Shop OBEX P1 Docs P2 Docs Learn Events
DVD Laser Diode And Optical Block (Future Laser Photoplotter) — Parallax Forums

DVD Laser Diode And Optical Block (Future Laser Photoplotter)

idbruceidbruce Posts: 6,197
edited 2014-05-12 05:00 in Propeller 1
Hello Everyone

Has anyone tried controlling a CD writer laser diode and optical block with a Propeller?

Bruce

EDIT: 02/17/2012, It now appears that a CD laser diode is incompatible with common films, having a wavelength near 750 nm. A DVD laser diode would be much more suitable for common films with a wavelength closer to 650 nm. As a result, I have removed CD from the title and replaced it with DVD.
«134

Comments

  • PublisonPublison Posts: 12,366
    edited 2012-02-10 10:34
    idbruce wrote: »
    Hello Everyone

    Has anyone tried controlling a CD writer laser diode and optical block with a Propeller?

    Bruce

    Hey Bruce.

    Is this for your proposed photo plotter? By "Optical Block", do you mean a focusing block?

    Are you going for a vector or raster scan photo plotter?

    If your going for a raster scan, your basically re-inventing a laser printer, spinning mirrors and all that. But it sound like you want to do direct to board. I'm not sure the wavelength of the CD laser would match the board requirements. I'm sure you have done some research on this.

    As far as a vector Photo Plotter, I don't think a Laser Diode would work. You need to vary the light intensity against the speed of the table, (or movement of a light head). Using stepper motors for a table and a stationary light source, the light intensity would have to vary with the ramp up speed of the steppers, as exposure is a function of how long the light source is actually on the film. A 45 degree plot would need more intensity than a 0 or 90 degree plot.

    I built and serviced Gerber Photo Plotters during the 80's, and I can say they were a little complicated. Joe Gerber was a genius, and between him and a great batch of engineers, they had it worked out. BTW, our lowest price Photo Plotter at the time was $50,000.

    Let me know if you have any questions. I'll try and answer them.

    Jim
  • LeonLeon Posts: 7,620
    edited 2012-02-10 10:41
    If it's for a raster photoplotter a red laser would be more suitable, as special red-sensitive film is made for them.
  • PublisonPublison Posts: 12,366
    edited 2012-02-10 10:51
    Leon wrote: »
    If it's for a raster photoplotter a red laser would be more suitable, as special red-sensitive film is made for them.

    You correct Leon. At Gerber we did a 2000 DPI Red Laser Photo Plotter back in the 80's which would do 4/4 using Kodak or Dupont film. I don't know what is available now to do the same.


    Jim
  • LeonLeon Posts: 7,620
    edited 2012-02-10 11:31
    Kodak still makes suitable film for 670 nm red laser diodes.
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-10 13:03
    I think someone has used a bluray laser to directly write to PCB laminates. And I have seen at least one drum based plotter using the film Leon mentions: http://pico-systems.com/photoplot.html

    Last year I was contemplating the former idea as I only ever do tiny pcbs.

    You can just treat the laser diode as a diode but just treat it carefully. I don't think it would be worth touching the rest of the components. Sam's laser FAQ is a great source for understanding how to drive lasers and there is also quite a bit on youtube and instructables.

    Graham
  • idbruceidbruce Posts: 6,197
    edited 2012-02-10 17:44
    Publison, Leon, and Graham

    Thanks for the responses.

    Yes this is for my proposed photoplotter.

    I am still in the process of researching the idea and I must admit that I am not that sharp about the definition of raster or vector graphics. However I will be aiming the RS-274X Gerber format, and the Gerber RS-274X Format User's Guide, which will be my main reference, is geared for raster plotting. It is my intention to expose film for this particular purpose, as compared to direct imaging of a PCB.

    Creating a Cartesian coordinate XY table is a piece of cake, but as you all realize, it is the optics and the driving software that will be the grunt of the project.

    I am sure there are many software solutions available, but I am also sure that they are probably pretty expensive. If the software is out of my price range, I may attempt to write something based upon the Gerber RS-274X Format User's Guide, which can be found here, http://mcglothin.us/RobotScrapbook/CommunityPcbMill2005/IntroductionToGCodes/rs274xc.pdf.

    As for the laser optics, I am still gathering information about this subject, before attempting any hook up that might damage the assembly that I have. According to what I have read pertaining to laser photoplotting to film and laser diodes, the laser diode from a CD player or writer should have the correct wavelength for exposing most available PCB films. I believe the optics will be the most difficult part. For my current research pertaining to lasers, I am now refering to these web pages: For the overall concept, I am now lookiing over the following to patent documents:
    • Gerber - Photohead With Flashing Beam - Pat. No. 4,363,539, which can be found at the United States Patent And Trademark website.
    • Gagnon - Photoplotting Apparatus - Pat. No. 4,809,028, which can be found at the United States Patent And Trademark website.
    There are many other patents pertaining to the subject, but I chose these two as my starting point to gain a perspective of this project.

    It will probably take me some time to get a good start on this project, because of the other irons in the fire, but I try to obtain research for future projects whenever I am interested or have a little free time. Out of necessity, I will have to purchase photoplots until I get it built.

    My intention is to limit the gantry movement to 4" X 5", with a 1" margin around the plotted image, leaving a 3" X 4" exposure to match my exposure box/cylinder.

    Thanks for your interest and input.

    Bruce
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 04:01
    Here is some further info pertaining to your project.

    If you intend to print on film then seriously consider the drum based system that I linked to. You can even buy commercial plotters that work just like this. Because of good old c = pi.d you don't need a big drum even for fairly large films. You just need a decent gear ratio to drive the drum from a stepper and then a single suitable linear axis. The software is no more complex as the relationships are all linear in fact I have done engraving on cylinders by plugging in the rotary axis into the y axis drive before adjusting the steps/mm constant.

    As for optics, if you intend to use a laser diode then the only optics required is a single focussing lens if the beam is already collimated or something to collimate and then focus, that could be just a pair of lenses. A decent red laser pointer with a small lens taken from a CD player would probably work fine. I'm assuming that you would move the laser diode around rather than attempt "flying optics". There are loads of people desperately trying to burn things with tiny laser pointers, this is basically the same issue, making a small focussed dot, you just won't want quite as much power.

    As for software I think you can get gerber to g-code converters already and then you could use a PC running turbo CNC or mach or create something with the propeller.

    Graham
  • LeonLeon Posts: 7,620
    edited 2012-02-11 04:36
    Bungard makes a nice drum photoplotter:

    http://www.bungard.de/index.php?option=com_content&view=article&id=43&Itemid=60&lang=english

    It costs over £6,000 from Mega UK.

    I think that the need for a darkroom with a low-cost photoplotter will turn off many people. The nice thing about my process using an inkjet printer is that I don't need any special facilities.
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 04:50
    Yep, that's the commercial one I had seen. I expected the film attachment method might have been a little more classy for £6k, masking tape!!

    Jon Elson's as linked to above is basically the same (even uses tape to hold the film). Very simple mechanically would not cost anything like as much as the commercial one even if you took everything off the shelf and did not touch surplus. Actually for a lot less money you could be using one of these for the linear axis: http://dunkermotoren.com/default.asp?id=118&lang=2 and a harmonic drive between motor and roller.

    Graham
  • LeonLeon Posts: 7,620
    edited 2012-02-11 05:00
    I was a bit surprised at that. It's what I use to attach the slightly over-sized piece of JetStar film to an A4 carrier sheet for printing.
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 05:11
    Graham
    If you intend to print on film then seriously consider the drum based system that I linked to.

    When I first started researching homebrew photoplotters, I came across that same web page, and I did seriously consider a drum based system. However at this point, I have a lot more knowledge pertaining to linear based systems as compared to rotary systems.

    I was thinking something more similar to the laser photoplotter shown exposing film in this video, but only on a much smaller basis: http://www.youtube.com/watch?v=JG5qUhn9sSM

    I have not really dug to deep into the patent documents yet, but I believe the laser light is passed through mylar or another type of film which contains various line widths, apertures, pads, etc... This film is probably rotary and contains numerous images. If you need a certain pad size, simply index to the appropriate image on the film and flash for exposing the image. I am not 100% positive, but I believe it must be something similar. Just like a photomask for exposing boards, only in this instance, it exposes the film to different shapes and line widths.

    Bruce
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 05:24
    idbruce wrote: »
    When I first started researching homebrew photoplotters, I came across that same web page, and I did seriously consider a drum based system. However at this point, I have a lot more knowledge pertaining to linear based systems as compared to rotary systems.[/URL]

    What is there to know? Rotary things go around, that's about the long and the short of it. The steps per mm will be the steps per rotation / circumference. Simple! For a single rotation you can even use a spring to remove backlash.

    I don't think the patent is particularly relevant, what you want to do is basically plot with light onto a film like a CNC machine, I don't see why that is complicated, they would use masks in the past because they did not have computers.

    If you wanted to get something working quickly then you could output the artwork as a BMP/TIF of the correct resolution and then read that directly with the propeller, then process the pixels line by line, they tell you when to turn the laser on and off. You could run any old code to move the gantry part and sync the turning on and off as appropriate with another cog, then turn the roller by a distance equal to one line of the image and repeat. The same could be done with a gantry if you insist. This is what I planned to do. If you use the vector gerber files I think you have to do a lot more coding to make sure the correct parts are filled etc.

    Graham
  • LeonLeon Posts: 7,620
    edited 2012-02-11 05:25
    That's the technique used with the original Gerber photoplotters, with a Xenon flash lamp. I don't think that it's used any more.
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 05:36
    Please refer to page 12 of the Gerber RS-274X Format User's Guide. It is sample code where an aperture is selected, and then a flash command is sent. I cannot think of any other reason for this then the one I stated above.

    Bruce
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 05:41
    Yes but you would not actually implement apertures physically, you would emulate them in software. So you have a definition of a pad that can be applied at a position. But they will not have actual apertures in the machines these days, why would they.

    If you had explained your comments about the patent were related to understanding the software and not the hardware I would not have commented.

    Graham
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 05:47
    But they will not have actual apertures in the machines these days, why would they.

    To me, I would assume it would be much faster to expose light through an image on film, as compared to creating an aperture with a very fine dot. Additionally, there would have to be software routines to create the apertures in addition to an aperture command.

    Bruce
  • LeonLeon Posts: 7,620
    edited 2012-02-11 05:50
    That would be the case with a Xenon flash lamp, but you will be using a small dot of laser light and building up the image from the aperture.
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 05:55
    idbruce wrote: »
    To me, I would assume it would be much faster to expose light through an image on film, as compared to creating an aperture with a very fine dot. Additionally, there would have to be software routines to create the apertures in addition to an aperture command.
    Bruce

    By the same logic we should stick to type writers or those IBM printers with the cool balls with all the letters on!

    Photoplotters almost certainly print like a printer does, line by line, think about when you need to fill a large area, you could have a selection of apertures and try and build it up but better to just use the fine dot line by line. The film will be sensitive and it will not take long.

    And even if some still use that method I don't think it means you should.

    Graham
  • LeonLeon Posts: 7,620
    edited 2012-02-11 05:57
    I think that the Bungard web site has plotting speeds for that photoplotter I mentioned.
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 06:00
    Okay I agree because you will have irregular shapes and pours that cannot be compensated for on film, so it will have to be seperate routines. So I am now assuming that the RS-274X data is sent to a software interpreter to carry out the proper commands, such as Mach3, etc....
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 06:02
    If you go for the high resolution BMP approach I mention above you could do it all on the propeller with an SD card. Weekend project!!

    But it looks like the commercial plotter has software to interpret the gerber before it sends something to the printer (probably some compressed version of the BMP).

    Graham

    p.s. I mean those weekends with lots of caffeine and pizza
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 06:08
    Considering this approach, that will have to be one very fine beam of light :)
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 06:13
    Spot size is about 1.22.lambda/NA

    NA is the numerical aperture of the lens: http://en.wikipedia.org/wiki/Numerical_aperture

    Its the focussyness, big lenses focussing closely have a higher angular spectrum and produce smaller dots.

    Basically you can get close to micron size but 50um might be more suitable. A longer focal length and lower NA will be less sensitive to the distance between the film and the lens due to a longer focal "waist" or depth of field.

    Graham
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 06:14
    There are a lot of C++ routines and various GDI functions in C++ that would be great for creating the masks. But translating the RS-274X format to GDI would be a real task. I would assume that instead of sending a bitmap, the commercial photoplotters send it to the GDI.
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 06:28
    Graham

    On the other hand, perhaps you are correct, and perhaps it would be best to use a monochome bitmap and just interpret the bitmap file. If the pixel is black have the laser energized, if it is white have it deenergized.

    Bruce

    EDIT: Makes perfect sense to me, probably the easiest way to go. So if a person wanted to create a PCB for a friend from RS-274X data, the RS-274X data could be used to create a monochrome bitmap by sending the data to the GDI to create a bitmap, and then send that bitmap to the controller for interpretation. HMMMM

    EDIT: Pixel by pixel exposure. Sounds pretty simple to me.
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 06:48
    idbruce wrote: »
    Graham

    On the other hand, perhaps you are correct, and perhaps it would be best to use a monochome bitmap and just interpret the bitmap file. If the pixel is black have the laser energized, if it is white have it deenergized.

    Lost my post due to pressing the wrong button.

    You could even create a simple compressed bitmap format. For each line the numbers represent the numbers of pixels on and then off alternately:

    40
    10,20,10
    10,20,10
    40

    Would be a 20X20 box (assuming each line is 10 pixels heigh) with a 10 pixel boarder. Easy to create the file on the PC and would make processing at the propeller end easier too.

    Graham
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 07:36
    Graham, Leon

    Here is an interesting link that has access to a 274X to tiff converter http://members.optusnet.com.au/eseychell/ with much more interesting stuff.
  • LeonLeon Posts: 7,620
    edited 2012-02-11 07:46
    I knew about some of Adam's work.
  • Graham StablerGraham Stabler Posts: 2,510
    edited 2012-02-11 10:21
    Very interesting document about laminating film resists, very handy if you want to work with unusual laminates.

    Graham
  • idbruceidbruce Posts: 6,197
    edited 2012-02-11 15:31
    This page confirms the suspicions that I had earlier.
    http://www.artwork.com/gerber/appl2.htm
Sign In or Register to comment.