980nm fine focus lens

Hi all,

I'm new here so forgive me if I haven't been able to find the answer, but I'm chasing a lens to focus a 980nm diode down to the smallest spot I can at around 50m.

I'm looking to make a low angle scanner for about 50m distance with a very small spots size for drawing thin lines at a large distance. Any suggestions on galvos and optics would be great. Especially if I can use the same optics for other wavelengths such as 670nm and 532nm.

What purpose does an infrared line at 50 meters have? Tell us more about your project, please.

It servers as an active sensor for a robot. Some monitoring equipment looks for the line and measures the distance to it. The accuracy and precision depends on many factors, but the thickness of the line is extremely important because the sensing equipment can measure down to micro meters but if the line is centimeters in size, it's not much good for figuring out an accurate distance. The other problem is, that as the line gets thicker, the power needed to make it visible increases. So ideally, a very narrow line, 1mm or less is ideal. Less power requirements and better accuracy.

Projected onto what? In what sort of environment? What is your budget? How much space are you limited to? Why 980nm? Is this an upgrade you're making, or is nothing built yet?

Can you elaborate on the sensing equipment? I am quite skeptical of the micrometer claim.

Projected onto dirt, concrete, grass, metals, (that's part of the experiment really), Budget is tight at under 1k. Space isn't too much a of a problem but strength is, it is to be mounted on a moving platform. 980nm at this point because it is significantly safer given the time and dose since it is possible it will be used in a populated environment with at least animals in the vicinity and possibly humans as well. Why other wavelengths then? easy, because given a few wavelengths, I can tell what it is that I've projected onto. It is a scale up of what I already have working with a Powell type lens. I've had all this working on a small scale but scaling up the current system seems impossible given the lens I have can't focus down fine enough to give accurate results at such a distance.

The sensing equipment is on loan from the university. It is radically more accurate than micrometer, but given the calculations for surface reflectance, noise and diffuse positional shift I expect in some surfaces, micro meter is all I can expect. The sensor of use is a narrow spectra camera that I can select the wavelength of detection with a control on the pass band between 1 and 10nm. With a long focal length lens on this thing, it's all up to how well aligned everything is and the right Signal to Noise Ratio and of course, the dependance on seeing the intensity distribution without saturating the scene.

I'd be happy to get a 1mm line at this distance. I can do the subpixel interpolation to reduce it to micrometer accuracy.

The simplest way I can think of is a particularly large cylinder lens as your only optic.

Leaving one axis alone will allow it to diverge at its typical 40 degrees or whatever and form a line that way. Then with a cylinder lens, focus the other axis at 50 meters. It would need to have a long focal length and be quite large. I don't think you'll find a cylinder lens big enough, but maybe some sort of fresnel solution? Or maybe you could make your own by bending a piece of plexi, fitting it against a flat one, and filling it with water or something like that.

There are other ways to do it, but it would require something like collimating, reshaping, expanding, and collimating again which gets hairy.

I have access to a small CNC but it's not capable of milling glass. Can I get away with perspex? It doesn't have flood coolant or a high speed output so the best I can mill on it is perspex with a good cutter.

I just haven't got enough experience in designing lenses to know if this machine is too small for the job.

Water absorbs IR at 980nm, it wouldn't pass but I can make a solid perspex piece up to 153mm x 300mm x 60mm

Any links to the lens makers equations? I know the output angles of the diode, I figured I'd have to settle for only shaping one axis but one is fine I think. Even if it does diverge on the other at some angle well above, so long as the line is thin, it will be visible to the camera.