WT3 piezo test

[marty]

this is a screen capture of wt3 and a test i done on 2 piezos, the green line is a goldwood piezo 1016, the purple line is a cpc farnell piezo mpt 016.
the 2 horizontal lines along the top are something to do with phase response.
i dont know much about wt3 yet, just the basics.
can i use this software to identify good and bad piezos, and if so what do i need to look for
what does a bad piezo curve look like?
thank you

[Chris_Allen]

Interesting to see that the Goldwood has a big response dip around 3K, where the CPC doesn't. I take it that all the response curves of the good piezos come out like this?

[Sydney]

Early last year ( when the issue of leaky piezos came up ) I ran impedance sweeps on 2 "1016" style piezos I had, using WT2.
( * I can send them to you if you like - PM please )
I used a 10 ohms resistor in series ( for stability when using capacitance load and to closer emulate a likely "real world" implementation ).
That said - my sweep - from 200Hz to 20Khz, has 300ohm impedance up till approx 1500Hz where it begins dropping.
The decent is fairly smooth WITHOUT the strong peaks that your sweep has, and the phase line was NOT flat.
( So I don't know why the discrepancies )

[Bill Fitzmaurice]

Interesting to see that the Goldwood has a big response dip around 3K, where the CPC doesn't. I take it that all the response curves of the good piezos come out like this?

Those being impedance charts they don't necessarily reflect response. Impedance sweeps might show good/bad, but to do so one would first have to identify a good tweeter and a bad one and then measure both to see the difference, if any.

[marty]

ok heres another chart, i have 1 bluearan piezo that i tested, that is the orange line at the top, they did not sound as loud as goldwood or cpcs,
the sound was much lower, i only have 2 goldwoods, they are the green and purple lines. the yellow and darker blue line are cpcs 016, i tested a pile of these cpcs 016 and they produced similar curves to this.
i have 1 double piezo cpc mpt-177, that is the light blue line at the bottom, obviously was louder than the others.
i just used the 1k ohm resistor that came with wt3 for the impedance calibration.

[Sydney]

FWIW:
In audioXpress Oct 2008 - Testing Loudspeakers: Which Measurements Matter, Part 2 ( by Joe D'Appolito ), is listed possible causes for wrinkles in the impedance plot: A resonance, a vibration, or a standing wave.