ADC basic.

[fzn10]

Hi, I got given this project, in which I have to some hardware in it, iv never done hardware before and really need to get started.

Now my aim is to measure how fast a propeller is spinning, another student will worry about storing that data somewhere. I could do that by measuring the voltage produced by a turbine as the turbine is connected to an alternator. The voltage produced by the alternator will vary between 0 and 300mV(p-p) depending on the rotational speed of the turbine. My frequency will also be speed dependent and will vary between a small range of (0-5Hz). Im using a PIC16F87X PIC, that has an inbuilt ADC converter. What I need to know is that does the analogue input has to be DC. If not Could I just measure the AC, and write a program that just measures what the amplitude of the waveform is? Sorry for the broad questions, but any feedback will be appreciated.

Thanks

 

[dknguyen]

Adc

The ADC measures the instantaneous voltage so AC and DC doesnt matter...as long as the input amplitude is within the ADC's limits.

You would have to sample the signal twice at least twice as fast as its highest frequency component (to prevent aliasing) and then figure out from all the samples taken the amplitude of the signal (you can probably do differentation between all the signals and take the voltage value of the sample whenever the derivative is nearest to zero.)

You say your signal will vary between 0 and 300mVp-p? So you are saying your signal varies from -300mVp to 300mVp (sorry, but the wording is a little bit open to interpretation)? If that is the case, you will probably need to bias your signal since the ADC probably cannot take a negative voltage input reading without damage.

Since your signal's amplitude is so small, you probably also want to use the bias circuit to also amplify the signal so you can artificially increase the resolution- that is, amplify the signal so at its maximum amplitude it takes as much of the ADC's maximum input range as possible without exceeding it.

 

[philba]

If the alternator produces a reasonably clean AC signal, why not amplify, rectify, filter and use a schmitt trigger to generate a pulse train. counting pulses is easier than reading an ADC to determine frequency.

Of course an opto interruptor or hall effect device would be a cleaner way to measure RPM.

 

[Nigel Goodwin]

Yes, you don't need an ADC, it's completely the wrong thing to use!. As suggested, amplify and square the pulses and measure their frequency.