Project Logs

Collapse logs

• Alginment procedure

  Hexastorm • 01/23/2021 at 14:16 • 0 comments

  Made a video of the alignment procedure with new updated tool chain. I moved away from the PRU of the Beaglebone to the FPGA and replaced the ueye camera with an arducam. This sounds easy but still took me a year :-).

• DNA Laser printing & COVID

  Hexastorm • 01/19/2021 at 16:49 • 0 comments

  Recently, I was reading upon the character-by-character reconstruction of the BioNTech/Pfizer SARS-CoV-2 Vaccine .
  Most of us, will end up with this in their bloodstream so I thought it was rather interesting. It turns out there was a startup which produced DNA using laser printing . Austen Heinz gives a good presentation of this process below, scroll to time 10.24 min
  

  The startup Cambrian Genomics was founded by Mr. Heinz in 2014. The startup no longer exists and seems to have faced many difficulties. They did manage to raise over 10 million USD. Austen Heinz committed suicide at the age of 31 on May 24, 2015.
  His patents are still around, see US 10,822,605 B2 which got awarded in November 2020.
  As there are still people funding the patents, there must be people working on it. I looked into the patents for details on the required accuracies. You need optics which provide an accuracy of around 1 micron.  This accuracy could be possible with prisms, if you have a reasonable budget. Process uses laser induced forward transfer or laser catapulting.
  
  

  Quote from patent;
  

  Methods of using emulsion-PCR derived beads by the present assignee have been based on the fact that the bead itself serves as a decent ablation substrate for the approx 1 microJoule nanosecond laser pulse used to induce transfer.
  Beads from 1 to 10 micrometers are easily and specifically ejected by focusing light into their plane.
  Although ejection optics disclosed by the present assignee can target features as small as 600 nm, it is desirable to generate slightly larger, more disperse 1 micron clusters for use with initial prototyping.

• Freecad Assembly Design

  Hexastorm • 01/06/2021 at 15:51 • 0 comments

  I have made an assembly design of the laser scanner in Freecad using the new assembly 4 workbench.

  I am really happy with this workbench. It really helps with the open-hardware concept of the project; without these tools it would not be possible to give other people the ease of use to adapt and add their modifications.  Companies like Prusa Research are open source but not open-hardware as they don't provide access to the assembly drawings. They might do that in the future with these new tools.
  

• volumetric 3D printing

  Hexastorm • 01/06/2021 at 14:19 • 0 comments

  As described on Hackaday and Sci-hub, a new volumetric printing method has been described which uses two colors.
  A light sheet with a 375 nm laser diode and DLP projection between 450-700 nm.
  

  The prism scanner can also be used to accomplish this. Basically, you can intersect the projection of the prism scanner with a laser sheet or line.  The prism scanner could operate at 550 nm and the light sheet at 375 nm. It would allow you to solidify resin at the intersection and not at the top of the resin bath.

  Note that if you project a resin bath from above (say from air). The top surface of this resin bath is NOT even. In fact it is wobbly, due to all sorts of surface effects. This would be circumvented by this new technology, as the top of the resin bath is not solidified.  An alternative for this would be to use a process akin to Continuous Liquid Interface Processing (CLIP) (here you also don't solidify the top due to oxygen interaction).
  Back to Xolo, they might infringe patent US10843410B2, there seems to be no patent in Europe... The patents claims something very similar to what Xolo is doing.
  

• Patent portfolio of Luminar Technologies

  Hexastorm • 12/05/2020 at 12:49 • 0 comments

  Austin Russell is the world's youngest self made billionaire at age 25. Mr. Russell founded Luminar Technologies, trades a Nasdaq under LAZR, a company which builds Lidar scanners for cars.
  Luminar's lidar scanners use a combination of a rotating polygon mirror with a galvo mirror. The laser source uses a wavelength of 1550 nm. This allows one to use a higher laser power without risking damaging the retina.
  Luminar currently has a good patent position in the US market and a very weak one in the European market.
  

  Pulsed laser for Lidar system: US 10520602 B2, US10012732B2
  Group of patents which protect the laser source used in a Lidar system. Luminar seems to have a monopoly on using laser sources between 1400-1600 nm and repetition rates up to 100 MHz.
  In Europe patents were filed under the numbers, EP3411660A1 up to A4. These have all been rejected due to lack of novelty. Personally, I think it is strange the US patents got accepted.
  There was already prior by Trimble, see my earlier post, this system also operates a wavelength of 1600 nm and has an even larger scanning range. It was announced at Intergeo in 2016.
  
  Low profile lidar scanner with polygon mirror: US10571567B2
  Patent which protects a certain embodiment which uses a rotating polygon mirror with a galvo mirror.
  No patent seems to have been filed in Europe.
  
  Group of software patents; US10627521B2, US10209359B2, US10254762B2
  There is a collection of US patents which protects scan patterns, adaptive pulse patterns and monitoring the vibration of the car to improve the scan patterns. Software is not directly patentable in Europe.
  
  Plurality of digital converters; US10267918B2
  Still pending in Europe under a different title under number EP3602128A1-A4, search report requires amendment before April 2021. It seems likely that it will result in some sort of patent.
  

  Pulse Encoding; US10545240B2 
  Still pending in Europe, search report was not positive see EP3602112A1

  Summary
  I have looked into a lot of patents of Luminar Technologies, but couldn't find a single one which is accepted in Europe yet. Most patents are rejected, maybe one or two will be accepted in a much weaker form.  So far, building and selling your own car Lidar scanner in Europe still seems possible in the US, a non-free nation with respect to Lidar scanning, you would face many legal problems

• Icestorm meets Hexastorm

  Hexastorm • 11/17/2020 at 16:22 • 5 comments

  Finished new controller board for laser scanner and can now stream data via a ringbuffer to prism scanner. No more micro-controllers from now on but only FPGAs :-). Thanks to Claire Wolf, Migen and Litex team. Code can be found here.
  

• Dyeing textiles with lasers

  Hexastorm • 11/06/2020 at 10:41 • 0 comments

  On Tuesday 3 November 2020, I gave a talk on coloring textiles with lasers. Lasers allow you to locally tune the diffusion of a colorant into the textile by applying heat. My aim is to reduce waste and create a more sustainable world with this technology and promote my prism scanner :-). You can watch the video here.

  .

• Dutch state fighting open-hardware

  Hexastorm • 10/19/2020 at 11:28 • 1 comment

  I created an open-hardware project and got part of my inspiration while working for the Dutch State (TNO). The core idea is that a laser bundle is moved by rotating a prism.  The Dutch state got a patent for a plurality of laser bundles but not for a single laser bundle. For the printed circuit board application, it founded LDI Systems in 2015. This failed and they wasted multiple million tax dollars. I only spent 10K dollars on a working system and paid taxes. I thought they would leave it there. But TNO requested another subsidy from the Dutch government (NWO) to pump 45K euro into a new company AM systems BV for the 3D printing application. I have contacted NWO in this regard as I don't see how the original issues can be solved.  I also wonder how much "own" money is brought in (company is largely owned by the Dutch State). A company is not subject to tender law but the government is.  As such the company might be use to circumvent it. I am also not aware of employees in this company and know they contacted an optical consultancy. My work is free within the constraints provided by the typical licenses MIT and GPL. But I think is strange the state is still sponsoring this project. Hackaday gave me 3K and the Dutch states gives a chosen business developer 45K to explore their failed project. That's not really fair and in that sense the state it is sponsoring "closed hardware".
  Often the government throws away the result if unsuccessful, if it was open source one could at least learn something. Anyhow, I filed an official complaint and have talked to NWO.
  

  Links are no longer online, you can still find it in Google Cache. NWO stated it updated its website.

• Hackaday Prize 2020

  Hexastorm • 08/06/2020 at 13:40 • 0 comments

  Although, I do not plan to participate in the Hackaday Prize 2020. It is inspiring to look for sustainable applications of laser technology.
  In 2019, the United Nations set up a campaign for sustainable fashion. Most of the clothes are manufactured in Asia. This process requires a lot of water and produces toxic waste.
  Luckily, Dr. Laura Morgan, looked for ways to dye textile with lasers, see her extensive PhD thesis.
  

  She showed that lasers can be used to dye textiles. Results can be seen in the image below.
  

  The laser allows one to locally tune the diffusion of colorant into the textile.
  She points out multiple benefits for the environment. Production can be done locally,
  waste can be reduced and new colors can be added to existing products allowing reuse.
  

  In her laser experiments, Dr. Morgan uses 2 Joules per square centimeter. Results are improved with multiple passes, e.g. 15 passes. A carbon dioxide laser is used in her experiments.
  She also shows other applications; fading linen, increasing the absorption of wool and applying 3D texture via heat distortion.
  

  I think textile coloring could be a nice angle for the Hexastorm. If I have the time, I will do some experiments with a 450 nm laser at 10W. This is much safer than a carbon dioxide laser as infrared light is invisible. The process still requires a washing step but the chemicals involved are a lot less dangerous than the chemicals used for PCBs. The application I see is adding patterns to existing products. I think the process is too slow due to insufficient laser power for large scale production.

  Which brings me to an update of my current progress.
  At the moment, I am mostly busy with creating the FPGA electronics for the scanner.
  I have made a new board as I made a mistake in the previous board. It turned out that I couldn't use the SPI programming port for sending laser data to the FPGA.
  I also wrote new software, this has so far only been tested virtually.
  

• FPGA Pi Hat blinky

  Hexastorm • 06/11/2020 at 15:38 • 0 comments

  Managed to make the LED blink on my custom FPGA board. It was quite a challenge and I am glad I didn't go for a BGA package.
  I placed the ICE40 chip on the board with drag soldering. The other components you can hand solder except for the oscillator which requires hot air ( used AYOUE 852) and paste. In the next iteration, I plan to use mainly hot air (it is much faster). Used TS100 as a soldering iron, the Loctice GC10  for paste, flux and some desoldering wire to get the solder on the TQ144 component right. Biggest headache is that I didn't realize I had to send a wake up packet to the flash ram. I tried the python script ice zero prog to flash the memory but it didn' t work; turns out the wake up packet is implemented in icezprog but not in the old python script (OMG!).
  Also found out the board has too many caps and resistors, icezero uses less see schema
  

View all 68 project logs