Help finding a displacement sensor

[Rhythmtech]

I am trying to work out the best way to measure displacement of a shaft that will be moving in and out with about 30mm of travel at about 60 times a second.  I will need to be to the nearest .1 mm or better and be able to capture the change in direction fairly accurately. +/- 5 or 10 volts DC output and fairly rugged. 

My first though was some sort of laser sensor pointed at the end of the shaft. 

Anybody have any ideas or suggestions?

[JohnS_AZ]

Get out your checkbook! 

http://www.omega.com/pptst/LD620.html

[Zad]

My first thought would be a capacitive or inductive sensor, but this usually needs an additional arm to be attached to the device under test. An alternative might be a binary Gray code optical scale attached to the object and a linear array of photodiodes focussed at the shaft.

[DJPhil]

That's a lot of motion for a mechanically coupled solution to handle on the cheap. LVDTs are available with insane ruggedness, but they cost quite a bit. I'm not sure if the low end LVDTs would hold up, but they might be just fine. The ruggedness of any approach can be very good. An LVDT will have little trouble with dust or debris if it's bolted to the shaft (most are designed this way). An optical solution is more vulnerable, but is likely to be a lot cheaper and have no problems with wear.

What information do you need? Just the peak time on the shaft (high and low), zero crossing, or do you really need 0.1mm analog accuracy on the complete travel?
The expense of the good linear displacement sensors comes in making it linear and accurate. If you don't need the median values you can save a lot of money.

Do you have transverse access to both resting points?
If so, optical interrupters might be a good idea, either hooded IR or laser. You can use interpolation in software to estimate median values if you need them. If you're not using a micro, well, it might be cheaper to put one in than to get a good sensor.

[NiHaoMike]

A linear quadrature optical sensor should work nicely.

Capacitive encoders, similar to the ones in electronic calipers, might also work. (I have a cheap set of calipers that can resolve down to 0.01mm.)

[TopherTheME]

If you're measuring a displacement of something at 60Hz you're going to have a hard time finding a displacement sensor that can operate at that speed. I would look into using an accelerometer and then integrate its signal to find position.

[gonnafail]

I think what you are looking for is a laser interferometer. I have used these successfully in the past. They are not cheap though but very precise.

best of luck.

[NiHaoMike]

Is there a rotating part whose rotation is correlated with the shaft travel? If so, you could use a resolver. That's how modern automotive ignition systems determine piston position.

[tecman]

The most common transducer as you have described is an LVDT.  It is a non-contact magnetic device.  Characteristics are very rugged (used in aircraft for years), works over a very wide temperature range, high accuracy and the response you are looking for is no problem.  It requires an oscillator/demodulator to derive the signal.  They can be had for $300+.  Look for a surplus unit if you need cheaper.

Paul

[Rhythmtech]

Thanks everybody. Looks like the LVDT is the way we are mostly likely going to go