Catalog
Overview of demodulation techniques
Most people don't carry a dual trace scope or phase angle voltmeter around with them. So your going to need some method of reading and interpreting your LVDT output signals. There are a number of commercially available solutions for this but this page is concerned with more “cost driven” projects. I've built all of these and their total cost ranges from $5 to around a $100.
AD698 (The I don't know how it works and I don't care method)
The Analog Devices 698 and 598 ( for five wire LVDTs) are meant to be all-in-one solutions. If you use Visual Basic at home then this is definitely the direction you'll want to pursue first. They provide an adjustable excitation and proportional DC output. What happens in between is not our problem. Free sample are even available for the AD598.
Potential issues:
Excitation is limited to just over 20 volts.
The price for the DIP (if your a coffee drinker) version is sky-rocketing.
AD630 (The way your Grandfather demodulated his LVDTs)
This is another fairly good method for setting up your LVDT. The price is still slightly less then that of the AD698 and nearly as simple to configure. Although, this is a bring-your-own excitation proposition. Everything is pretty well spelled out in the data sheet. We're just routing the LVDT signal to either an inverted or non-inverted amp depending on the state of the excitation. Also, the phase lead/lag network is far less important then you might think.
The PSoC way
Alright I lied, I haven't actually tried this one yet. But it's just so much more interesting then my other versions. Here is a link to the EDN article where I found it. The PSoc is apparently, an other microcontroller but with configurable analog components built in to the IC. Instrumentation amps, multiplexers, ADC, DAC, various filters you name it. This definitely warrants further investigation for any hobbyist. Also you'll want to look at www.cypress.com
Frighten your coworkers with the mystical Double Subtraction technique.
This is one of my favorites, probably because its discovery was totally accidental (much like the fabled combination of chocolate and peanut butter). The secondary output of the LVDT is subtracted from the excitation signal. This leaves a signal proportional to the movement of the core and centered around the level of the excitation. Rectifying both the subtracted signal and the original excitation, you can remove the offset. Tah-dah!
This is more of an algorithm than a circuit, although we've made hardware versions just for fun. I have this one running in LabView 7 at work and it's the most noise immune version I've tried so far. I'll provide the code upon request. rstraight2000@yahoo.com
