Piezo Testing

[mtglass]

To reinforce the need for testing piezos, I had put 2 arrays of 6 together, and one piezo, although it measured the same as the rest on a calibrated SPL meter, it didn't have the same high end frequency reproduction - we found it using a spectrum analyzer and pink noise. That 16% made an enormous difference in the end result. Once I knew what the problem was with the array, I could hear the difference in the bad piezo, but my ear wasn't good enough to hear the frequency difference when I was just testing single piezos using my ears and an SPL meter. So my process is:

1) Test one piezo to set a baseline:
1.1) Use pink noise at a set level
1.2) Check the SPL output: same distance, and orientation, with a SPL meter
1.3) Check the frequency output: same distance, and orientation, with a RTA
2) Check the next piezo, by rinsing and repeating the above.
3) Wire the piezos together in pairs to test polarity using test leads. (I haven't found any issues here, it may be un-necessary)

So far, about 20% fail the above tests, but I don't have enough data to give a better idea of how many Goldwood GT-1016 piezos will be within tolerance.

Having a very quiet test room helps. One piezo isn't very loud, and I should probably push more voltage to them, but I get nervous.

Happy speaker building!