 micron and measured how many pixels it moved. The program from the camera measures pixels exactly on the camera matrix, not on the laptop screen :) Therefore, 4 pixels per 1 micron is a fairly real value, I'm sure of it. Of course, it could be 3.5 or 4.5 pixels.

But maybe I misunderstood. You are probably right.
 
Interesting problem, but helping the aggressor make war parts isn't on my list of things to do.
 
Lately I've been experimenting with a lens from some kind of projector. In principle, I managed to get the desired magnification, but the image quality is not very good yet.
All the elements just stood on the table, so the external light spoils the quality.
I tried two schemes of work: lighting through a narrow gap (the gap between the jaws of the caliper) and lighting the contour of the diamond insert.
Below are a few photos of my super precision optical setup)))))) Obviously, I need to deal with external light, in the near future I want to at least insert pipes from a vacuum cleaner between the elements.

1. Here you can see: a light bulb - a green light filter - a caliper with a gap between the jaws of 0.22 mm - a lens from the projector.
DSC_8522.jpg

2. An image of a light strip on some sheet of metal. With a gap size in the caliper of 0.22 mm, the width of the projected strip is 4.5 mm, an increase of 20+ times.
DSC_8528.jpg

3. Here I am already starting to use the camera. The distance between the lens from the projector and the camera is about 1 meter to achieve the desired magnification of 20x.
DSC_8530.jpg

4. In order not to make such a huge construction, I inserted the camera into a small handheld microscope with an x8 objective.
DSC_8531.jpg

5. And here I am looking at a spent diamond insert. I did not have a new insert with me (I did experiments not in the workshop with the machine).
DSC_8536.jpg
 
One of the best results so far with the observation of the light gap. Slit width 50 microns. As you can see, the resolution on the camera turned out to be about 3.5 pixels / micron.

2022-02-28_17-58-10.jpg
 
Finally found the time to test the lens I just received. Its characteristics are very close to a similar lens from Mitutoyo - Ф=65mm, x10 magnification. Only it costs not $1200, but only about $33 :) Today I assembled a simple stand from a sewer plastic pipe, a camera and a lens. I wrapped it all up with duct tape.
The resulting resolution is 3 pixels/micron. Not great, but not bad, just enough. Below is an image of the measure of the reference line, between the two small lines there is a distance of 10 microns. The length of the red line is 306 pixels. 306/10=3.06 Now, apparently, we need to defeat laziness and make a good steel pipe with precise seats for the lens and camera.

2022-03-15_18-17-28.jpg
 
Hello. I will continue the story of my experiments to deceive reality :)
Last week I completed the optical tube, and together with our technologist, we adjusted the diamond inserts on the face mill. I must say right away that the result of the work of the mill turned out to be good. In my experience, the accuracy of the inserts in the face mill is within two microns. In my experience, the accuracy of the inserts in the face mill is within two microns.
But, as you can see from the photo below, the picture quality is far from ideal. The edge of the plate is not clear.
Our rechargeable flashlight, which has always been used to expose inserts, turned out to be disastrously weak for this optical system. As a result, I had to use the flashlight in my smartphone, which, it seems, can send an SOS signal if I find myself on a desert island.
What I plan to do in the future:
- buy a powerful source of directional light and fix it so that the light beam is directed exactly at the center of the lens
- add a green or blue light filter to the light source
- I need to try to shine on the insert from the other side. Now the light hits the back surfaces of the insert, and it seems to me that the light can "go to the sides" and because of this the edge is not so sharp. Perhaps, if you shine on the front surface of the insert, which is strictly perpendicular to the light beam, the result will be better.
- I need to try to work in reflected light, and not in transmitted light, as it is now. It is possible to apply a combination of reflected and transmitted light.PXL_20220413_151843643.jpgPXL_20220413_151852406.MP.jpgPXL_20220413_151901760.jpg
 
A light stop of a diaphram or just a piece of black construction paper with a small hole located at your lens focal point in the tube will help a lot.
An easy way to find this distance off the lens is to go outside and point the lens at the sun.
Now move it up and down to find the smallest and brightest spot behind the lens on something that will not start on fire.
Do not use your finger when lens pointed at the sun !!
This is where the stop needs to be. The clear sky sun is a great low cost point source of collimated light.
What this does is allow only the somewhat straight light beams through and tosses out those at angles giving better focus and depth of field.
As such one needs to pour more light into the system.
Ideally one would use a infinitely small pinhole but that not practical and no lens is perfect.
Also at this tube length you are far outside the lenses corrections so a color filter would help but not as much as a stop.

IMO a very good effort. From the pics I think I have and use the same camera family and software.
Bob
 
I forgot to say - many thanks to everyone who on the previous page of the topic offered to look in the direction of photo projectors. This lens is designed just for them. Its regular distance between the object of photography and the photographic plate is 790mm or so. An additional detail is visible on the pipe next to the camera. Without it, if the camera is directly screwed to the pipe, the distance will be exactly equal to the regular one. And with the extra extension piece, it's 100mm larger, which should increase magnification. I made this part interchangeable, because I was not sure that this distance could be increased without compromising quality. In the course of work, I tried to screw the camera directly to the pipe and get the standard distance between the object and the camera, but this did not affect the blurry edge of the insert.

A light stop of a diaphram or just a piece of black construction paper with a small hole located at your lens focal point in the tube will help a lot.
An easy way to find this distance off the lens is to go outside and point the lens at the sun.
Now move it up and down to find the smallest and brightest spot behind the lens on something that will not start on fire.
Do not use your finger when lens pointed at the sun !!
This is where the stop needs to be. The clear sky sun is a great low cost point source of collimated light.
What this does is allow only the somewhat straight light beams through and tosses out those at angles giving better focus and depth of field.
As such one needs to pour more light into the system.
Ideally one would use a infinitely small pinhole but that not practical and no lens is perfect.
Also at this tube length you are far outside the lenses corrections so a color filter would help but not as much as a stop.

IMO a very good effort. From the pics I think I have and use the same camera family and software.
Bob
This is an interesting tip, thanks a lot! I will definitely try to do this. The lens has a built-in adjustable aperture, but I forgot to try to change it during operation - it was completely open.
 
I forgot one more important thing for someone who decides to repeat my design.
The pipe from the inside must be made matte black. Of course, it was possible to give this part for anodizing, but the pipe is made of aluminum, and the bushings at the ends of the pipe are made of steel. I wasn't sure if anodizing two different metals would work well. So we just rolled up a sheet of matte black paper and stuck it inside through the hole for lens.
 
Last edited:
Black construction paper or matte poster board is a common optics trick. If your angle is shallow enough, sand blasting, even if left silver, can be surprisingly effective.
 
Flock paper is common. Black anodize does not work at all.
One can also do "light stops" in places so the camera never sees the tube reflections.
Long tube here.... How to kill all reflected or bouncy stuff you do not want hitting the sensor.
The OP is up against some hard walls and trying stuff so I have to say a big huge hats off to that.
Bob
 
A handy method of killing side reflections in a tube is to use a coiled spring (around 10mm pitch, maybe 1 or 1.5mm wire thickness) inside the tube. The spring should be flat black. Common is bras with a black finish.
 
Another is forming grooves similar to a 60° thread form on the inside of the tube, then applying a matte black finish. (I would like to try some of that "blackest material ever" stuff). And yes, this is very important to kill image artifacts and give good contrast. I remember reading that in the olden days they would even use carbon black, or soot. A good method to get it was a rich acetylene flame.
 
Last edited:
As we go off into neverland.
Best optics design/ray tracing software for such projects at low cost and for easy to understand and use?
Zemax and Olso not so simple for this type project.
Now there must be basic type stuff out there for free that is far above the basic lens formula.
Bob
 
Another is forming grooves similar to a 60° thread form on the inside of the tube, then applying a matte black finish. (I would like to try some of that "blackest material ever" stuff). And yes, this is very important to kill image artifacts and give good contrast. I remember reading that in the olden days they would even use carbon black, or soot. A good method to get it was a rich acetylene flame.

Going down the rabbit hole here:

You might be thinking of Aktar’s products. Sort of like fine velvet in appearance. Some not as dark as you would think from the media, the rest, generally hard to maintain except in an enclosed setting. IE, ok inside an optical system, but don’t leave a piece sitting uncovered on a dusty workbench, or even on a not so dusty workbench.

If you stick to their products that work well at shallow angles, looking at them under a microscope they are all pretty rough. Some in a structured way, others more random.

Good baffle design along with a sharp thread like profile on the walls solves the problem about as well. You just need to worry about the rounded tips on the thread pattern.
 
I think he was referring to Vantablack, which is a proprietary black pigment with better than 98.5% absorption across the visual spectrum. The particles are actually arrays of nanotubes, meaning it is very porous at the microscopic level rather than a solid chunk of mineral or crystal. Light is absorbed within the nanotubes, rather than reflected.
 








 
Back
Top