Piezo Voltages into a buffer - EDITED

[Kisen]

Hi,

Below is a circuit created to detect impact on a Piezo. Apologies for the lack of quality in the image, its a copy paste of the software im using.

The divider circuit on the left will centre my piezo voltage and when all components are configured correctly should give me an ADC output on the lower buffer of +/- 1.65v around the centre voltage of 1.65v - So 1.65v is my 0v when i think of it as AC.

The output of the upper buffer does the same at the moment and its output is fed to a comparator. What i would like to achieve is an output on the upper comparator of 0-3.3v (or 0-5v) So i need to shift the voltage back to 0v rather than it being halved (if that makes sense)

The reason for this is that the comparator is used as a trigger. With the compared voltage being either above or below 1.65v i can only capture the upper or lower AC wave. I want to trigger on the first rising or falling edge, but currently its anyones guess which edge will be seen first.

Can anyone help?

Thanks

 

[alec_t]

Welcome to ETO!
I can't make out the text on your pic. Can you post a sharper one?
When piezos are struck hard they can create high voltages. I think you'll need some input spike protection on your circuit.

 

[Kisen]

Hi,
Here is a clearer image of the piezo circuit. It has input spike protection using the 2 schottky diodes. Changing R2 and R17 allows me to set the amount by which the voltage is clamped.

But my question relates to the buffer.

The output of this buffer when the piezo is untouched is 1.65v - which is the same as 0v. As the piezo is hit, it fluctuates between 0v and 3.3v.
What i would like to do is move the 1.65v (zero) to 0v but still have both sides of the AC wave.

I am thinking that perhaps instead of doing this, a window comparator will be just as good.

 

[rjenkinsgb]

Capacitively couple the piezo output to the comparator circuit.

Use eg. 10K to 0V from both inputs, and a something like one megohm from negative input to positive power, to put a small offset between the inputs and set a trigger threshold.

Then use two schottky diodes, from the input capacitor to each comparator input - so the positive input sees a positive edge, and a negative edge goes to the negative input.

That should give you a positive pulse from either polarity signal.

----
However, from experience with piezo discs as sensors, the output will be several volts for even a moderate input.
I've used them with just load & series resistor and diode direct to PIC ADC inputs, no active processing at all.

I believe you could just feed the piezo, with a load resistor across it and protection diodes to supply & ground, to the input of such as a 74HC14 schmitt trigger as well as an ADC input, and the output of the HC14 to an input capture set to trigger on either edge.