Angle Transducers? Bird Tracking

You don't say how the tracking is to be done - presumably you are thinking in terms of observers manually aiming a sight of some sort. Other ways are possible of course - of which the most interesting might be using a pair of video cameras.

So, if the problem is to work out where a sight is pointing, then I would consider using high quality potentiometers (called servo pots), and then simply digitising the output voltage. You can get simple A/D converters that plug into a parallel port. The voltage from the pots can be sent over a simple cable if the environment is electrically 'quiet' - quite likely in this case. If it is noisy (near civilisation !) then a simple amplifier/buffer at each obs point would be needed [ The signal needs to be sent differentially ]

Alternatively you could do the whole thing digitally with encoders etc - which would cost more, but has the potential to be more reliable, repeatable and possibly more accurate. Of course, with people in the loop doing tracking, high accuracy is probably a moot point ..

regards,

Dave

Dave, Thank you.

You're right. I expect the observers to manually point at the birds while looking down a stick or tube. Tracking accuracy is not critical since we're talking about distances of less than 1000 feet. If the computer is able to track and calculate distance, it will be easy to use a very constant feet per second calculation to help correct errors and reject bad readings.

The biggest challenge for the observers, the mechanics, and the tracking software will be when the birds cross straight over head, or suddenly change course as they reach an observer. The vertical measurement must go well beyond 90 degrees. The horizontal must turn countless circles.

servo pots? ok, I'll look into it. I don't think we need more than a couple dozen samples per second. Could it work to run analog for the 100 to

Considering that we'll use four to six pots, how would you interface to a PC? I'm not familiar with capturing multiple A/D readings in to a Parallel port. Is there a supplier or circuit description you could refer me to?

When you suggest doing the whole thing digitally with encoders, what parts are you thinking of? Are you suggesting digital signals between the observation point and PC?

transducer?

You may need a triple gimbal - the obvious dual design can suffer from 'gimbal lock' as you pass the zenith !

Servo pots are simply high quality potentiometers - most of them don't mind going round and round (but they only measure ~ 300 degrees or so).

If the environment is electrically quiet, then you may get away with simple cables. If not then I would suggest a simple op amp circuit (placed at each sight) to drive the long cables differentially. I'd make it battery powered for simplicity.

See

Well, there are an infinity of possible solutions ! Starting with a digital angle encoder, you should note that most of the less expensive ones are incremental, so right away there is a calibration issue. In practise you would need a simple micro to keep track of the encoders steps, and send a readout down the cable to the PC. On balance I'd probably use serial (RS232/485) comms here.

In practise I would guess that the observer will be a significant source of errors anyway, so the analog approach is probably perfectly adequate, as well as easier to debug in the field !

If this were my problem I'd start by seeing how good the observers were at aiming. I'd test this by strapping a video camera to a sight to be aimed. By looking at the video afterwards you could see just how accurate the tracking process was. This would set you an error budget that you could use to design a suitable system.

regards, Dave