Help with Inductive Proximity Sensor

Have you thought about using a crank position sensor from a automobile. Sounds similar to what the one on a GM vehicle does. It handles temperatures from below zero to 250 - 300 degrees I believe.

Hall effect - Wikipedia, the free encyclopedia

These are good ideas, but we need a pattern on this of slightly larger than what they provide. I have not looked at the Hall effect sensors at all. I actually think our loop works well, but we need something new to drive it.

There are a number of hits when you Google inductive proximity sensor

http://www.gepower.com/prod_serv/products/oc/en/bently_nevada/proxprobes.htm

The best sensors I have found for this are from a company called Proxitron (www.proxitron.com). Because of the metal wheels, it sounds like you are using this with some type of rail system. I have specified the large "hockey puck" style inductive sensors into this type of application and they work from -30F to 180F without any problems. You can get roughly 4" (100mm) away from the target with some of the larger units.

Over what sort of distance do you need to detect the wheel?

A simple metal detector circuit (many on the web) will detect iron and steel easily so long as the distance does not exceed the diameter of the search coil, so if you build a 25cm coil, it will clearly detect a wheel at 25cm distance (probably more, but for reliability stay with that)......

Make a coil with a diameter slightly more than the maximum detection distance.....Pulse induction type circuits are a good place to start with.....

PICAXE PICs are probably the easiest/cheapest to program and use, All software that you need is free..... Check out their website at:-

Picaxe PICs

This will be used to detect a train wheel rolling over a point on the rail. None of the sensors mentioned meet the requirements. I want to design a frequency shift detector for this application. I think this will provide the best results. These sensors are also subjected to things that will require them to be smart enough to "retune" when they are subjected to a change over time.

There are many details I don't know about this application, but it seems to me that a simple PZT accelerometer would easily detect a train wheel passing over it. This would produce a huge spike in the accel's output. A simple rectifier would provide a DC output, and a threshold detector could provide a positive "wheel found" criterion.

PZT has a wide temperature range, and protection from water should be easy to figure out.