Seeing the traffic light project got me to thinking that it might be interesting to expand the project by adding induction loop traffic actuators (exactly like how many signals actually work). However, I only have a really nebulous idea of how these actuators work: cars are metal, they change the inductance of the loop buried in the pavement. Great, now how would I go about measuring that change in inductance?

One way I've thought of would be to basically set up the loop as part of an oscillator and somehow measure its resonant frequency. I really don't know the best way to do that, however, so I'm asking here.

I have a couple of those devices for detecting vehicles. Basically, they are just metal detectors. They sit there and hum away continuously and when enough metal comes into its field, it affects the frequency it is humming on and the internal circuit detects that shift. This is caused by affecting the inductive properties of the loop coiled in the ground, usually 3 to 5 wraps of plain insulated wire.

Well it's good to know that I was right about how they work, but what I'm more interested in is what property specifically would I want to measure (which voltage) and would I have to have the MCU poll it continuously and then analyze the results over some time window or would it be possible to design a secondary circuit which can just 'flip on' which the MCU can measure?

Hi http://www.sm0vpo.com/_visitors/tech/pocketmetaldetector.htm I have built this circuit and it works great. I think you should be able to adapt it for what you want to do. a simple diode detector on the output would give you a voltage to sense.

That is neat circuit, and what a clean page so easy to read.

Ralph

Interesting. I think I'll try that one.

Thanks for the website. Not sure if I want to do digital mixing, but I've gone and sort of engineered a prototype circuit in the java circuit simulator here: http://www.falstad.com/circuit/

And you can get the import string for the circuit here: http://pastebin.com/amm7gcm9

It's all analog right now and the values for various components are completely made up (I'd likely have the oscillators running at much higher frequencies). It works in the same basic way as the above circuit. There are two oscillators on the left, one being a reference oscillator. I then pass the output from the oscillators into a summing amplifier. The output from this amplifier is sent through a high pass filter to cancel out the DC component of the signal which is then fed into a precision rectifier. The output from this is sent through a band pass filter to filter out the high frequency oscillations and any DC offset. This way, if both oscillators have the same frequency (but possibly out of phase), the output should remain close to 0. As the frequency difference (controlled by changing one of the inductors) increases the amplitude of the output increases as well.