Generating an AC waveform from a car battery

[LM741]

hi all. hope everything is well.

Just some advice please:

want to measure the displacement of a car's shock absorber's piston. Plan on doing this using a linear variable diferential trasformer (LVDT). Those of you that are familiar with such a device, know that its chief operation is based on a change of induced voltage in one of two secondary coils.
the problem: As we all know - the only way we can generate this magniti flux is by creating an AC waveform i.e. the LVDT requires an AC input.
How does one generate this from a car battery (as the device will sit in the shock absorber whislt the car is travelling). There is a device that sort of does what i want - but it does too much: http://www.edn-europe.com/article.asp?articleid=383.
I just want somethig that generates an AC wave from DC..
maybe just a mere DAC?

any feedback would be greatly appreciated!

thanks very much!
John Peter Criticos

 

[waht]

All DC to AC concerters are bascially oscillators.

A simple project would be to build a 555 timer square wave oscillator at 60 Hz if that's what you need, then connect the output to a high power transistor such as 2N3055 for more current, That would give you about 10 Volts AC, then if you want to step it up just hook it up to a step down transformer in reverse. Also you would have to filter out the output from the third harmonic 120 Hz for a purer sine wave if required.

Hope that helps.

 

[LM741]

Awsome! thanks a lot what- that's what i was sort of looking for! i'll do some research and come back to the forum to question some more!
thanks again
regards
John

 

[berkeman]

I don't think I'd use an LVDT for the function your are targeting. There are simpler linear and rotary sensors that could be used. Can you describe some of the other techniques that you considered, and why you chose the LVDT? Have you considered the noisy electrical environment of the car power system, and taken steps to keep that noise from muddying your data?

 

[LM741]

I thought about using a variable separation capacitor/inductor - but that's kind of the same thing isn't it? either way - I am still going to have to play with oscillators and phase demoulators/descriminators aren't I? Can't use a displacement sensor that works with a change of restistance coz a car always inhibits or is exposed to heat - and this of course - will cause undesired changes in resistance!
Do you have any suggestions- what sensor do you think is best suited for the application stated above?
Thanks very much Berkeman!

 

[LM741]

another reason: capacitive displacement transducers can only measure very small displacements - like the the displacement of that a microphone digram due to the pressure fluctuations caused by travellling sound waves. Maybe there are ones that can measure large displacements - but perhaps would be very expensive?

 

[triden]

They also make Linear Resistive sensors that are a long shaft that moves in and out. As the shaft moves, the resistance changes and that can be read as displacement. They are a lot cheaper than LVDT's too

 

[berkeman]

Can't use a displacement sensor that works with a change of restistance coz a car always inhibits or is exposed to heat - and this of course - will cause undesired changes in resistance!

Well, you have a good point about the harshness of the automotive environment, I guess. You might be able to protect the sensor some with some flexible booting/enclosure or something, to help cut down on the environmental exposure. And if you used resistive position sensing, you would do it ratiometrically, so that the heating effects would not matter. Like, you would sense the wiper position as the ratio of the two resistances to the two ends of the resistive material from the wiper position, not just the resistance to one end from the wiper.

I was also thinking of using a relative position sensor, with limit switches to give you an absolute reference. I don't know if it would work for your project, though. You would just need to sense quadrature ticks on a moving shaft or something, or turn the linear motion into rotary motion and use a rotary shaft encoder or quadrature encoder to give you the relative motion ticks (the quadrature info from the two sensors tells you what direction you are moving in). But if this sensor has to power-up and know right away what the absolute position is, then relative position/motion sensing won't do it for you.

I'd vote for a resistive sensor if you can protect it well enough, and use the ratiometric technique I mentioned. Sounds like a fun project!

 

[LM741]

fun? not really :) - don't get to build it! just have to do a hardcore report on it! - must include a final circuit design (with filters, ADC,DAC, amps etc) Can i ask why you so against an LVDT? really don't think anything to with resisatnce is good option in this type of app.
Another thing: Know of any free software programs that i can build/simulate such a circuit? Bored of PSPICE!
thanks

 

[berkeman]

Okay, I guess you've convinced me. I wish I knew what they use for racing suspension telemetry transducers -- do they use LVDTs as well?

I googled suspension transducer tutorial, and the 2nd hit was for LVDTs:

http://www.macrosensors.com/ms-lvdt_faq-tutorial.html

There's a lot of good info on that website, including applications information. You might already have all that stuff already, but if not, I hope it helps.

So back to your original question at the top of this thread -- you'll make the AC waveform with an oscillator circuit that is running off of a regulated voltage that you make from the car's power. You'll probably regulate the battery voltage down to 5V or something like that, depending on what the LVDT circuits typically look like.

 

[Averagesupernova]

As far as I know you can also get linear pots that aren't really pots. They send out 2 square waves in quadrature. X pulses per inch.

 

[LM741]

Berkemen >> Thanks a lot! I owe you! will take a look at that site! appreciate it!

thanks Averagesupernova

John