1
\$\begingroup\$

I'm new to ECE stuff, but I was wondering what kind of equipment I'd need to have an 8x8 or more grid of lasers being controlled with specific on/off patterns.

I'd want 5mW lasers in an 8x8 grid. It seems like I need an Arduino/Raspberry Pi, breadboard, wires, power supply, 8x8 matrix (or make one myself) laser diodes, collimating lenses... Does this sound about right? I see lots of articles for LED multiplexing, but nothing for lasers. Is there a substantial difference?

\$\endgroup\$
6
  • \$\begingroup\$ What kind of lasers? What voltage and how much current do they need? Is the current regulation included in the laser component or do you need to provide it? \$\endgroup\$
    – The Photon
    Commented Jan 28, 2015 at 20:27
  • \$\begingroup\$ Why a grid of lasers? A grid of LED's can make letters or such. But you can't really look at a grid of lasers. And if you shine it on a wall, well you'll want angular tweakers for each laser... Pointing angle of each will be somewhat different. \$\endgroup\$ Commented Jan 28, 2015 at 20:32
  • \$\begingroup\$ @ThePhoton Something like these, so 3V <40 mA current. Doesn't look like there's regulation in it. \$\endgroup\$ Commented Jan 29, 2015 at 19:32
  • \$\begingroup\$ @GeorgeHerold It's for scientific applications, I don't want to make letters. I want to make binary patterns with straight line paths with high intensity light red light. \$\endgroup\$ Commented Jan 29, 2015 at 19:32
  • \$\begingroup\$ @BenjaminCecchetto There are ways to avoid having to mount up a grid of diode lasers here, but they depend on your application. For instance, do you need to generate your light pattern only at a specific distance from the source, or does it have to look the same over a wide range of viewing distances? How big does each beam have to be? What is the separation between beams? Do you need to switch each beam on/off independently? How fast do you need to switch? \$\endgroup\$ Commented Jan 30, 2015 at 11:26

2 Answers 2

1
\$\begingroup\$

Summarizing the specs from the discussion so far: Working distance 30 - 100 cm Spot separation "as close as possible" but presumably the beams need to be well-separated Total power 8x8x5 mW, approx. 400 mW Independent switching of each beam, "not terribly fast"

Basically any solution will be challenging to implement for a beginner. There is a major time-money tradeoff as well. Given the questioner's level of experience, it's hard to imagine setting this up for much less than $5k if it's meant to get done within a few months.

I think the best solution is to get a single high-power laser and use a spatial light modulator (SLM) to make the pattern. You can imagine the SLM as being subdivided into an 8x8 array. Then each region of the SLM can be set up as a mirror with variable focal length. Switching between focal length of 30 cm and infinity will change the intensity at the spot. Alternatively, you could switch the SLM pattern to divert the beam entirely away from the target.

Get laser safety training before starting this project. Laser jocks have an old saying: "Do not look into laser with remaining eye".

Step by step:

  1. Set up a low-power diode laser and controller (e.g., a model HL6756MG laser). Use a telescope to expand the beam so that it nearly fills the SLM.
  2. Set up the SLM and test out some patterns. A lot of SLMs can be controlled over VGA just like a computer monitor. The SDE1024 SLM from Cambridge Correlators seems like a good bet for you.
  3. Program the SLM with a Fresnel mirror array and check out the quality of your spots on target. A Fresnel mirror is the reflecting counterpart of a Fresnel lens (http://en.wikipedia.org/wiki/Fresnel_lens).
  4. Switch out the diode laser for a higher-power model (e.g., the SLD1332V, with 500 mW power).

And you're done.

Thorlabs is probably the best place to purchase generic optics equipment. They would have all the diode laser and controller stuff. I only suggest Cambridge Correlators for the SLM because it's cheaper and specifically designed for beginners.

\$\endgroup\$
0
\$\begingroup\$

There are multiple ways to do this, I will outline one that I have seen demonstrated before.

If you want to multiplex one laser you could create an 8xN grid with 1 laser, 8 mirrors, 1 motor, 1 collimator, and a controller to control timing and firing of the laser.

Mount the mirrors evenly spaced on a cylinder or octagonally shaped object, and attach it so the motor will rotate the mirrors. Adjust the vertical angle of each mirror to achieve the rows in the grid. The rotating cylinder will achieve the horizontal angle to form the columns by adjustment of the rotation speed and the firing rate of the laser. A collimator can collect the dispersing rays and form a fixed grid, the size of which is limited by the collimator. The requirements of the controller might depend on the firing rate. Some perhaps tricky parts of this design may be: adjustment of the mirrors (involves some fine tuning), keeping track of the angular rotation of the motor so you know which mirror is currently in the firing line. On the down side the firing rate per pixel is limited by the rpm of the motor (120Hz for 7200 rpm). On the pro side it is probably cheaper and less tedious than buying and assembling 64 lasers.

Here's a laser projector project with AVR that uses this principle http://heim.ifi.uio.no/haakoh/avr/ (without the final collimator).

\$\endgroup\$
3
  • \$\begingroup\$ Thanks, but I do not want to multiplex one laser. I need the intensity contribution from many of them at the same time. \$\endgroup\$ Commented Feb 5, 2015 at 15:03
  • \$\begingroup\$ So, basically no multiplexing then ? Well for 64 simultaneously (but independently) controlled lasers, you should probably control them with an FPGA due to the width of parallel IO. FPGA is easily 10x more difficult than a microcontroller though, for a beginner. \$\endgroup\$
    – HKOB
    Commented Feb 5, 2015 at 15:47
  • \$\begingroup\$ What about a chain of serial to parallel shift registers? Connect that to an SPI port and you're good to go. \$\endgroup\$ Commented Mar 8, 2015 at 21:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.