LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

High microphone background noise using 9233/4472

I am doing sound measurement using PCB ICP microphone with preampifier and always get high background noise. The sound chamber backgound noise is about 20dBA and my measurement is close to 40 dBA when using NI 9233. I also tried NI 4472 and it reduced the background noise to 36 dBA which is still pretty high since my interested signal is around 42 dBA. Here is my system config:

 

Microphone: PCB 377A02, sensitivity 20mV/Pa

DAQ card: NI 9233/NI 4472

Computer: Dell Latitude with windows XP

 

My questions are:

 

1. What is the lowest signal level that 9233 or 4472 can measure?  For 9233, the input range is +-5V and dynamic range is 102dB so I assume the lowest input voltage 9233 can handle should be about 0.1mV. Is this correct?

 

2. For a 20 dbA signal, which is about 0.0002 Pa and Microphone output is about 0.004mV, can  9233/4472 handle such low level signal? I did not find any gain setting for both devices.

 

Thank you very much for your help.

 

0 Kudos
Message 1 of 8
(3,249 Views)

When you say the 9233 is .1 mv are you talking 10-7 v? The voltage coming out from the microphone .004 mv. Do you mean 4x 10-9 v?

 

The reason I ask, 9233 has 24 bit resolution and should resolve to 7x10-7 V. If you are talking about the output of the the microphone in millivolt, then the value you are expressing is 4x10-6 V,

 

If the latter is true, you have a range roughly of 10.

0 Kudos
Message 2 of 8
(3,241 Views)

0.1 mV means 10-4 V and 0.004mV means 4X10-6V in my original post.

 

The spec of 9233 shows its dynamic range is 102 dB which means we can not get 24 bit from that device 'cause 24 bit means about 140 dB.

 

Also according this post, " the smallest signal that you can see on the NI 4472 is 10 µV".

 

I am quite confused what exactly the smallest signal that 9233/4472 can handle. Did anyone successfully use those 2 cards to measure 20 dbA noise? Please share your experience.

 

Thank you very much.

0 Kudos
Message 3 of 8
(3,217 Views)

mrblue wrote:

1. What is the lowest signal level that 9233 or 4472 can measure?  For 9233, the input range is +-5V and dynamic range is 102dB so I assume the lowest input voltage 9233 can handle should be about 0.1mV. Is this correct?


I'm not quite what calculation you did here.  Because the 9233 is a 24 bit device, it is capable of detecting changes in the 0.6 µV range.  However, this does not mean that a given voltage measurement will be accurate to 0.6 µV.  One of the key considerations for any of our Dynamic Signal Acquisition devices is the difference between absolute accuracy and relative accuracy.  DSA devices are not designed for absolute accuracy.  That is to say, a 24 bit DSA device measuring a 5.000000V signal may return a value of 4.995000V.  However, because of the high relative accuracy of the device, each channel will show readings very close to the 4.995000V and when the reading changes the device will be able to measure that change very accurately.  So, if that initial reading of 4.995000V changes by 1 µV, the device will detect that change very accurately and return 4.995001 V.  This is the purpose of DSA devices and why they are used for sound and vibration applications, where relative accuracy is of more importance.  If you would like a rough absolute accuracy specification I'd refer you to this KnowledgeBase.

 

Both the 9233 and 4472 modules should be capable of measuring readings in the range you are testing.  I would recommend reviewing this article for issues with your noise floor.  You can find the specifications for the spurious free dynamic range on the specifications page for the 9233 and 4472.

Seth B.
Principal Test Engineer | National Instruments
Certified LabVIEW Architect
Certified TestStand Architect
0 Kudos
Message 4 of 8
(3,199 Views)

Thanks, Seth. Your note about absolute accuracy and relative accuracy is very helpful.

 

I want to clarify the calculation I did for 9233 since I am still confusing.

 From the definition of Dynamic range: 

Dynamic Range (dB) = 20 x log (Maximum Voltage / Minimum Voltage).

 I get the formula for Minimum Vlotage

Minimum Voltage = Maximum Voltage / ( 10^(Dynamic Range /20))

 

For 9233, the maximum voltage is 5v and Dynamic range is 102 dB. Therefore,

  

Minimum Voltage = 5/(10^(102/20))=3.97X10-5V=39.7 µV

   You mentioned 9233 is capable of detecting changes in the 0.6 µV range, this imply the

Dynamic Range (dB) = 20 x log (5 / (0.6X10-6))=138.4 dB

 Why in the datasheet of 9233 the dynamic range is only 102dB? Also, Why 4472 is 110dB and 4461 is 118dB? Those three cards all use 24 bit AD. Please advise.

Message Edited by mrblue on 12-15-2008 04:04 PM
0 Kudos
Message 5 of 8
(3,161 Views)

You are indeed correct, I was mistaken in my calculation in that I was only considering quantization error.  Let me get clarification on this issue and get back to you.

Seth B.
Principal Test Engineer | National Instruments
Certified LabVIEW Architect
Certified TestStand Architect
0 Kudos
Message 6 of 8
(3,140 Views)

I've gotten some clarification on this issue.  Your equation is mostly correct, however the maximum voltage should be 10V, because the 9233 has a ±5V range.  Thus, the full range is 10V.  As a result, the smallest signal that you will be able to detect with the 9233 is 79.4 µV.

 

The reason that the dynamic range is different for each of these 24 bit devices is that the noise floor of all these devices is higher than the step size at 24 bits.  At 24 bits, the smallest step size over a 10V range is 0.5 µV.  However, the noise floor of the device due to many other conditions is well above 0.5 µV.  As a result, resolution no longer dictates the smallest detectable signal; instead it is the noise floor of the device.  The noise floor is dependent on many things, including the circuitry on the board, your sampling rate, and the voltage range you are sampling in.  As a result, we specify the noise floor (which is also your dynamic range) based on what rate you are sampling at.

 

Does this answer your question?

Seth B.
Principal Test Engineer | National Instruments
Certified LabVIEW Architect
Certified TestStand Architect
0 Kudos
Message 7 of 8
(3,130 Views)
Seth, thanks for the clarification and it did answer my question.
0 Kudos
Message 8 of 8
(3,100 Views)