T 4.1 two piezos crostalk problem

Lestra

Active member
There are two piezos mounted on acoustic snare, one under the mesh head and second on the shell. Works ok but with crosstalk, false triggering. What is the best general approach to overcome this problem in the code? I am using Teensy Piezo Drum code, modified for two sensors. Thanks.
 
Doubt its crosstalk, they are both picking up the impacts. Probably needs sensitivity tuned to prevent false positives. First step is look at the waveforms with a 'scope to see exactly what's happening.
 
I am looking for a way to separate signals from the sensors; physically, it's not possible to a greater extent. There is a difference in signal strength because the strike on one sensor is stronger than the vibration that is transmitted to the other, and there is also a slight delay. The third approach that comes to mind is the one used by Roland in their electronic drums, which involves reversing the polarity of one sensor and then solving the problem through software. I have a lot of experimentation ahead of me, but before that, I wanted to check if any of you guys have experience with this issue.
 
I have two piezo sensors positioned one above the other, with some degree of physical isolation. When the first sensor is struck, its signal is stronger than that of the second sensor, which receives the hit partially through mechanical transfer, and vice versa. So, there is crosstalk but also a difference in amplitude. The second sensor is connected in reverse, with its positive terminal to ground and negative terminal to the analog pin. Both signals go through a conditioning circuit according to the instructions suggested by Paul Stoffregen here:
https://forum.pjrc.com/index.php?threads/test-code-for-piezo.48308/post-161683

The idea is to use the amplitude difference and to analyze and sum (or subtract) the two signals. This way, when the first sensor’s signal is stronger, the difference will be positive, and when the second sensor’s signal is stronger (since it is connected in reverse), the result will be negative. The issue is that Teensy doesn’t handle negative voltage values, so I would appreciate guidance on how to achieve this. I've learned about the DC offset method, but I’d like to know if it’s compatible with the existing conditioning circuit and whether it might pose any risk to the Teensy itself.
 
Bias the piezo to midrail using a voltage divider thats 1:1, and ensure there is a high value resistor across the piezo to discharge it to 0V dc (piezo's are capacitive). Then you'll have a mid-rail signal you can work with.
 
MarkT sorry for delay.
 

Attachments

  • circuit.jpg
    circuit.jpg
    62.5 KB · Views: 3
Back
Top