Sampling rate affecting frequencies of piezo actuator

[andersne]

Hi, 

I have a system setup with a piezo actuator and amplifier, a accelerometer and power supply attached to a cantilever beam and NI USB-6251. The aim for the setup is to see how different frequencies generated by the actuator effects the cantilever and what input the accelerometer detect for a given frequency/amplitude. My problem is that it seams clear that the waveform generated is affected by the timing settings. For instance will the actuator produce two different frequencies when I change the sample rate for the system. This is validated both by the accelerometer and notable change in pitch (sound from when the system is running). I've used different kinds of LabView setups all having the same result. Tried different examples VI's like the Voltage - Continuous Output.vi and simultaneously augured signal from the accelerometer, also the figure 2 setup from https://www.ni.com/en-no/support/documentation/supplemental/06/timing-and-synchronization-features-o.... What's the reason for this problem and how can I avoid it?

System is tested with frequencies between 100-300Hz, 600Hz at the most. Sampling rate mainly between 1000-4000.

[GerdW]

Hi andersne,

---

@andersne wrote:

My problem is that it seams clear that the waveform generated is affected by the timing settings. For instance will the actuator produce two different frequencies when I change the sample rate for the system.

System is tested with frequencies between 100-300Hz, 600Hz at the most. Sampling rate mainly between 1000-4000.

---

For me it is NOT clear why changing the sample rate should influence the (sine?) signal frequency - as long as you obey Nyquist and Shannon… (Well, there is mechanical behaviour of your actuator, when it "jumps" from one sample (aka position value) to the next.)

Using sine signals of 600Hz with a sample rate of 1-4kHz is somewhere between "not obeying Nyquist" to "following Nyquist by the words, but using too less samples per period"! What about using a sample rate of >=10kHz? This would ensure atleast 16 samples per sine period (at 600Hz)…

---

@andersne wrote:
What's the reason for this problem and how can I avoid it?

---

As long as we don't see any code or measurement data we cannot help much…