Input frequency from microphones does not vary

[imperial-aero]

Hi,

I'm working on a project that involves generating signals to a loudspeaker and measuring them on microphones. After I've run my VI (attached), I now find that the analog inputs from the microphones have the same frequency throughout (50Hz, which is the mains frequency), regardless of what the frequency of the loudspeakers are (I'm expecting the frequency of the microphones to be similar to that of the loudspeakers).

I'm not sure what the problem might be. I think the microphones are working correctly as it does detect sound. They are normal tie-clip microphones and are battery charged. So I'm now wondering whether the problem might be in the method the DAQ devices were set up, or the input terminal configuration of the channels. I'm using a PCI 6229 device with 16 RSE input channels for the microphones.

Another problem which I've observed (different from the above) is that when I set my VI to save data, the loudspeakers makes a 'pop' sound every second or so. It does not produce this sound when I don't turn it on, which is very odd.

Thank you for any help on this.

[imperial-aero]

I've done the following test, and my conclusions regarding the problem is given as well.

- After changing my input terminal configuration of the VI to NRSE I found that one of the ai channels (ai8) was actually producing the correct frequency. The other channels produced no readings that were just fluctuations around 0.

- So I connected all of the 16 microphones, one by one, to this particular channel to check whether they gave the correct frequency readings. This did show that all the microphones were working. (when I did this test, I had disconnected all the channels except that channel I was testing on)

- Then using a function generator, I checked each and every ai channel separately and was surprised to find that all the channels did actually give the right signals. So, I came to the conclusion that the individual channels are working (when I did this test, I had disconnected all the channels except the channel I was testing on)

- However, when I connected one microphone to one of the channels, with all the channels left unconnected, it did not produce a signal, except the channel that was already shown to be working (ai8).

-Then I connected all the microphones up again and to my surprise I found that the previous ai channel (ai8) that was reading the correct signals, now did NOT. However, I found that another channel  (ai4) was now giving the right frequency readings, which was very strange.

- I disconnected one of the microphones that was connected to a channel that wasn't reading any signals and then I connected the function generator to that channel. The channel which the function generator was connected, and the channel (ai4) now both gave input signals.

So after all this, I've come to the conclusion that the microphones are working, the individual channels are working. The reason why I was obtaining 50 Hz was because somehow the input terminal configuration setting to RSE caused it to display the mains voltage on my VI.

But I can't figure out why all the microphones when connected together isn't working. And also, why is one channel working, and why does the channel that does work keep changing (i.e. in my test from ai8 to ai4). I'm afraid I don't know much about circuitry or how the devices were setup initially.

Any further suggestions will be very much appreciated.

Thank you.

[Tom O]

Hi imperial-aero,

It sounds odd that you're getting such inconsistent results from your different channels. I'd like to know a bit more about how you're connecting the microphones, to which pins etc. in case that's the problem. Also, when you test them, what are you using - your own program, or Measurement and Automation Explorer test panels? Finally, are you using the same grounding modes each time?

It sounds like these inconsistencies are to do with your hardware setup, so if we can troubleshoot and check that as much as possible we should be able to work out what's going wrong.

Regards,

[imperial-aero]

Thanks for the reply Tom. I just found that it was something wrong with the BNC Connectors I used to connect the microphones to the BNC-2090A terminal. It was very odd, because I found that if I put the microphone jack all the way into the pins, then I wasn't able to get any signals on the oscilloscope. But if I put it half way in it does give me some signals. So, I'm guessing it's something to do with the stereo and mono of the microphone. Now, I want to know whether there is some sort of BNC connector that will be suitale for my microphones. Also when I run my LabVIEW VI, I find that the microphone readings aren't the same (even though I expect them to b the same). Could it be an error in the microphones (since they are just normal tie-clip microphones)? I guess in that case I will need to find a method of calibrating them. Regards,

[Tom O]

Hi,

Glad you've made some headway with getting your signals in!

When you say the readings aren't all the same, how different are they? I guess depending on the microphone it will vary greatly according to its position in the room, the direction it is facing, etc. What are you doing to ensure that the readings should all be the same?

Regards,

[imperial-aero]

I have 8 microphones which are equally spaced around the circumference of a tube (at the same horizontal location). Infront of the microphones I have a set of loudspeakers (again 8 loudspeakers equally spaced around the circumference at the same horizontal location). Therefore, I expected them to have the same amplitude. I have attached a drawing of the setup.

WhenI obtain the readings from the microphoes, they vary quite a bit (I can even observe this on my input waveform graph).

I guess I'll have to do a more detailed check on the microphones (i.e. have them in the same location) and measure them, to be exactly sure whether they are giving different signals. I'm glad that, atleast, I know my LabVIEW code is working and that I'm able to get inputs and outputs. I now need to make sure that what I'm getting is accurate.

[Tom O]

Hi,

Thanks for the experiment render, it looks interesting!

It sounds like your NI hardware and software is set up fine, so your inaccuracies are likely to be with your setup or microphones (I imagine there are relatively broad manufacturing tolerances on tie-clip mics, for example). I hope you get it sorted out, let me know if there's anything else I can do to help!

Best regards,

[imperial-aero]

Hi Tom,

I've done more tests on the microphones to check whether their signals were actually different. I placed 8 microphones directly above a loudspeaker (forming a tight circle) to ensure that they read the same signals (very crude, but the differences due their locations should now be minimised). Then using an oscilloscope, I measured the peak-to-peak amplitde of each one of them. I found out that, with the exception of 2 microphones, all the rest had roughly the same readings.

I tried to obtain the same observation on my computer (using the VI I made). But now I find that, they all have different mean values, i.e. some of the graphs were vertically displaced (attached pic). I think if they had the same mean, then the graphs should be identical. Also, with the loudspeaker off, I still got this displacement in the readings (pic attached as well).

Do you have any idea why the graphs are displaced, whether it is a hardware or a software problem.

Also, I wanted to ask you what  units the amplitude on the waveforms graphs are? Are they the voltage reading? Because the amplitude obtained from LabVIEW were different to that of the oscilloscopes.

Thank you.

[Tom O]

Hi,

The offsets you are seeing could be caused by a couple of things:

• You have a grounding issue with the connections and grounding mode you are using, leading to changes in the relative potentials of each microphone. If you want me to check your connections, please draw a quick diagram of how they're connected.
  
• The microphones give different voltages when "neutral" due to differences in the mics or manufacturing tolerances.
• Differences in air pressure on each mic (I guess this is a bit of a long shot!).

If you change the mics around, do the offsets stay with the microphones or the channels? e.g. mic0 on chan0 gives 0.5V offset, mic1 on chan1 gives 1V offset; after swapping the two mics, do the offsets swap as well?

Finally, if you find that these inaccuracies are unavoidable, you could capacitively couple the microphones to the module with a large capacitor. This will block the DC offset of the microphones and only allow the AC signals through.

Regards,

[imperial-aero]

Thank you for the reply Tom.

After discussing the problem with my supervisor, I found out that the offset shouldn't affect my results as I'm just calculating the amplitude and phase spectrum using FFT. I reckon the offsets are because of the microphone tolerances more than anything else (since they are normal tie-clip microphones), so I'm going to leave it as it is.

However, I've got another issue with the connections. This time I'm getting noise from the signals. Here's a link to the problem I've asked on here (haven't had any respones yet).

http://forums.ni.com/ni/board/message?board.id=250&message.id=38253#M38253

I suspect that this is due to some grounding issues. The odd thing is, I'm getting the noise only when I connect two or more coaxial cables to my terminal block. I understand that it'd be pretty difficult to say what the exact cause of the problem is just by reading my post, but if you could tell me some test I could carry out to identify the problem, it'd be a great help.

Cheers.