I've completed two laser microphone components that use the triangulation technique in conjunction with a high-bandwidth photodiode. These can be used to pick up vibrations remotely on any reasonably reflective surface. A production run of 12 is almost complete. Click through for pictures!
These sensors are being integrated into an embedded, battery-powered setup using the cortex m4f microprocessor. I will be using this platform to perform DSP on several channels of these laser microphones, with a few application specific algorithms. I'll write a little more about those later. In the meantime, imagine using these sensors to map the vibrations along any surface at a distance, and having a tiny computer capable of turning this field of information into a signal that is "optimal" for different definitions of the word. For example, we could extract a signal with the most musically useful material, diminished noise, etc. This of course only scratches the surface of what is being worked on, but hopefully you get the point.
First, the most important image of the series-- the larger boards with the black IC sockets are the instrumentation amplifiers for the photodiodes. There are 11 here, nearly finished. The twelfth is completely finished, and I'll show that one off in the next image. The smaller boards with the metal transistors are current amplifiers for the laser diodes. This allows me to control the optical power of the laser using a control voltage, a function which will come in handy if I want to do frequency shifting or time multiplexing, or some combination of both.
The next image shows the prototype model of the laser triangulation transceiver. There is no housing at this time, because a calibration solution is being sought out. The grey component on the left is the emitter, a cheap laser pointer. The black component on the right is the photodiode, which acts as a pickup when combined with the amplifier circuit shown in the bottom right. These are very similar to the transceivers that were developed in this post.
I had some issues scaling the signal appropriately to take advantage of the ADC's bit-depth, so I've put together a prototype of an intermediate stage to condition the signal. This is merely a simple voltage level shifter and amplifier. Once the ranges of shift / scale are optimized, these will be duplicated several times, and housed in a single box for easy calibration of many channels.
For an example of what these things can do, check out my water surface recordings.