# Need formula for core detection

I need to detect when an irrigation head is functioning, non-invasive. The heads are the Toro brand and when activated a cylindrical shaft extends several inches above ground. This shaft is a non-ferrous metal, I assume to be polished aluminum. It measures 1.75" in diameter. My thoughts are to wind a coil about 2" in diameter of say, 50 turns of # 22 awg. This coil will sit slightly above and centered with the head. I will energize the coil with a square wave derived from a pic chip at say 1khz. Will I be able to detect a change in the applied waveform when the head's shaft enters the coil? Does the core (head shaft) need to be of a ferrous material? Is there a formula I can use to calculate the coil parameters?

• Why not use a physical switch that gets depressed when the shaft extends, or a IR (or other IR) beam that gets broken when the shaft extends? – pfyon Jun 3 '11 at 14:46
• I'm assuming the activation is by water pressure, so there's nothing electrical you can monitor. But could you perhaps detect vibration when it's in operation? For your coil idea, I'd expect more effect at a higher frequency. You might be able to make an LC oscillator at RF and then build a frequency counter in the mcu, possibly after using an external prescaler chip to divide the rate down. But this may generate RFI - as your original idea could. – Chris Stratton Jun 3 '11 at 14:51
• How about a capacitor instead? You could make one with some aluminium foil. When the core emerges it will go between the plates of the capacitor and the capacitance will probably change and you could detect that using one of the libraries for capacitive buttons. – AndrejaKo Jun 3 '11 at 15:13
• I also thought of a switch. Perhaps a micro switch with a roller wheel. That may be what I need to do, although it is somewhat invasive. As for IR, I think when the sun is out it will create havoc. Vibration is something I could consider, however I am first looking for a solution that I can see will work. A lot of experimentation would be an issue. The capacitor might work, but I would think it would need to be built close to the heads shaft, creating all kinds of alignment issues. All good ideas, thanks! Any others? – SteveR Jun 3 '11 at 15:42
• If you go with IR, first you would want to shield it, but you'd also want to use a scheme where you take two readings, one with the IR source on and the other off, and compare the difference. Actually, for almost every active scheme under consideration, power consumption would be minimized if you sleep for as long as permissible, wake up and do a quick active test (or 2) and then go back to sleep. How long depending on the allowable lag in detection, from milliseconds possibly even to a few seconds? – Chris Stratton Jun 4 '11 at 1:41