piezo signal conditioner for e drum

Audioguru

Joined Dec 20, 2007
11,248
I think the piezo signal voltage will be so high that it destroys the input of the opamp. Use clamping diodes.
The opamp gain is 11 times which is much too high.
The diode forward voltage throws away a lot of the signal envelope, use an active peak detector (with the diode inside the negative feedback loop) instead.
 

Thread Starter

seayaker

Joined Jan 27, 2009
74
I think the piezo signal voltage will be so high that it destroys the input of the opamp. Use clamping diodes.
The opamp gain is 11 times which is much too high.
The diode forward voltage throws away a lot of the signal envelope, use an active peak detector (with the diode inside the negative feedback loop) instead.
Dear sir, I am not an electronics expert or even a hobbyist, I've been a professional musician for over 50 yrs. I built this unit with the intent to use it professionally but I need to get it to register the hits more accurately. I have been struggling with this for a long time. I can put circuits together but I don't have a scope or any test equipment. I would be externally grateful if someone could put up a diagram or add to the schematic above so that even a dummy could build it. I have several mcp6004s if they would work. Thank you.
 

Sensacell

Joined Jun 19, 2012
3,432
Here is what I did- it's polarity insensitive, I had issues with the signal phasing, this circuit doesn't care, you get a clean pulse from both excursions, positive and negative.

The design is oriented towards lowest latency and adjustment range.
It produces many pulses each hit, you need to add de-bounce to resolve single hits.

The MCP6024 works from 2.5 to 5.5 Vcc, the circuit should work on 3.3V but the output will be 0 - 2.7 V - a diode drop less.

This circuit works great using those little brass piezo beeper disks - glued to anything that vibrates or gets whacked.

The output impedance of the piezo is high enough that the input protection network in the opamp eats any over-voltage without harm. If you have a giant piezo unit, it would be wise to add a diode from the piezo to Vcc to clamp the input.
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
I see on the datasheet that the Cmos MCP602x rail-to-rail opamps have built-in voltage clamping diodes on their inputs, then the diodes I added are not needed.
I put the rectifying diode inside the negative feedback loop to cancel its forward voltage and my negative feedback causes the opamp to produce a linear representation of the signal envelope and its amplitude, not just simple full blast squarewave pulses from the comparator outputs shown on the Piezo Trigger circuit.
 

Thread Starter

seayaker

Joined Jan 27, 2009
74
I see on the datasheet that the Cmos MCP602x rail-to-rail opamps have built-in voltage clamping diodes on their inputs, then the diodes I added are not needed.
I put the rectifying diode inside the negative feedback loop to cancel its forward voltage and my negative feedback causes the opamp to produce a linear representation of the signal envelope and its amplitude, not just simple full blast squarewave pulses from the comparator outputs shown on the Piezo Trigger circuit.
Thank you all for replying, I'm using 20 mm piezos going into a teensy 3.6 board which has the DA converters and sends velocity sensitive signals to the software. It works with pull down resisters and clamping diodes but when I put a diode in series,(either before or after the op amp) it kills the velocity sensitivity. It works as is, the only problem comes when I play fast or drum roll, It doesn't register every hit. For example if I play bobady, bobady, bobady, bomp, it will skip the second hit of each bobady phrase as if it's still reading the first hit. There are 21 pads so the circuit in post #7 would cost over $50 just for the MCP602s,
 
Last edited:

Audioguru

Joined Dec 20, 2007
11,248
At first you said that the envelope of the signal is important, like the piezo is a microphone or vibration sensor then you want to record and playback those sounds.
But now you say the signals are digital and you are trying the comparators circuit in post #7, with the 0.6V rectifier threshold to overcome and higher levels are always at full blast like guns, not drums.

You do not understand that a comparator simply switches its output on and off and is not linear to variable levels. You also do not understand that a rectifier diode does nothing until the voltage becomes 0.6V or more and you do not understand that when I put the diode inside the negative feedback loop of an opamp then that problem is cancelled.

Isn't $50 almost nothing compared to buying 21 drum pads?
 

Thread Starter

seayaker

Joined Jan 27, 2009
74
At first you said that the envelope of the signal is important, like the piezo is a microphone or vibration sensor then you want to record and playback those sounds.
But now you say the signals are digital and you are trying the comparators circuit in post #7, with the 0.6V rectifier threshold to overcome and higher levels are always at full blast like guns, not drums.

You do not understand that a comparator simply switches its output on and off and is not linear to variable levels. You also do not understand that a rectifier diode does nothing until the voltage becomes 0.6V or more and you do not understand that when I put the diode inside the negative feedback loop of an opamp then that problem is cancelled.

Isn't $50 almost nothing compared to buying 21 drum pads?
I don't know if you are confusing me with someone else or what but I never said anything about recording and playing back or anything else that's in your first paragraph. All I'm trying to do is make the piezo peak signal read more accurately.
 

Audioguru

Joined Dec 20, 2007
11,248
All I'm trying to do is make the piezo peak signal read more accurately.
Then you want the rectifier diode inside the opamp negative feedback loop like I showed. Then the amplitude of each peak will be exactly what is played. The circuit in post #7 switches its output to maximum (less the 0.6V diode voltage) or nothing which is not accurate.
 

Thread Starter

seayaker

Joined Jan 27, 2009
74
Then you want the rectifier diode inside the opamp negative feedback loop like I showed. Then the amplitude of each peak will be exactly what is played. The circuit in post #7 switches its output to maximum (less the 0.6V diode voltage) or nothing which is not accurate.
So you're saying I should try the circuit in post #5 ? if so which op amp? what resistors? All schottky diodes? You're right about all the things I don't understand, as I said in post #4. I have no background in electronics. This is the first and only circuit I ever tried to build. I don't have an assortment of components, I have a few things left over from my failed attempts but I have to order most of these. So if you think the circuit in post #5 would work could you please fill in the component values?
 

Audioguru

Joined Dec 20, 2007
11,248
I do not know what the piezo signal must do. You said the function is to send "velocity sensitive signals to the software". Your schematic shows a "percussive envelope" and you say you want the peaks to be accurate.

1) How can a drum beat have "velocity"? When you hit softly then the output amplitude is low and when you hit hard the output amplitude is high. Then the envelope is important since it will have low, high or in between amplitudes and a simple diode in series will throw away the first 0.6V of the amplitude so a precision rectifier circuit I showed is needed.
2) What is the maximum amplitude so that the amount of opamp gain will be known? Resistor values determine the amount of gain.
3) Do you hit a piezo sensor or do you hit whatever the piezo sensor is mounted on?
4) Does the piezo and its mounting resonate for each hit like real drum membrane and must the resonance need to be eliminated so one hit does not make many outputs?
5) Do you simply want to "trigger" the software to do something for each drum hit?
 

Thread Starter

seayaker

Joined Jan 27, 2009
74
I do not know what the piezo signal must do. You said the function is to send "velocity sensitive signals to the software". Your schematic shows a "percussive envelope" and you say you want the peaks to be accurate.

1) How can a drum beat have "velocity"? When you hit softly then the output amplitude is low and when you hit hard the output amplitude is high. Then the envelope is important since it will have low, high or in between amplitudes and a simple diode in series will throw away the first 0.6V of the amplitude so a precision rectifier circuit I showed is needed.
2) What is the maximum amplitude so that the amount of opamp gain will be known? Resistor values determine the amount of gain.
3) Do you hit a piezo sensor or do you hit whatever the piezo sensor is mounted on?
4) Does the piezo and its mounting resonate for each hit like real drum membrane and must the resonance need to be eliminated so one hit does not make many outputs?
5) Do you simply want to "trigger" the software to do something for each drum hit?
View attachment 174503 View attachment 174504

concept.jpg View attachment 174506

piezo_whack-300.png

The piezo simply sends a signal to the board which converts it to digital and sends a midi signal to the software, the midi has information about which note to trigger and at what velocity. The piezos are on hard plastic, it is not a reverb issue. All we're concerned with here is to get the peak and only the peak to register accurately. I have no way to measure the amplitude.
 

Audioguru

Joined Dec 20, 2007
11,248
Most of your attachments produce an error and do not work.

The opamp with a diode in its negative feedback loop rectifies the signal from the piezo and gives an accurate amplitude if the gain of the opamp is not too high which would cause clipping.

So I guess the piezo signal's amplitude is all you need, it triggers the digital hardware to make the drum beat with the correct frequency and amplitude.
 
Top