measuring frequency response in piezo pickups

[mmichael]

Hi:

How would you measure frequency response in a piezo musical instrument pickup?

I've made several and am interested in doing some comparative analysis. I taped a pickup to a monitor and recorded some white noise, but the curve looked really strange. Granted the monitor isn't perfect and the location of the pickup will affect how it flexes, but I'm guessing this arrangement isn't ideal.

Thanks.

 

[Nigel Goodwin]

First point, you need to use pink noise, not white noise.

Or use individual frequencies and plot a graph.

 

[Brion55]

Your source would have to be very flat over the range you are wanting to test. attaching it to the monitor would pickup all kinds of box resonance and such and not much from the high frequency transducers. Probably not very reliable. I would think some kind of frequency sweep as Nigel suggested would be the way to go.

 

[mmichael]

Thank you very much for the responses. I wondered why the white noise sloped to the right in my software analyzer. I looked it up and now I know.

I am going to try plotting a graph of individual frequencies. I also thought I might suspend a thin piece of wood in front of my monitor and tape the piezo to it, then recording as the pink noise vibrated the wood, hopefully simulating a guitar soundboard. I realize the vibration of a guitar soundboard relative to an attached piezo is complex, especially when its played with a human body contacting it, but it seems like an interesting thing to try.

Thanks again.

 

[Nigel Goodwin]

Thank you very much for the responses. I wondered why the white noise sloped to the right in my software analyzer. I looked it up and now I know.

I realise 'pink noise' sounds a little silly, but like you say, now you now!.

 

[fdeck]

Here is another possible idea. An ideal piezo pickup is approximated by a current source in parallel with a capacitor. Departures from ideal might show up by comparing impedance plots for the two pickups. This has the disadvantage and advantage of not attempting to devise an ideal source of mechanical vibration.

 

[mmichael]

How exactly would you recommend measuring the impedance of the piezo (for comparison to the current source in parallel with the capacitor)?

Thanks.

 

[Nigel Goodwin]

How exactly would you recommend measuring the impedance of the piezo (for comparison to the current source in parallel with the capacitor)?

Same way as you would anything, measure the output with no load, then apply a load and measure again - use ohms law to calculate the impedance.

You could also use a variable resistor as the load, and adjust it until the output is halved, then simply measure the value of the pot on your meter. Exact same principle, but avoids any simple maths.

I would suggest using a scope with a x10 probe (10Meg) or x100 (100Meg) so as not to load the pickup.

 

[mmichael]

Thanks very much. These comments are providing a real education for me.

 

[phase_accurate]

Since these transducers are reversible you should also be able to couple two of them together and use one as an exciter while the other one is the DUT.

If your measurement is in dB then your actual response is the deviation from linear in dB divided by two.

Regards

Charles