Why does NI-9233 with IEPE microphone report 39dBA as minimum noise level?

[JDfromRI]

Hi,

I have been using an ACO Pacific 1/2" IEPE microphone and preamp (7052SYS) with cDAQ and a 9233 IEPE module. The ACO mic is very similar to the G.R.A.S. 46AE (50mV/Pa) but with a 20mV/Pa sensitivity. Using this set-up in a hemi-anechoic chamber, the SPL is reported as about 39dBA (using the Sound and Vibration standard  'sound meter' VI). Yet, the third party lab's signal analyzer reports the room at about 25dBA. I have partially verified the NI set-up with a 114dB 1kHz calibrator. The NI set-up indicates about 114.3dBA. So we are on the money there.

Is there a noise floor on the 9233 that I am dealing with? Any thoughts? Recently, my company has built its own hemi-anechoic chamber and we like the NI tools. We had hoped to use the newer 9234 with a cDAQ chassis since the 51.2kS/sec is sufficient for our measurement range of 20-20kHz. We already use cDAQ extensively here. 

Thanks,

-JD

[dsb@NI · Accepted solution]

From the 9233 specs:

Integrating Noise density:
A-Weighted Power = ( A weighting power factor for bandwidth ) * (Bandwidth) * ( Noise Level )^2
A-Weighted Power = ( 0.53 ) * ( 20000 - 20 Hz ) * ( 400 nV / sqrt(Hz) / 20 mV / Pa )^2
A-Weighted Power = 4.26u Pa^2 rms
Leq (A-Weighted) = 2.06m Pa rms = 40.2 dB(A) ref 20u Pa

Calculating from Noise at full scale range:
FS = Voltage Range / Sensitivity
FS = ( 5 / sqrt(2) ) V rms / ( 20 mV / Pa ) = 176.8 Pa rms = 138.9 dB ref 20u Pa
Noise = 138.9 dB ref 20u Pa - 95 dB FS - 2.73 dB(A) = 41.2 dB(A) ref 20u Pa

So the fact that you are reading 39 dB(A) indicates that your 9233 noise is meeting specifications. There is still an issue that some effort needs to be made to better match the signal range to the input range of the 9233. You can add some external signal conditioning to apply pregain (20 dB) to the microphone output before the 9233. You also have other DSA hardware options that include input gain.

[JDfromRI]

Doug,

Excellent post. Thank you for the detailed reply; this is essentially what I figured the situation was. Can you suggest how we might add 20dB pregain before the 9233/9234?

Thanks,

-JD

[Kurt_Veggeberg]

It depends on how many channels you want to measure in your anechoic chamber. GRAS Sound and Vibration offer a 1/2" preamplifier that will work with a 1/2" pre-polarized microphone capsule and provide up to +20 dB gain. It requires at least 3 mA of current and each channel of the 9233/9234 provides only 2 mA.   You have to tie two channels on the 9233/9234 to provide enough current which is not ideal.     If you plan on making Sound Power measurements in your anechoic chamber, you will probably need 8+ channels.

You might want to look at moving to the PXIe-449X series dynamic signal acquisition modules for making measurements in your anechoic chamber.   These modules offer more dynamic range (113 dB vs 102 dB) as well as gain settings for the same basic price per channel as the 923X modules and and program the same way.    These come in 8 - 16 ch. versions per module.      With these, you would be capable of measuring down to the noise floor of the microphone (probably about 20 dbA).

Here is a case study about using these modules in an anechoic chamber.    

http://sine.ni.com/cs/app/doc/p/id/cs-13433

Kurt Veggeberg

Business Development Manager,  Sound and Vibration

kurt.veggeberg@ni.com

[Parthiban_SM]

Hi Doug

DSA hardware options link in the post was not accessible. Could you please provide correct one

-Parthiban  

[dsb@NI]

Here is a link that works today:
Sound and Vibration - DSA Hardware Options

Note, this link takes you to the landing page for all sound and vibration hardware options. From there you are welcome to select your platform of choice and investigate available modules.

[Parthiban_SM]

Thanks Doug. I have visited this web page before. the reason I asked here is, I thought you might had specify the hardware which is measure below 40 dBA in the previous link. My application requirement is to measure the noise from 20dBA to 120dBA. Please suggest the suitable hardware for this application from starting range

-Parthiban   

[dsb@NI]

Any of the NI Sound and Vibration modules could measure SPL from 20..120 dBA when perfectly matched to a high-quality microphone. Practically, it is challenging to find a perfect match for a specific application, so we have to see if we can match the expected output range of a microphone to the input range of the acquisition module.

First, let's assume you have acquired a microphone that meets your application requirements. For example, your low noise and high range requirements motivate a low-noise, high-sensitivity microphone (such as https://www.pcb.com/sensors-for-test-measurement/acoustics/specialty-microphones/low-noise). For this microphone, the inherent noise is 5.5 dBA and the nominal sensitivity is 450 mV/Pa. Given these specs, for your application (20..120 dBA), the lowest signal output from the microphone is ~90µ Vrms and the maximum signal output from the microphone is ~9 V rms. 

Let's take a look at the PXIe 4464 as a possible fit for your application. 
At the ± 10 V pk range:
Idle-channel noise <= 12.9µ Vrms

Full-scale range = 7.07 Vrms (sine input)

Effective measurement range (mic + acquisition channel) = 5.5 dB (limited by mic) to 117.9 dB (limited by 4464)

At the ± 31.6 V pk range:
Idle-channel noise <= 108µ Vrms

Full-scale range = 22.4 Vrms (sine input)

Effective measurement range (mic + acquisition channel) = 21.6 dB (limited by 4464) to 127.9 dB (limited by 4464)

Typically, the idle-channel noise is better than the published specification, but it may not be your policy to make purchasing decisions on typical specs.

Looking at a different microphone (https://roga-instruments.com/downloads/PDF/MM222-EN.pdf). For this microphone, the inherent noise is 15 dB and the nominal sensitivity is 50 mV/Pa. For your application (20..120 dBA)

The minimum output signal is 5.6µ Vrms and the output at 120 dB SPL is 1 V rms.

At the ± 3.16 V pk range for the 4464:
Idle-channel noise <= 4.3µ Vrms

Full-scale range = 2.24 Vrms (sine input)

Effective measurement range (mic + acquisition channel) = 15 dB (limited by mic) to 124 dB (limited by 4464)

[Parthiban_SM]

Thanks for the illustration Doug, As a beginner I am learning something new from each of your messages 

-Parthiban