Sound Pressure axis scaling in Flexlogger

[Arvind_Auto]

Hello All,

I have been attempting to obtain sound level measurements using a PCB microphone and an NI DAQ setup (cDAQ-9184 with a sound and vibration input module, NI-9231). 

Flexlogger is able to detect the sensor with the right sensitivity values through the TEDS functionality.

I tried measuring a known sound pressure value using a sound calibrator emanating 94dB @ 1kHz however the frequency spectrum graph is not scaling to the expected dB value. I am attaching a snapshot.

The output value in dB is chosen in the y-axis tab. I am unable to find any other instance where I can edit any setting to scale appropriately.

Require help in figuring this out.

Thanks!

[esattern]

I'm not an expert in sound measurements but does the graph you attached look correct other than the amplitude?

An alternative you could try is adding channels to calculate the RMS and convert to dBs so you can control the scaling rather than relying on the graph to do them. You should be able to add a statistics channel to get the RMS and a calculated channel to get dB (see attached image). Then, you could map that calculated channel to the graph and keep the same frequency analysis settings you configured earlier.

Regards,

Eric - FlexLogger Product Owner

[Arvind_Auto]

Thanks Eric!

I got the right value on the calculated channel.

However, it doesn't seem to map to the spectrum graph though. The live value displayed is correct but it doesn't show up on the frequency spectrum graph. 

Any pointers on where to look?

Thanks,

Arvind

[esattern]

Glad the calculated channels worked!

Regarding the graph: my first guess thought be to double check that the filter settings on the graph match what you're expecting. E.g. I'm not sure how clean the signal is from your calibrator but maybe there's a better windowing type or you need to increase the frequency resolution? Are you still using the same settings as you posted in your initial screenshot? One of my coworkers linked me to this document that outlines some of the differences between the options we have: https://www.ni.com/en-us/support/documentation/supplemental/06/software-filtering--windowing---gener...

Next guess: maybe there's a software bug and the channel didn't get mapped correctly. If you delete the graph and add a new one to the screen, does that resolve it? If not, what about reloading FlexLogger? If either of those do work, I'd definitely want to know so we can figure out how to reproduce and fix it.

I'll ask around to see if I can find someone with more sound and vibration experience than me to take a look at this too.

[Arvind_Auto]

Hello Eric,

Thanks for your reply. I tried deleting the graph and opening new ones as well. I put a low pass filter in the calculations tab as well with a smaller block size and the time domain (high speed graph) seems to be appropriate. The frequency spectrum graph still eludes me. I tried different filters, resolution and I don't see the x-axis moving from the range it displays currently. Let me know if there is anything else that I could try.

Thanks and Regards,

Arvind

[Arvind_Auto]

Hello Eric,

Were you able to check on your end?

Thanks,

Arvind