Building A Laser Head With High Speed, High Resolution

A test exposure on cyanotype paper shows off the prototype’s resolution, around 100 microns.

Typically, when it comes to scanning a laser, it’s done with galvos or a rotating mirror assembly. However, these methods can be slow and cumbersome, or restricted due to existing patents. [Rick] aimed to find an alternative solution with the Hexastorm project, using a rotating prism to build a high speed, high resolution laser head.

The project currently uses a Beaglebone for the brains, with a polygon motor sourced from a photocopier used to rotate the prism at over 20,000 rpm. The project aims to be a proof of concept for rotating prism technology, which can then be adapted to specific tasks. With the promise of high speed and high resolution, the system could be used in fields as diverse as PCB manufacture, 3D resin printing, and even virtual reality displays. [Rick] explores these potential markets in a pitch deck, comparing to existing solutions in the marketplace.

If you’re interested in high performance optical systems, [Rick]’s work makes compelling reading. It’s not the first time we’ve explored cutting edge laser hacks, either. Video after the break.

9 thoughts on “Building A Laser Head With High Speed, High Resolution

  1. Resolution is somewhat relative, it all depends on the way it is scanned

    A raster scanner like this one can have a great resolution in one axis but will always lack in the other, unless the second axis is based on a resonating scheme at a specific angle it will be extremely hard to achieve high performance in both axes

    In raster scanning mode, a MEMS mirror with a XY resonating movement can be optimized for a resonating movement at specific angle and amplitudes, which can give good performances with a ratio between axes depending on the physical size of the mirror.

    In vector scanning, a XY galvo mirror set of ten years ago was usually 100k to 200k points per second at 8 to 10 degrees of incidence, so this must be better now

    It all depends on the use you seek, and I’m not counting the laser itself which has much parameters to take into account (profile, mode, beam characteristics, collimation, coherence length, etc…)

    1. The laser head is indeed discretized in one direction. MEMS mirror with resonating movement are not discretized in both directions. Still they require a scan lens, e.g. f-theta lens and my system does not. A mems mirror can also be slow and does not have constant speed. The mirror has to slow down at the end of sweep and revert direction.
      All the players both in the photo-polymer and the PCB market with laser direct imaging use either galvo scanner or rotating mirrors, e.g. Formlabs, Envisiontec, Orbotech and Manz.
      Rotating mirror can go very fast. For instance high speed cameras like the Cordin 131 HD (http://www.cordin.com/pdfs/Cordin131-HD.pdf) rotate at 7,500 revolutions per second. I have even seen high speed camera with mirrors rotating at 1.2 million revolutions per minute.

  2. Good point! The system is indeed discretized in one direction. A XY MEMS mirror is not discretized in both directions. Still it requires as a scan lens, e.g. f-theta lens, and can be slow as it has to slow down an change direction at the end of sweep. All players in both the photopolymer and the pcb laser direct imaging market use either galvo scanners or rotating mirrors for example; orbotech, manz, envisiontec and formlabs. Formlabs ditched the xy galvo scanner for a galvo raster scanner in the form 3.
    I have seen polygon scanners reach speeds of 1.2 million RPM. For instance the Cordin Model 131-HD reaches a speed of 7500 revolutions per second and is used as a high speed camera.

  3. “Rotating mirros” is still covered by patents? That would be a disgrace. Otherwise i see no upside to rotating prisms relative to rotating mirrors, just downsides such as higher weight, chromaticity, absorption,…

    1. Interesting points. Disadvantages are indeed that light of different colors refracts differently and the reflection of light when it goes from one medium to the other. Laser light has almost one wavelength so “chromaticity” is not a problem.
      Quartz does absorb light but this is irrelevant here. The main losses are due to reflection. This can be minimized by coating the glass. The reflection of a mirror is also dependent on the angle. Scanning mirrors need an f-theta lens; the prism doesn’t and has a telecentric projection. The key advantage is that If a mirror has a facet to datum error of one degree, the total error is two degrees. If a prism with a refractive index of 1.5 has a facet to datum error of 1 degree, the total error is 0.5 degrees, see deviation angle prism. This is key for the cross scan error. At the moment, the prism has the same weight as the mirror. I actually made it lighter as I was afraid it wouldn’t rotate.

  4. The fact that patents actually slow down development is irritating to say the least. The whole idea of patents was (and still is, in theory) to make inventions avail- and usable for a broad public. The need to look for second-best or even worse alternatives to what has been proven to work fine (and is covered by a patent) is the opposite of progress (except when you discover an even better method than the one you were trying to avoid in the first place).

    1. Thanks nitpicker, the practice of patents is indeed surprising. I do think innovators should be rewarded. The current practice is expensive, requires extensive legal expertise and deep pockets to obtain and defend a patent.
      In practice you can also get a patent if you can come up with a new application. So if no one was using polygon mirror, laser diode in photopolymer 3D printing. You can get a patent. Even thought this would have been obvious for any bachelor physics student looking into this topic. Still there is also a lot of randomness involved. It is impossible to know in advance if and how the patent will be awarded.
      I don’t think the rotating prism system is inferior. It is however the new kid on the block. Incumbents don’t like that and customers are skeptical. Which system is better depends on your metric. The criteria you use and the weight you give them.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.