Transparent Polygon Scanning

From RepRap
Jump to: navigation, search

TPS (Transparent Polygon Scanning) is a laser scanning technology. The technology can be used for additive manufacturing. The first transparent polygon scanner with a single laser bundle was described by Lindberg in 1966. Rik Starmans developed the first transparent polygon scanner using a single laser bundle and named it Hexastorm, which is a pending trademark. Hexastorm is shown in figure 1.

Figure 1

The Reprap movement started with Fused Filament Fabrication (FFF) technology. TPS is very different but to put it in perspective. The smallest element of a FFF enabled printer is the nozzle which is a circle with typically a 300 micrometers diameter. In the proof of concept laser scanner with TPS technology, this is a circular laser spot with a 25 micrometers diameter. The standard speed of a FFF printer is 50-80 millimeters per second. TPS enabled printers are able to reach a spot speed of 6.8 to 34 meters per second. The maximum scan length is 24 mm. The speed of a whole scan line is 16 to 84 mm/s.

Description

The bleeding edge of Hexastorm can be found in a blog on Hackaday. A video is available on YouTube. The current transparent polygon scanner consists out of a laser diode which is collimated with an aspherical lens. The bundle is then focused by a cylindrical lens in a direction parallel to the prism. The bundle is deflected by refracting it through a tilted transparent plate. The bundle is finally focused by a second cylindrical lens orthogonal to the scanline. The position of the laser bundle is monitored by a photo-diode. A Beaglebone is used to ensure the correct timing of and stream data to the laser diode. A schematic side view and top view of the system are shown in figure 2 and 3. The two cylindrical lenses ensure the bundle is circularized and the cylindrical lenses minimize the scan error per facet due to manufacturer imperfections of the prism.

Figure 2
Figure 3

A substrate can be solidified by moving the optical head in a snakelike pattern over the substrate as shown in figure 4. A more advanced scanning technology is infinite field of view.

Figure 4

The system has four advantages; it has a high optical quality, is cost effective, scalable for industrial applications and has extensive freedom of use. The system has a high optical quality as the bundle is in focus over the full scan line and incident at 90 degrees. The transparent polygon scanner gives a flat field projection and the system is telecentric. Details are provided in the physics section. The system is cost effective as it is able to achieve this quality without an expensive f-theta lens. A reflective polygon scanner needs a thicker disk as the f-theta lens typically has a longer focal length than the second cylindrical lens. A longer focal length requires a larger collimated bundle diameter for the same focused spot size, see physics section. A reflective polygon scanner also needs a system to correct for facet aberrations if the facets are not perfect. Thirdly, the system is scalable. The maximum optical power of laser diodes becomes less as their wavelength becomes shorter. As a result, laser diodes need to be combined to give more power. It is, however, not possible to combine more than two lasers into a single bundle without interference. The electromagnetic field only allows for up to two polarizations. Companies like Kleo Halbleitertechnik therefore sell systems which use 288 laser bundles, see video. Companies which use single bundle transparent polygon scanners, like Hexastorm, have a scan angle of 90 degrees. Kleo uses multiple bundles per polygon and therefore has a scan angle of 45 degrees or less, for details see physics section.

Figure 5
If multiple transparent polygon scanners are used in a combined fashion, see figure 5, a substrate can be solidified at once and the snake like pattern is not needed. This might be desirable in industrial production systems. In figure 5, the scanheads are stacked both horizontally and in the "depth" of the image. The depth stacking is denoted by using a different color. This stacking is needed in two directions to ensure that the scanlines overlap and the invidual scanners do not colllide. In figure 5, also a camera is depicted which can be moved under the transparent polygon scanners to align the scan heads. The camera would monitor the location of the scanline for each scanhead. This can then be accounted for by the algorithm that calculates the data sent to the scanhead. If a laser fails in an array, the array can be fixed by replacing the scanhead or part of the scanhead, e.g. polygon or laser bundle with aspherical lens.

Finally, the system has an extensive freedom of use, it is open and not patented. Various companies have created patents in the field of reflective polygon scanners. These companies overlooked transparent polygon scanning, although its possibility was indicated by Lindberg in 1966. This allows the open-hardware revolution to enter other industries such a PCB manufacturing.

Specifications

The Hexastorm, the first single bundle transparent polygon scanner in the world, has the following specifications:

  maximum scan length: 24 mm
  typical scan length: 8 mm
  wavelength: 405 nm
  revolutions per second: 67-350 Hertz
  spot diameter: circular, 25 micrometers
  cross scanner error: 40 micrometers
  laser driver frequency: 2.6 MHz
  optical power: 300 mW
  facets: 4

The scan speed of a transparent polygon scanner is not uniform and varies slightly. The scan speed at the center is 80 percent of that at the edges of a scan line of 8 mm. The non-uniform scan speed is mitigated by using a high speed laser with a 50 MHz pulse rate. The lower power at the edges could actually be useful, as it exposes a zone which is often illuminated twice.


Prior Art

Earlier scanners can be split into two groups; transparent and reflective polygon scanners. Laser scanners can come with a single or a plurality of bundles. The first patent for a transparent polygon scanner with a single laser bundle was filed by Lindberg in 1966. Rik Starmans build the first practical transparent polygon scanner with a single laser bundle. The Netherlands Organization for Applied Scientific Research (TNO) has filed a worldwide patent application for a transparent polygon scanner with one or more bundles WO 2015/160252 A1. TNO only got a patent for more than one laser bundles through a transparent prism, i.e. a plurality which is not one, in the United States. The US most likely rejected the single laser bundle claim as TNO only developed systems with a plurality of bundles and there was already prior art by Lindberg. Jacobus Jamar was the one who first looked into the topic of transparent polygon scanning at TNO and introduced Rik Starmans to the problem. TNO build two machines with a plurality of bundles; the Lepus Next and the Argos. The first reflective polygon scanner with a single laser bundle in additive manufacturing was used by the Institute of Physical and Chemical Research (RIKEN) in 1997. In 2015, Envisiontec got a patent US 9079355 B2 for a reflective polygon scanner with a single laser bundle in additive manufacturing to protect its Scan, Spin and Selective Photocure (3SP) technology. KLEO Halbleitertechnik sells the Speedlight 2D. The Speedlight 2D is a system which uses 9 reflective polygons and 288 laser diodes to solidify a substrate with a width of 650 mm. The reflective polygon has 32 facets and rotates at a speed of 50.000 rotations per minute, see patent US8314921B2.

Business Case

The market for laser scanning technology is extremely large. Possible applications are; laser direct imaging of printed circuit boards, additive manufacturing, laser cutters, self driving cars, photocopiers and object scanners. Already in the field of 3D printing applications can range from sintering powders to solidifying polymers or egg whites. The analysis was simplified by listing exposure technologies and light sources in the 3D printing and PCB market. This should give the reader a quick overview of what is available.

Alternative Exposure Technologies

The following alternative illumination technologies can be distinguished;

  • Polygon scanner with refractive F-theta lens and one laser bundle
    • Used by: Envisiontec, Orbotech
  • Polygon scanner with reflective F-theta lens and one laser bundle
    • Used by: Next Scan Technology
    • Notes: Reflective lenses probably make the lens lighter than a glass alternative. This might simplify the fabrication of large lenses. Reflective lenses might also be beneficial at low wavelengths. Light gets absorbed by glass in deep UV.
  • Reflective polygon scanner with multiple laser bundles
    • Used by: Manz
    • Notes: the polygon tilt angle is smaller than 45 degrees, most likely costs 1 million euro's
  • Transparent polygon scanner with multiple laser bundles
    • Used by: LDI Systems
    • Notes: LDI started as a spin-off company from TNO.
  • Mask illuminated by LEDS
    • Used by: Microtec
    • Description: There are three modes of projection; contact, proximity and projection. Key is that you can make a mask with large features and project it to smaller features. The method is unbeatable in feature size and can get down to features of 7 nm. Companies like microTEC use mask technology to print structures. Operators are used to manually align masks. Companies like Idonus produce equipment to exposure these masks with UV Led lithography. A homogeniance of 3 percent can be achieve and rays are parallized up to 1.8 degrees. Sources costs in the order of 20K euro. Automated mask aligners also exist e.g. MLA150
  • XY UV laser
    • Used by: Bungard, Kloe, 4PICO
    • Description: A laser head is put on a gantry stage and the substrate is exposed in a snake like pattern. Speed up to 400 mm/s are obtained. The substrate can also be rotating. The technology is slow but can obtain very high accuracies up to 300 nm with a 405 nm laser. Kloe produces the Dilase 3D and 4Pico the laserwriter such as the PicoMaster200. These technologies seem to be used in areas such as microfluidics and to produce masks. Bungard produces a machine named the laser direct which also uses a laser mounted on an xy-stage to produce PBS.
  • Digital Micromirror Device (DMD) illuminated with LEDs
    • Used by: Ucamco and Prodways
    • Notes: Ucamco uses multiple beamers adjacent to each other. This is expensive, as a result Prodways translates the beamer and illuminates a 45 degrees mirror, see Moving Light technology. If the DMD is illuminated with laser diodes this can result into multiple-slit interference. Recently, Texas Instruments developed a DMD chip for an infrared laser; the DLP 650L NIR. Ucamco sells beamers which use three wavelengths peaks. This can be advantageous if different resists are used or the edges need a different exposure.
  • Galvanometer scanner with laser
    • Used by: 3D Systems, Materialise, Formlabs
    • Notes: due to inertia galvanometer scanners are slower than polygon scanners, 3D systems uses Nd:YAG laser with f-theta lens
  • Resonant Galvanometer Scanner
    • Notes: Can reach a line speed of 16 kHz, but the line speed is not constant, see 1 and 2.
  • Galvanometer scanner with parabolic mirror
    • Used by: Formlabs in Form 3, see Youtube
    • Notes: The galvo-mirror is used to obtain a constant line speed without an f-theta lens. The parabolic mirror is used to get the spot into focus over the full scan line.
  • Liquid Crystal Display (MSLA Technology)
    • Used by: Structo
    • Notes: Structo uses an array light source and projects through a digital mask. This technology can be scaled. It is, however, very energy inefficient see technical details. Cooling is a challenge. The technology cannot reach low wavelengths, i.e. below 400 nm. The light engine has to be in close vicinity of the reservoir. This could be an advantage for a Continuous Liquid Interface Processing (CLIP) like technology.
  • Acousto-optic deflector (AOD)
    • Used by: LPKF
    • Notes: 100 kHz position switching, no moving parts, sub-nm positioning
  • Optically Addressable Light Valve (OALV)
    • Developed by: Lawrence Livermore National Laboratory
    • Notes: The light valve is optically addressed by a projector at 470 nm which sets the transparency of the OALV. An OALV does not requires that the beam has a single mode and low divergence. The fine feature size of single mode laser beams is often limited by the thermal diffusion length. Lawrence Livermore uses two sources; laser diodes to heat up the powder and a Nd:YAG laser to initiate the process an melt the powder, see article.

The Grating Light Valve, sold by companies like Silicon Light Machines, can be used for mask-less lithography below 15 micrometers, i.e. 2.5 micrometer features and was omitted. The MEMS scanner was developed by Philips for pico projectors in mobile phones. This was unsuccesful. Philips sold the technology to InnoLuce who developed it further for mobile lidar. Innoluce was then bought by Infineon. For a description of LIDAR, see the lidar section in this article.

Alternative Light Sources

The following light sources have been considered;

  • Light-Emitting Diode (LED)
    • Used by: Ucamco and Prodways
    • Wavelengths: 405, 395, 385, 374, 365
    • Frequency: set by other element in the optical path, e.g. the refresh rate of the DMD chip
    • Power:<4 watt
    • Price: 5 euro's per LED
    • Note: LEDs offer less contrast and depth of field than laser diodes but can be combined as they do not produce coherent light. Texas Instrument seems to have a monopoly on DMD chips. Projection systems are sold by other vendors; for example, the LUXBEAM Lithography System sold by Visitech. Wintechdigital sells the PRO4500 with the following specifications; 5.5 Watts, 405 nm and 58 micrometers for 2500 euro's. DMD chips can handle less optical power at shorter wavelengths. For wavelengths below 405 nm, the power limit is currently 4 W per chip DLP9000UV.
  • Laser Diode (LD)
    • Used by: Manz, Envisiontec
    • Wavelengths: 405, 395, 375 nm
    • Frequency: 50 MHZ
    • Price: 22 euro's at 405 nm, 3870 euro's at 375 nm
    • Power: 0.4 W at 405 nm, 70 mW at 375 nm
    • Cooling: Air is sufficient, SLD3237VF can operate at 80 degrees.
  • Diode-Pumped Solid State Laser (DPSSL)
    • Used by: Orbotech
    • Wavelength: 355 nm
    • Frequency: 80 MHZ
    • Power: 24 W
    • Price: 190k euro's
    • Vendor: Coherent
    • Sizes: LASER 305 x 200 x 1100 mm, power supply 482 x 177 x 505 mm,
  • Nd:YAG LASER
    • Used by: 3D systems and Materialise
    • Wavelength: 355 nm
    • Power: 1 W
    • Frequency: < 1MHZ

The femtosecond laser, which can be used in two-photon polymerization to focus light in space and time and trigger a non-linear reaction, was thought to be too expensive for large-area photo-polymerization.

Physics

Please note that all equations on Reprap.org currently do not work which results in <math> statements. They can be rendered with LaTeX.
In the following, an analytical description of the system is given. The section starts with a parameter definition. Hereafter, the following properties are discussed; polygon, spot ,transparent parallel plate and polygon tilt angle. All the equations are also available in a Python script. This script can be used to quickly obtain the properties of the system. The calculations are verified with a numerical simulation.

Parameter Definition

The polygon rotates about its center, i.e. the point inside the polygon that is equidistant from each vertex. The substrate moves under the polygon in a certain direction. The smallest angle between the illumination direction and the substrate movement direction is defined as the polygonal tilt angle.

  • <math>\alpha</math> denotes the static polygonal tilt angle. This is 90 degrees in the Hexastorm.
  • <math>I</math> is the angle of incidence of the optical beam on the transparent polygon
  • <math>I_{max}</math> is the maximum angle of incidence used during illumination
  • <math>f_{efl}</math> is the effective focal length of the lens used to focus the bundle
  • T defines the thickness of the polygon, T is equal to 2r.
  • r defines the inradius of the polygon
  • a defines the polygon side length
  • R defines the circumradius of the polygon
  • v is the number of vertices of the polygon.
  • n is the refractive index of the polygon, quartz is used with a refractive index of 1.47
  • d is the diameter of the aspherical lens
  • <math>\lambda</math> defines the center wavelength of the laser diode beam

Polygon

Figure 6.

In figure 6, a regular convex polygon is shown with the following parameters; r is the inradius, R is the polygon circumradius and a is the polygon side length. In figure 6, the number of vertices, v, is equal to 8. Earlier, we defined 2r to be equal to T. The number of facets of the polygon has to be even for opposing planes to be parallel. If the number of facets is uneven, there will be an edge crossing during illumination which makes the polygon unsuited for scanning.

  • <math>a=T\cdot tan(\pi/v)</math>
  • <math>R=\dfrac{a}{2 \cdot sin(\pi/v)}</math>
  • The interior angle of a simple polygon with v vertices is <math>180-360/v</math> degrees.

For an octagon, <math>I_{max}=90-(180-\dfrac{360}{v}) \cdot 0.5</math> which is 22.5 degrees.

Spot

Light emerges from a small optical window from the laser diode and as a result diverges. There are several ways to focus the diverging beam. The choice is a trade-off between spot quality and cost. A laser can be focused directly with an aspherical lens. This is cheap but will give an elliptical spot. The laser cannot be tightly focused as aspheric lenses typically have a small diameter. A better spot can be obtained by first collimating the laser diode with an aspherical lens and circulating it hereafter with an anamorphic prism pair. The bundle can then be focused with an achromatic doublet. This is a more expensive solution but can give a smaller and circular spot.

Lens Alignment

The aspheric lens position accuracy is determined by the optical magnification of the whole system. The emission point typically has a size of 0.5 micrometers by 1 micrometers. The size can be measured via the Fraunhofer diffraction pattern. The emission point accuracy is assumed to be +/- 80 micrometers. This was estimated from similar laser diodes. For a 50 micrometers spot, the aspheric lens has to be placed at an accuracy of 3 micrometers. The magnification is 50. The aspheric lens can be purchased mounted with a M9 thread and screwed into position with a thumb screw mounted on the lens.

Rayleigh length

The spot is defined to be in focus in twice the Rayleigh length; <math>z_r=\dfrac{2 \pi w_0^2}{M^2 \lambda}</math>. <math>M^2</math> is called the beam quality factor. Most collimated single TE mode laser diode beams have <math>M^2</math> of 1.1 to 1.2, source Sun, Haiyin.

Spot size

The spot size of a collimated and circulated bundle focused by an achromatic doublet is <math>w_0=\dfrac{4\lambda}{\pi}\dfrac{f_{efl}}{D}</math>, where D is the diameter of the collimated bundle. The spot distance is <math>f_{efl}</math> from the achromatic doublet.

The spot distance of a laser diode directly projected by an aspheric lens is given by the thin lens equation; <math>\dfrac{1}{f_{efl}}=\dfrac{1}{s_1}+\dfrac{1}{s_2}</math>. <math>S_1</math> denotes the distance between the laser diode and the aspheric lens. <math>S_2</math> denotes the separation between the aspheric lens and the spot. The magnification, M, is given by <math>M=-\dfrac{s_2}{s_1}=\dfrac{f}{f-s_1}</math>. As the emission point is not square, the spot will be elliptical.

The smallest spot which can be formed is given by the Airy disk; <math>w_0 \approx 1.22\dfrac{s_2 \lambda}{D}</math>. Here D is the width of the collimated beam at the lens.

Transparent Parallel Plate

The properties of a parallel plate are described by Smith and Wyant. The most important properties of a parallel plate are summarized in this section. Typically, a parallel plate is used to transversely shift a collimated bundle. It can be concluded from Snell's law that the refracted bundles are parallel with the incident bundles. For a converging beam, a parallel plate also gives a longitudinal focus point displacement away from the source and optical aberrations. The optical aberrations increase if I is increased. Via an analytical calculation it is ensured that the Strehl ratio is above the Rayleigh limit at <math>I_{max}</math>. This effect starts to play a role for spots with a diameter below 30 micrometers.

Displacement

  • longitudinal displacement <math>=\dfrac{n-1}{n}T</math>
  • transversal displacement <math>=T sin I(1-\sqrt{\dfrac{1-sin^2 I}{n^2-sin^2 I}})</math> denoted as <math> \tau </math>.

The spot speed can be derived by differentiating the transversal displacement with respect to time. Let's assume the angular speed is constant, i.e. <math> \frac{\partial I}{\partial t}=c </math>. The chain rule can then be used to determine the spot speed <math> \frac{\partial \tau}{\partial t}(t)=\frac{\partial \tau}{\partial I}(I(t))\frac{\partial I}{\partial t}(t)</math>. The speed at the center is smaller than the speed at the edges of a scan line. As a result, the amplitude at the center should be smaller and is ideally corrected for by the laser diode driver by varying the pulse frequency or current.

Strehl ratio

The optical performance of the system can be evaluated via the Strehl ratio. If the Strehl ratio gets below a threshold, the aberrations will become dominant and the system will not image properly. As a result, it must be ensured via calculation that the Strehl ratio is larger than some acceptable limit, e.g. the Rayleigh limit of 0.71. Literature provides us with the Seidel coefficients of the main aberrations. These are used to determine the Strehl ratio. In the following, a quick overview is given. The f-number or <math>f_\#</math> equals <math>\dfrac{f_{efl}}{D}</math>. A transparent plate does not have a Petzval field curvature aberration. As a result, the projection of the transparent polygon is telecentric. The third order Seidel aberrations are listed below using Wyant. To simplify checking, the equations are listed with the page and equation number from James Wyant's book titled Basic Wavefront Aberration Theory for Optical Metrology.

  • spherical wavefront aberration <math>=-\dfrac{T}{f_\#^4}\dfrac{n^2-1}{128n^3}</math>, page 42, equation 72
  • coma <math>=-\dfrac{TI}{f_\#^3}\dfrac{n^2-1}{16n^3}cos\theta</math>, page 44, equation 75
  • astigmatism <math>=-\dfrac{TI}{f_\#^2}\dfrac{n^2-1}{8n^2}cos^2\theta</math>, page 45, equation 77

The coefficients can be used to calculate the optical path difference W defined on page 17.

<math> W(x_0, \rho, \theta) = \sum_{j,m.n}^{}W_{klm}x_0^k \rho l cos^m \theta</math> with <math> k=2j+m </math> and <math>l=2n+m</math>

The coefficients <math>W_{klm}</math> can be determined from the Seidel aberrations and table 2 on page 17. X is <math> \rho cos \theta </math>, for more details see Wyant.

The optical path difference can be used to calculate the wavefront error <math> \sigma </math>, see equation 62 on page 37.

<math>\sigma^2=\dfrac{1}{\pi}\int_{0}^{2\pi}\int_{0}^{1}\{\Delta W(\rho,\theta)-\Delta \overline{W}\}^2 \rho d\rho d\theta</math>

The wavefront error is converted to lambda wavefront error via division;

<math>\sigma_{\lambda}=\dfrac{\sigma}{\lambda}</math>

Finally, the Strehl ratio can be calculated with the first three terms of its Taylor series, see Wyant page 39 equation 67.

<math>\text{strehl ratio} \approx 1- (2 \pi \sigma_{\lambda})^2 + (2 \pi \sigma_{\lambda})^4/(2!) </math>

As can be seen from the equations, the Strehl ratio is minimal for <math> I </math> is <math> I_{max} </math>.

Polygon tilt angle

If a polygon is made thicker, multiple laser diodes can hit the same polygon. As can be seen in figure 7, the polygon axis must be tilted as otherwise all the laser diode projections will completely overlap each other. The polygon tilt angle can be 45 degrees if the scan length per laser is longer than the separation between two lasers. Kleo is not able to do this as the image gets distorted at the edge of a lens and the bundle exits the lens at 90 degrees. In Kleo's system, the polygon tilt angle is less than 45 degrees. This can be proved as follows. The scan length per laser is denoted as <math>S_L</math>. The scan length per laser orthogonal to the substrate movement is written as <math>y_{length}</math>.

<math>y_{length}=sin(\alpha) \cdot S_L</math>

The distance S between two subsequent lenses, i and i+1, orthogonal to the substrate movement equals

<math>y_{i+1,i}=sin(90-\alpha) \cdot S_{i+1,i}</math>

Overlap between diodes requires;

<math>y_{length} - y_{i+1,i} \geq 0 </math>

As a result, for 45 degrees we have <math>S_L \geq S_{i+1,i}</math>. This shows that Kleo cannot use 45 degrees. In the above analysis the polygon tilt angle is set equal to the tilt angle of the laser diode lanes. This is not true, in reality the angle will distort slightly due to movement. This effect is very small and has been neglected.

Figure 7.

Simulation

The equations were verified with an open optical ray tracing and lens design framework named Rayopt. Using this framework, it can be verified that the system is telecentric and the system is in focus over a plane, i.e. it has a flat field projection. The script has been made available by Hexastorm here. A picture of the output is shown in figure 8.

Figure 8.

Experiments

A CCD camera was placed under the Hexastorm. A neutral density filter was placed on top of this CCD camera. The laser was focused onto the camera chip. A single picture of a static and three pictures of a dynamic exposure are shown below. A script has been provided by Hexastorm which can be used to determine the spot size from these pictures using OpenCV. A pixel measures 4.65 by 4.65 micrometers in these pictures. The pictures are stored as bitmap to ensure analysis is still possible.

Static exposure

The static exposure of a single spot is shown in figure 9. The movement of the spot is not shown. This movement is very small but noticeable. It does not effect the resolution of the current Hexastorm.

Figure 9.

Dynamic exposure

In a dynamic exposure, the cross scan error and jitter can be determined. The jitter is the scan error in line of the scanning direction. The cross scan error is the scan error orthogonal to the laser scanning direction. In the current Hexastorm it is not possible to determine the facet. If this was possible, one could correct for a scanline shift per facet. As a result, it is important that subsequent laser lines overlay each other if the laser head is not moved. The pictures were taken at 67 revolutions per second. Two features are shown. A feature made with 2 and a feature made with 4 laser pulses at a laser frequency of 100 kHz. Early versions of the Hexastorm had a large cross scan error due to improper mounting of the polygon. This was later improved. Currently, the cross scan error is in the order of 40 micrometers. An early exposure with a large scan error is shown in figure 10. The improved version with a low cross scan error is shown in figure 11.

Figure 10.
Figure 11.

Finally, it is also possible to project only at a single facet. This improves the quality but reduces the scan line frequency to one per revolution. A picture of a single facet illumination with a hexagonal prism is shown in figure 12.

Figure 12.

The Hexastorm is very stable in the scanning direction. The polygon must be very well made to achieve overlay of multiple facets. It might be beneficial to make both the polygon axis and the polygon prism out of quartz. Similar to a reflective polygon, the prism can then be clamped instead of glued on top of the polygon base.

Polygon cross-scan error reduction

One of the characteristics of a polygon scanner is cross-scan error which is a deviation perpendicular to the scan line. This is also referred to as wobble or dynamic track error. There is non-repeatable wobble error from the motor bearings. There is also repeatable error from polygon facet to datum error. There is no such thing as a perfect polygon scanner. There is always some wobble error. This can be fixed via optics, active alignment and software.

Optics

  • collimate laser bundle with aspherical lens
  • focus laser into one direction with cylindrical lens
  • refract through prism
  • focus laser into other direction with cylindrical lens

This will not only reduce the cross-scan error but will also allow one to make the bundle more circular. Note, this is different than the technique used in reflective polygon scanners where the collimated bundle is compressed to a line and after reflection is expanded to an ellipse, see patent. The sideview of this embodiment is sketched in figure 121. A top view is sketched in figure 122. The photodiode has been moved and now measured the reflected bundled and not the transmitted bundle. The idea is that this created more room for the lenses.

Figure 121.
Figure 122.

Active alignment

Next Scan Technologies uses a reflective f-theta lens to produce a line. This lens is large and has fabrication errors. The lines are therefore not perfectly straight. To compensate for this error a galvo-scanner is used for active alignment. The Hexastorm could achieve active alignment via an Acoustic Optical Deflector (AOD). The optical path would be; laser, collimation lens, first cylinder lens, prism, AOD and second cylinder lens. The AOD would compensate for fabrication errors in the prism; like non-planarity. It would ensure that all the rays of the different facets are parallel. As such they will be focused in the same focal point by the second cylinder lens. Instead of an AOD a reflective galvanometer scanner could be used for active alignment.

Software

  • only use one facet
  • correct on facet basis with correction table

This has several disadvantages;

  • cannot reduce non-repeatable motor bearing errors
  • speed is clipped, in case of one facet

Additional Claims

In this section, we want to quickly list some possible applications of transparent polygon scanners to generate prior art. This list has been created to limit patent claims and extend the freedom of operation of Hexastorm in for example the US market. Envisiontec was able to patent a reflective polygon scanner although there was already prior art by the Institute of Physical and Chemical Research (RIKEN). To prevent this, I have created an extensive list of possible applications. Creating patents would have been very expensive. Established companies are also able to sue patent holders to clarify their operation freedom. This can create very tricky situations for bootstrapped startups. These ideas, however, have not been widely tested. It has been quickly drafted to generate prior art.

LIDAR

LIDAR is a surveying method that measures the distance to a target by illuminating that target with pulsed laser light and measuring the reflected pulses with a sensor. In the following two technologies are considered; MEMS and reflective polygon scanners. In both systems a laser bundle collimated in one direction and focussed in the other direction is incident on a mirror. As a result, a narrow line is produced in the far-field. This line diverges. A camera measures the time of flight and position of each laser spot. Reflective polygon scanners have been used in LIDAR systems for over 30 years. The large aperture, wide scan angle (up to 120 degrees for reflective), linear scan speed and high scan rate of polygon scanners has provided long range and high resolution for surverying (Faro), terrain mapping (Riegel) and mobile applications (Valeo Scala in the Audi A8). A challenge of a polygon scanner is its resonance frequecency; ideally this is above 2KHz. Due to the larger size than MEMS, polygon scanners have lower resonance frequencies. Mirada Technologies claims to have solved the resonance frequency challenge by immersing the polygon in a liquid and move it with a magnetic field. MEMS manufactures claim to have a lower price point and be more reliable than polygons. Infineon is active in mobile applications. MEMS have a smaller scan-angle. A large mirror requires a larger cavity which is hard to fabricate. A transparent polygon scanner could also be useful in LIDAR. The operation would be very similar to reflective polygon scanners. The shorter scanline might be beneficial for some applications and could be extended with an optical transformer such as a mirror or lens.

Foil

A transfer substrate simplifies the application of a layer. Coating can be made easier with a blade which is more suitable for viscous, i.e. filled, resins. Lithoz uses a rotating disk in its LCM technology, see video. Admatec uses a foil see patent. In the past, TNO tried to patent a foil coater see EP20090164821. This most likely failed due to a patent from Charles Hull, US5637169. In any case, we do want to list a explicit example of how a foil could be used in combination with a transparent polygon scanner. The foil would be made in contact with the part during the illumination. The two images show how the foil should be applied in up projection or down projection.

Figure 14.

The rest would be very similar and easy to implement for a skilled observer, see TNO's description or Hull's description. The coating layer applied on the foil might be applied very precisely so areas are not coated twice or to block interaction between the foil and an already coated section; see figure 4. It is also straight forward to outline how this could be used in the case of multiple laser diodes. The scan head is still moved over a part. The scan head is moving relative to the part. The foil is moving however and as result the the foil and part are static with respect to each other. The foil may be made from Teflon or Teflon AF. It might also be beneficial to add a glass plate between the foil and the laser to create a flat reference substrate, see figure 15.

Figure 15.

Algorithms

The polygon facet can be detected by giving a facet a marking and detecting this marking while the polygon is rotating. This marking could be created by coating the edge of a single facet and measuring its reflection with a second photodiode. If the facet number is known, one can correct for scan errors which are facet dependent.

Laser bundle position

Not for all lasers it might be possible to measure the laser bundle directly with a photodiode. In these cases, it might be beneficial to add a second lower power laser bundle on the same facet and in the same direction or on a different facet (for instance orthogonal to the high power beam). That position could then be monitored with a photodiode.

Immersion lithography

The smallest feature you can pattern with a light source is dependent on the wavelength. Typically the smallest feature is in the order of half the wavelength. The wavelength is dependent upon the medium the light travels trough. Companies like ASML temporarily coat the substrate with a liquid to lower the wavelength and increase the resolution, also known as immersion lithography. As such, it might be beneficial to use a similar technology in the case of transparent polygon scanner. For example a liquid could be applied on the foil shown before. This liquid would not be used to get solidified but would only be there temporarily and used to increase the resolution. Another option would be to remove the whole foil and move a transparent plate over a substrate which is coated with a liquid. A transparent polygon scanner would then illuminate through this transparent plate.

Figure 16.

Bioprinting

The Aether1 is an example of a bioprinter. The Aether uses UV light to solidify liquids during printing, see video. Currently, the UV light not only exposes photo-polymers but also the cells. The UV light damages the cells. We claim that the Hexastorm is used as an UV exposure source in a bioprinter. The Hexastorm could ensure that only the photo-polymers are exposed and not the stem cells. The printer would deposit liquid or cells with syringes and selectively expose them with the Hexastorm.

Laser sources

As alternative to a laser diode, a quantum dot laser, C02 laser, femto-second laser or fiber laser can be used. The usage of a different laser source might be beneficial for laser cutting, printing or sintering of metal, plastic or any other sort of powders. The European Union is developing a 500 W multi-MHz laser which would enable the Hexastorm to sinter metal powders.

OCT

Optical Coherence Tomography (OCT) is an imaging technique that uses low-coherence light to measure samples based upon the principle of light interference. It is used in the medical industry to detect cancer in tissues and diseases in eyes, e.g. the Cylite of Hewlett Packard. OCT is typically used to obtain information from a sample. In 3D printing it has been used to verify a print. For example, Photoncontrol used Optical Coherence Tomography and Raman spectroscopy to test the quality of bioprinted tissue, see 1. A startup, called Inkbit (MIT), is using it to create samples accurately. They print droplets with an inkjet head and then verify the position of these droplets using among others OCT.
OCT has also been used to detect the adhesion between layers in a 3D printing process, see Non-destructive testing of layer-to-layer fusion of a 3D print using ultrahigh resolution optical coherence tomography.
Due to the interest in this area, I decided to elaborate upon how OCT can be used in combination with a transparent polygon scanner. I claim that in figure 2 a transparent polygon scanner is used instead of a galvo scanner. I claim that in the optical path of the Hexastorm a beam splitter is placed after the aspherical lens and before the first cylindrical lens to enable the scanner for optical coherence tomography. I claim the use of a transparent polygon scanner for wavefront measurement in Ophthalmology and Optometry. The vertical measure of an eye ball, generally less than the horizontal, is about 24 mm. The current scanhead has a scanline of maximum 24 mm, making it already close to dimensions required for eye ball measurement. I claim that possibly two transparent polygon scanners are used in ophthalmology and optometry, to move the bundle in two directions.
An OCT enabled transparent polygon scanner might also be useful for 3D printing. Imagine that a small percentage of the bundle is scattered to the reference mirror and most of it used to go to the sample. It will then be possible to sinter powders or polymerize liquids at the sample location. A small portion of the beam will be reflected and refract back to the beam splitter and interfere with the reference beam at the photodetector.
This allows one to measure the photopolymerization or sintering process during printing. I can imagine this is especially usefull if the layer height is less than the wavelength. I can also imagine that this is usefull during a process akin to Continous Liquid Interface Processing (CLIP). I would, however, emphasize that the process I envision is legally different. See the section on Carbon 3D; among others I use a transsparent polygon scanner instead of Micro-mirror Device (DMD). I label this process as Hexaforming as CLIP is a trademark and the process I describe is different.

Carbon 3D

Carbon has a technique which it denotes as Continuous Liquid Interface Processing (CLIP). Carbon uses a Digital Micro-mirror Device (DMD) to illuminate a photopolymer through an oxygen-permeable window made of a fluorpolymer such as Teflon AF. Teflon AF can be sourced from Biogeneral. NASA described how a teflon AF sheet can be made. A supplier fo chemicals can be found here.
The permeation of oxygen through the window creates a persistent liquid interface, nicknamed "dead zone" where photopolymerization is inhibited between the window and the polymerizing part. Oxygen inhibition was an effect that was already shown to play a role by Denkari et al. in 2006 for silicon release coating invented by John Hendrik, see US7052263 (B2). The "dead zone" in silicone release coating is so small that a peeling is needed to release the part from the transparent window. Hessel Maaldrink sped up the process by adding a force feedback sensor EP2043845B1. Using Teflon AF and a polyuerethane Tumbleston et al. 2015 where able to extend the "dead zone" to approximately 30 micrometers. As such, the part does not have to be peeled of from the optical window during the process and stair stepping is minimized. This allows for the production of flexible parts. Furthermore, the "dead zone" speeds up the viscous flow between two parallel plates, part and window, for the application of a new layer, see WO2014126837A2, improving the print speed.
I will now try to create prior art to support circumcenvtions of the Carbon patent to extend the freedom of application of the transparent polygon scanner marketed as Hexastorm.
After studying the WO application of CLIP, I noticed that the European patent is different from the US patent. In the European patent, EP2956823B1, claim one states ".. irradiating said build region through said optically transparent member to form a solid polymer from said polymerizable liquid while also concurrently advancing carrier away ..".
As such, I claim irradiating said build region with for example a transparent polygon scanner while not concurrently advancing away the part, but discretely. The part is exposed and moves after full exposure. Moving during exposure is also not possible as Hexastorm exposes a line and not a plane.
DMDs have a pattern/pixel rate of up to 20 kHz. Laser diodes can achieve a refresh rate in the order of 50 MHz. At 50.000 RPM and six sides, a transparent polygon scanner exposes at line rates of 5000 Hz. With a laser diode, the refresh rate is so much higher that it might be possible to alter the polymerization over much smaller distances. Stair stepping would be minimized even though the part is moved disretely.
If it is not possible, the procedure would still allow for the production of flexible parts.
The US patent, US 9216546B2, is wider in scope as claim one specifies "A method of forming a three-dimensional object, comprising the steps ...". The formulation using "steps" in US patent differs from "concurrently" in the European patent.
In the US the process is also under patent if the part is not moved during exposure. Carbon as a result markets its intellectual property as "digital light synthesis technology", although Continuous Liquid Interface Processing (CLIP) seems more applicable in the European union.
Key in the US patent is that parts are produced upside down and moved away from a build surface which is not air. This is peculiar as the original patent by Hull in 1986 specifies both up and down projection in figure 3 and 4 respectively.
As such, I claim the use of oxygen-inhibition down projection, where the top of a part is up, using a transparent polygon scanner. Again, the fast exposure of a laser diode might minimize stair stepping and the "dead zone" will simplify coating. Additionally, using air instead of teflon AF layer reduces costs.
After studying the following literature, dip and blade coating patent US5651934 of Charles Hull, flows in thin film coating by Christian Kushel, Zerphyr coating as described in US6159311, curtain coating as described in EP0928242 and the book Liquid Film coating by Kistler, I claim the following.
1. Firstly, the use of a "dead zone" in 3D printing to facilitate the coating of liquids in down projection photo-polymerization where the top of the part is up. I will now explain this.
Boundary conditions during coating are important. A part solidified up to air provides a non-consistent liquid // solid boundary condition. During for example blade coating the coater can collide with the part. A dead-zone would prevent this and create a more consisting wetting of the substrate during e.g. Zephyr coating as it is entirely liquid.
2. Secondly, I claim that an array of transparent polygon scanners is integrated in the Zephyr blade. I claim that possibly in the Zephyr blade a Teflon AF film is partly applied between the liquid and air interface. I claim that in the Zephyr blade the pressure of the air is monitored. I claim that an opening is provided to actively supply liquid to the Zephyr blade using a pumping mechanism.
3. Thirdly, I claim that the teflon AF film or part moves parallel to the plane of illumination and not only orthogonal as in the CLIP patent. Transfer substrates are used by Admatec, Carima and described by TNO in EP2272653. Especially, I look at the figure provided at the front page of US2012007287. I claim that the film in this figure denoted by 10 is teflon AF and the exposure module denoted by 9 is an array of transparent polygon scanners.

Swarm printing

I claim the use of a transparent polygon scanner in so called swarm printing. Liquid can be applied via various ways such as extrusion.

2D Imager

Via a transparent polygon source you can create a line from a laser spot, i.e. you can map a point (0 dimensions) onto a line (1 dimensions). Naturally, with a second transparent polygon with has its rotational axis orthogonal to the first you can map the line to a plane (2 dimensions). The second rotating prism would be thicker and typically rotate at a lower speed. This would partially mitigate the disadvantage, that it is much thicker. As a result, you form a 2D image plane. This image plane can then be translated with an optical transformer, such as a single lens

Bird control

Companies like Bird-X use lasers to scare birds at public air ports. Birds can get accustomed to the pattern and the pattern might be dependent on the type of bird. As such, it might be beneficial to alter the pattern of the laser with a transparent polygon scanner for bird prevention.

Satellite Communication

Companies like Inmarsat use laser bundles for optical communication with satellites, see video. The transparent polygon scanner can be used to transform a 0D laser spot into a 1D line or 2D plane. This plane or line can then be transformed with lenses and used to communicate with other satellites. For a DIY example with a microphone and a speaker see video.

Food printing

To get microwaves meals pass the FDA, a prototype meal with egg whites is typically made. These prototype meals consist out of salt and liquid egg whites. After microwave exposure, the FDA slices the prototype meal and checks whether the exposure is good enough. These microwave meals are made by hand. The Hexastorm could expose the liquid egg white with infrared radiation and trigger a maillaird reaction. If the process is combined with inkjet or needles more complicated egg white meals can be made. A layer of liquid egg white is coated an solidified with the Hexastorm. A 3D omelet can then be made by stacking multiple layers.

Tube measurement

Companies like LAP laser use reflective polygon scanners to measure tube diameters. Industrial giants like Vallourec use a combination of two of these scanners to inspect high collapse tubes. In a similar fashion, a transparent polygon scanner could be used to measure the diameters of tubes with a diameter smaller than the line length. Two transparent polygon scanners could be used to measure the diameter of larger tubes if they expose only the edges of these tubes. It must then of course be known that the tube diameter fluctuates less than the line length. So a tube with a diameter of 1000 mm can be measured if the diameter fluctuates less than e.g 10 mm. Keyence sells a module with 3-cmos chips and without a polygon scanner for details see the folder of the LS9000.

UV printing

In the printing industry UV LEDS are combined with UV curable inkjet heads. This process is also known as UV led curing. I claim the use in a commercial machine of one or more transparent polygon scanner with a UV inkjet head. The transparent polygon scanner could be used to vary the UV dosage. A UV LED can only provide an uniform UV dosage. It could also be that materials are applied to the substrate which are very sensitive to UV light. In this case a transparent polygon scanner could prevent that these materials are illuminated. It could also be that different materials require light of different wavelengths. In that case multiple transparent polygon scanners with possibly lasers of different wavelength could be used to apply the right wavelength and power to the right material. This claim applies to both 2D and 3D printing processes.

Particle Analyzers

Lasers scanners can be used in particle analyzers . A bundle is projected on a cell and the resulting image is measured with an image sensor. In specific, I claim that a transparent polygon scanner is used to move the laser bundle in such a particle counter or analyzer. Furthermore, I claim that a structured laser bundle is used which has been developed by CERN. Finally, I claim a transparent polygon scanner is used in a so called flow cytometry.

Rail Road Inspection

To inspect defects in rail roads laser scanners are used. The Hexastorm produces a shorter line length than most scanners. This could be mitigated by inspecting smaller things; e.g. small defects in airplane wings, smaller tubes / cables or surfaces. Like in rail road scanners, I envision that multiple scanners are used to illumate objects form multiple sides. The distrubed line would then be imaged by a camera. The image is then be used to calculate the shape of the substrate. I also envision a scanning system which consists out multiple scannes. An object shape is measured by a first scanner and a camera, the object is then marked by the second scanner. Furthermore, I claim the laser scanner first sends out a low-intensity line. This line would be used to record information of the substrate. The second line produced by the same scanner at higher power would be used to alter it. I can envision that this is beneficial for among others laser engraving and 3D printing.

Marking of medicine

Instead of the Videojet 7810 2-Watt UV laser marking system, the hexastorm with a UV laser is used to deliver high-contrast cold marking permanent codes enabling product lifetime track and trace for pharmaceutical, medical device and cosmetic manufacturers. Possibly a camera is added to detect the products to be marked.

Inkless printing

The Yes!Delft startup Tocano developed Inkless printing, see video Their idea was to print paper without ink and engrave it with laser light using reflective polygon scanners. The disadvantage of reflective polygon scanners is the long scan lines; it limits throughput. This is not a problem with toner printing as not much power is needed, but can be a problem with engraving. I claim that with a transparent polygon scanner, as it has shorter scan lines, a higher throughput can be created. Multiple Hexastorms would be used in tandem to illuminate a substrate such as a piece of paper. A possible supplier of laser diodes is Lumics. The diodes have a wavelength of 1 micrometer and a power between 3 and 7 W.

Laser Microscopy

In microscopy laser scanners are used to illuminate the substrate. The reflected light, which can be of a different wavelength, can then be imaged by a camera. I claim that a transparent polygon scanner is used to move a laser bundle over a sample in a microscopy. Possibly, a semi-transparent mirror is added to analyze reflected light. This mirror can be added between the sample and the transparent prism possibly after the last cylinder lens. A DIY version of a laser scanning microscope can be found at instructables.

Dual side PCB exposure with the use of transparent plate

Industrial giants like Manz, Orbotech etc. produce PCBs on a side by side base. First the top side is exposed and then the bottom side. I claim a machine which illuminates a PCB from two sides at once. To achieve this an array of transparent polygon scanners is used to expose the PCB from the top. The PCB is positioned on a transparent plate. As a result, the PCB can also be exposed from the bottom side. Possibly, the transparent plate is coated with Teflon or lubricated with a substance such as oil. This is to prevent damaging of the transparent plate.

Laser Induced Forward Transfer (LIFT)

Transparent polygon scanners might be used for laser induced forward transfer in bioprinting or deposition of viscous fluids. Poietis calls this laser-assisted bioprinting. I claim a laser induced forward transfer where a transparent polygon is used to move the laser bundle. An application might be bioprinting as demonstrated by Poietis.

Dual Photon

Nanoscribe uses a femtosecond laser to solidify resin via dual photon polymerization with a galvo scanner, see link. The advantage of the Hexastorm light module is its telecentric projection. This allows a machine to stitch lanes accurately without a telecenric lens. As such, it could be useful for a dual-photon process. I claim the combination of a femtosecond laser or the laser used by nanoscribe prior to 4 april 2019 with a transparent polygon scanner in dual photon polymerization. Dual photon polymerization can be used to create objects via 3D printing.

STED Lithography

STimulated Emission Depletion (STED) lithography is a way to expose 3 dimensional structures in resist, without a mask, with dimensions much smaller than the optical diffraction limit. This is achieved by a first ‘exposure beam’, with a diffraction limited Gaussian spot, followed by a second ‘depletion beam’. This second beam has a donut like intensity distribution and cancels the effect of the ‘exposure beam’, resulting in an effective exposure volume, much smaller than the Gaussian intensity distribution. STED lithography can get a lower resolution limit than 2-photon lithography. In literature, resolutions down to 1/5 of the diffraction limit have been reported. I claim a setup where 2 lasers, for example a continuous wave laser at 532 nm and femto-second Ti:Sapphire laser are coupled by for example two diochroic mirrors into one bundle. This bundle is then scanned via transparent polygon scanner. An overview of this technology where a galvo is used to move the bundle can be seen at link.

Flexographic sleeves

The Hexastorm can be used to print flexographic sleeves directly. The current process is indirect.

Detector bar to allign scanheads

If multiple scan heads are used, a challenge is to align these heads. The position of the laser in each head has to be known exactly. One way of doing this is by moving a camera under the scan heads and collecting position information per laser. The laser would project directly onto the CCD / CMOS chip and its position would be determined. This is however expensive as it requires an extra stage with camera. The chip has a finite size of for example 5x5 mm and has to be moved. Another solution of doing this would be to add a bar from diffuse glass, e.g. opal glass. The light would be scattered in this bar and reach the edges of it. At the edge of this bar there would be a photodiode. One might think that this bar must be narrow so the position can be detected up to 10 micrometers accurate in one direction. The current photodiode used to calibrate the laser is, however, also not narrow. You can simply use the rising edge of the signal recorded by the photo-diode used to monitor the diffuse opal bar. The stage upon which the scan head is mounted then moves in orthogonal direction to this bar. By turning on the laser and moving it over the bar. The position can be determined exactly in that direction. It might be needed to add a cap-around the bar to minimize stray light. This is also done for the photo-diode in the scanhead. Still, I need two dimensional information. I also need to know the position of the laser diode at the bar. To do this I could use multiple photodiodes along edges of the bar. These would all measure a signal if the laser hits the bar. The signals will however arrive at different points in time. This allows one to determine the position of the laser along the bar.

Transparent Galvanometer Scanner

I claim a galvanometer scanner which is made by a transparent prism instead of a reflective mirror. An advantage of a transparent galvanometer scanner is that it allows one to vary the spot speed along the scan line. The argument/applications would be largely analogous and follow from the above article.