Problems with sampling rate for PCB pressure sensor 105C13

[SOMA76]

I want to measure sound pressure inside a vessel containg water (ultrasound source is a t bottom) I am using ICP sensor from PCB 105C13 (specification sheet attached) with external power supply. I am using Labview 9 and PCI-6111E DAQ board.

For a sound input of 40KHz and 20 W I want to measure the rms pressure at different points in the vessel. I tried to measure the effect of sample amount and sampling rate on the rms pressure calculated. But as shown in teh attached sheet as soon as I increase teh sampling rate the rms pressure also increases. So it is difficult for me to get an accurate pressure.

Kindly guide me if someone has any idea.

[AndreasStark]

Hi,

since the output impedance of the sensor is small, there shouldn't be a problem with faster sample rates.

Did you select the right voltage range in your task? Refer to http://www.ni.com/pdf/manuals/370980a.pdf page 4 to see the available voltage ranges.

Maybe the 6111 is just not the right hardware. Did yuo check out the  NI Acoustic and Vibration Data Logger  at http://sine.ni.com/nips/cds/view/p/lang/en/nid/207807 containing the NI-USB 9234 Module.

Another Link: What is IEPE and is it the same as Integrated Circuit Piezoelectric (ICP)?

Best Reagrds,

[Henrik_Volkers]

How is your sensor powered and connected to your DAQ?

[josborne]

That is very strange.  My first thought is that you are seeing the effects of "aliasing".  Perhaps because you aren't sampliing fast enough.  But that probably isn't correct because your source signal is only 20kHz, and you are sampling way beyond 2x that frequency.  So I'm stumped.  Maybe you need a sound and vibration expert to comment on this.

If you find the solution, please post it.  I'd be curious to know what is going on.

[johnsold]

Where do you see a pressure change with sampling rate?  The graph you showed appears to have a constant sampling rate but a varying number of samples per reading.  Those are very different parameters.  Also it is not clear just what you are averaging.

Can you show some actual data?  I suspect that you may have some transient effect in the first few samples.  As the number of samples per reading increases, the effect on the average of the transient becomes less significant.  How do the first 5-10 samples compare with the next 2000 samples?

Lynn

[josborne]

I was confused by this, too.  Until I realized there are 2 pages in the PDF. 

Page 2 shows the changes with sampling rate.

[johnsold]

Thanks for pointing out the second page.

I still want to see some raw data.  I think that will be more informative than plots of averages over sampling parameters.

Lynn

[SOMA76]

Dear Lynn,

                 On page 2 of pdf you will see the effect of sampling rate. Since with the sound we have consecutively compression (positive pressure) and rarefaction (negative pressure). So in order to find the pressure at a given point we calculate rms (root mean square).

[josborne]

Quick question:  Why are you only taking 10 samples at each sampling rate?  At such high sampling frequencies, this means you are only sampling for a period of a couple microseconds or less.

EDIT:  Nevermind.  I misunderstood the text on your graph.  Looks like you are taking 25000 samples at each frequency.  Doing that 10x and then averaging?

[johnsold]

Are you seeing resonance effects? The specifications indicate that the resonance is > 250 kHz.  What is the bandwidth of the system? 

Does the pressure vessel have a resonance in the same range?

Lynn