Reproducing the Velador Experiment

Camera selection was initially based primarily on size and price, with resolution being a largely secondary concern.  The camera’s CCD chip can be damaged by the laser, even using an attenuated beam.  Price was judged more important than resolution for initial experiments due to the high probability of needing replacement cameras during the course of the experiment.

The resolution of the CCD chip used in Osadchey’s original experiment was 2 Megapixels.  The measured image spanned up to 96 pixels with that setup and the effect at least 6 pixels.   I initially estimated that this effect should be consistently observable by measurement at as little as 1/4 this span.  This would allow use of a CCD resolution as low as 125000 pixels, or 354x354, for detection and analysis.  This is sufficient to allow analysis using QBASIC software. 

Inexpensive low resolution, low memory CCD cameras are available that are suitable for this experiment.  It may be desirable to rig the camera switches to activate remotely, but real time data collection may be possible within the limits of camera memory.

The initial selected camera was a Phillips keychain digital camera with a maximum resolution of 288x352 pixels, which I expected to be marginally adequate for the desired observations.  This camera cost approximately US $20. Its plastic frame measures 1.5cm x 4.0cm x 5.7cm, with the CCD aperture offset by approximately 1.5cm frame-left and 0.3cm frame-top from the front center of the camera.  This camera is small enough to mount inside the hollow support beam.  The camera’s field of view is approximately 45 degrees.  It is capable of continuous video recording and remote image capture via USB cable.

This choice of camera, together with use of a shorter support beam, resulted in a pronounced decrease in sensitivity vs. the original Velador experimental design.  If the 2mm wide laser beam image spans 1/3 of the camera’s horizontal field of view, the predicted motion of 48 µm (for a 2.45m support beam) would span only 1 pixel on the CCD (6 pixels devided by four to account for decreased resolution, then devided by an additional factor of 1.5 to account for actual geometric performance of the CCD).

I have selected a replacement camera with 1.3 megapixel resolution - a Vision Quest VQ1005.  It is also a keychain style camera, with the same approximate dimensions to within 0.5cm.  It does not provide real time recording at resolution any greater than that of the original Phillips web cam.  Still image recording will be necessary for high resolution photos.  It will be necessary to disassemble the camera to rig a remote switch and remove the lens.

The camera and mounting hardware should optimally weigh only as much as the laser and its mounting hardware.  The camera’s mass is less than 50 grams, but mounting hardware is anticipated to be at least 200 grams, necessitating counterweights or similar measures.  It uses a single 1.5 V AAA battery, with an operating duration of less than 30 minutes.  The camera is compatible with a universal adapter on a setting of 1.5V.  Given that the power curve of a half-bridge universal adapter peaks at 15 V, extended use with a universal adapter should be withheld until after the first data series is taken.

Updated 10/15/07

I have obtained a new camera after burning out Camera 2 in a failed attempt to adapt it to a regulated variable power supply.  Camera 3 is a Digital Concepts model 57482/57483 4.1 Megapixel camera.  Its field at the 4.1 megapixel setting is 2304x1728 pixels.  I am using the 256 kB SDRAM chip salvaged from camera 2.  This camera does have a date function (which is set), and will assign dates to file names.  It  is compatible with 1.25V NiMH rechargable batteries, and uses two AA batteries.  

I intent to withhold tests using the regulated power supply until a few more trials worth of data are available.