Hi there,

I have a waveform (producing by a piezo) and I need to accurately start recording relative to the initial wave arrival.

That is there is a flat floor (0V) then the wavefront rises as the wave arrives. Is there a way to electrically trigger off a point somewhere on the wavefront the is always consistent relative to the wave arriving. The problems is that the wave will have different amplitudes therefore using a trigger halfway up the wave would not be useful as relative to the wave arriving it would not help.

Regards,
Lance

 

Pick a DC trigger point that is just above the noise level. That way it will trigger at the initial arrival of the wavefront, fairly independent of its final voltage peak.

 

I have it triggering at 3mV but that still seems to be too variable.

 

What is the piezo detecting?

 

It is attached to a metal spike that is being hit by a hammer .

 

What is the "recording" hardware?

If you are doing it with a microcontroller, just "record" all the time, and analyse the signal. Then if the analysis shows it was a valid hit sound it is OK, because you were already recording when the sound happened.

Basically it means you start recording BEFORE the event, so you always record the event start.

It's a common technique.

 

Nope all analogue signals unfortunately.

 

Hi there,

I have a waveform (producing by a piezo) and I need to accurately start recording relative to the initial wave arrival.

That is there is a flat floor (0V) then the wavefront rises as the wave arrives. Is there a way to electrically trigger off a point somewhere on the wavefront the is always consistent relative to the wave arriving. The problems is that the wave will have different amplitudes therefore using a trigger halfway up the wave would not be useful as relative to the wave arriving it would not help.

Regards,
Lance

If you can trigger at just above the noise floor, it should work fairly consistently. Some scopes have some really intelligent triggering, but it takes a while to figure it all out.

I take it this is not a piezo oscillator but some kind of impact detector?

 

@KL7AJ

I would agree with you, the operative word being "should". You are correct is is used to measure the start point in time for a speed of sound calculation.

I have attached two graphs showing my current setup (an inverting op-amp with 1000 gain acting as a comparator).

The output goes into a p-channel mosfet, and I was measuring from this gate drive voltage to ground. The mosfet has a turn on voltage of -1.5 volts, therefore taking this as the switch point and comparing the wave arrival point relative the -2V on the trigger gives the delay time between wave and trigger. Evidently the delay is not consistent for these three waves, and that is the problem I was trying to solve.

If I had a microcontroller then I would run a FIR filter to exaggerate the wave arrival (FIR filter would create a dip due to the sharp rising face), and measure the time between the dip and the trigger point then relay this information somehow. Unforuntately this particular circuit needs to be super low power (5 year life on 2xAA batteries) so that method is out of the question.

Potentially the problem is the circuit that is driving the trigger, ie it is still too analogue and needs to be digital, I am not sure.

Regards,
Lance