Vibrating wire sensor idea- How would you measure frequency?

Christof Eb. · 2012-03-09 12:12

Hi,
the idea is to build an experimental "vibrating wire force sensor". This is a wire string. Its frequency is proportional to the square root of the applied force. The vibration shall be amplified from a magnetic pickup. Either sigma delta adc or simply digital input to the prop. Sounds easy todo? There will be noise. How would you measure the frequency (somewhere about 200-20000hz?)?
I wonder if it is possible to do something like phase locked loop with the counters?
Resolution 0.01 per cent?
Some ideas welcome!
Christof

Phil Pilgrim (PhiPi) · 2012-03-09 12:32

How do you plan to excite vibrations in the wire? Do you think 200-20000 Hz is a reasonable range for a single wire? After all, each string on a stringed instrument has a different composition to accommodate its range, and the ranges are much narrower than 100:1.

-Phil

Duane C. Johnson · 2012-03-09 12:53

Hi Christof :

One could essentially "Pluck" the steel string with a magnetic pulse from a solenoid coil.
Then, immediately measure the frequency during the decay period.
Magnetic reluctance pickups can be directly coupled capacitivly into a prop pin.
Or, a capacitive pickup directly coupled into a prop pin.

I haven't done this with a prop but I have done this with another micro and it worked well.

Duane J

Rayman · 2012-03-09 13:04

I think I'd use electric guitar parts...

jmg · 2012-03-09 13:13

Christof Eb. wrote: »

How would you measure the frequency (somewhere about 200-20000hz?)?
I wonder if it is possible to do something like phase locked loop with the counters?
Resolution 0.01 per cent?
Some ideas welcome!
Christof

Here you go :

http://www.daqarta.com/dw_freq.htm

A couple of years back, we helped Bob expand the frequency counter feature...

If you are patient, you can get surprising results from a sound card.

Christof Eb. · 2012-03-09 13:48

Thanks for your comments! My question is really about ideas for fast and accurate measurement of a frequency somewhere in a hearable range. I don t know yet what range can be used. I actually want to try a non-wound guitar string and the coil of a relais or perhaps two of them. One as pickup and one to move the string. It would be cool, if one could use the same coil for pickup and picking.
Christof

groggory · 2012-03-09 14:48

Sounds like you're designing a microphone.

"fast and accurate measurement of a frequency somewhere in a hearable range"

I don't get what the source is for the vibrations in the first place. I think you need to fully define what exactly it is you're trying to do.

jmg · 2012-03-09 15:39

Christof Eb. wrote: »

Thanks for your comments! My question is really about ideas for fast and accurate measurement of a frequency somewhere in a hearable range. I don t know yet what range can be used. I actually want to try a non-wound guitar string and the coil of a relais or perhaps two of them. One as pickup and one to move the string. It would be cool, if one could use the same coil for pickup and picking.
Christof

See #5
By using a Sound card, or similar 'good ADC' scheme, you can interpolate to get a more precise zero crossing, than the sample-rate indicates. You can add a simple bandpass filter or tone control, to lower noise and get cleaner zero points.

Mark_T · 2012-03-09 16:00

Rather than thinking in the frequency-domain, consider the time-domain. Excite the string with a fast magnetic pulse, then wait for the "echoes" that return as the pulse races back and forth along the string - you can time several successive echoes to increase resolution. A bit like driving an ultrasonic distance sensor?

One issue with this approach is that you want each "ping" to die away substantially before the next, but the string may have a long decay time (certainly at the fundamental).

Tracy Allen · 2012-03-09 21:14

As pressure sensors these are "piezometers", used in drilling and maintaining wells, and in many other difficult locations because they can be so robust and accurate. Others are used for strain measurements, for example in bridge girders. A primo manufacturer is Geokon. There is good info there on how they work and on recent developments.

Here is a diagram from the Campbell Scientific manual for their vibrating wire interface, the AVW1:

Piezometer - (AVW1, Campbell Scientific).png

The interface works by scanning the drive frequency quickly across a range of frequencies, about 1 octave in the 1 to 10 kHz range, and subsequently looking for the ringing at resonance. The Q factor of these things is remarkable. They claim about 300 to 1000 cycles for the amplitude to die down to 1/e. (That number of cycles is equal to the Q factor.) A more recent interface also scans but then does what amounts to an FFT on the data to perform a spectral analysis for greater precision. All this is done over two wires, and two more are for a thermistor for temperature compensation. I see that Geokon now offers a module that self resonates and transmits back a frequency without needing an external excitation.

I had discussions with a customer about building an interface to these using the Propeller, but they did not get the contract so I never got to play with it. :-[ . My crazy idea at the time had been to feed a wide spectrum random noise from the Prop down the line, to excite resonance, like a sea shell.

Christof Eb. · 2012-03-17 11:55

Hi alltogether,
thank you for your comments!

I have posted now my experiment:
http://forums.parallax.com/showthread.php?138702-A-Physics-Experiment-Measure-a-Force-or-a-Weight-with-the-Propeller

I could use the method to measure time for a certain number of cycles to get the frequency.
Mark- I would like to use your direction of thought, but when I switch the same coil after the pluck back to reading there is a recovery time for the opamp. During this time there is no signal for several cycles.

Christof

Vibrating wire sensor idea- How would you measure frequency?

Comments