REW cal procedure using external SPL calibrator and meas mics with known sensitivity

[tiki]

Hello all,
I use as measuring equipment:
- Microtech Gefell MK 221 (50mV/Pa) and MK 301 for ultrasound (4,5mV/Pa) with the appropriate preamp MV225P48
- RME Fireface UC (Line outputs 3/4, Mic inputs 1/2, gain 0dB)
- SPL calibrator for the mics, giving 94dB SPL at 1kHz sine
- Win10 PC
Is there a difference in the cal procedure in comparison to the usage of an external sound source driven by the soundcard?
If I try the calibration using the REW SPL meter, the log sweep measurement shows way higher levels afterwards, if the mic still is connected to the calibrator (~120dB to 94dB SPL).
Maybe the 1kHz sine is the wrong signal for the cal procedure, which seems to use a log sweep too?
May I "tell" REW the mic sensitivity directly?

Thank you for your help!
Cheers, Timo

 

[John Mulcahy]

SPL calibration doesn't use a sweep, perhaps you are mixing it up with soundcard calibration? Just attach the calibrator, go through the SPL calibration procedure and tell REW the SPL the calibrator was producing.

 

[tiki]

Hello John,

Thank you for the quick response!
The soundcard calibration is done, see the attached screenshot.
I tried this recommended SPL meter cal procedure several times with reproducible results. I did not use an REW/soundcard driven sound source but an independent SPL calibrator with a steady 1kHz sine tone instead (94dB SPL in a small sealed chamber, which fits to the mic capsule only), see photo.
The following measurements show an 1kHz-level of >130dB SPL, whereas it should be 94dB SPL.
The Piezobuzzer is only for checking the ultrasound capability of the mic.

What am I doing wrong?
Thank you very much for the great software!

Cheers,
Timo

 

[John Mulcahy]

Are you following these steps?

1. Open the SPL meter
2. Click Calibrate
3. Choose Using an external signal
4. Enter 94 dB as the SPL figure
5. Click Finished

 

[tiki]

Yes, I followed that (simple) order, described in the REW help. Before that I did the soundcard calibration.
At least I believe, I did it right.
If REW produces an output (if not sine, it may be noise?), which should drive the sound source, this procedure may fail, because REW should expect the acoustic power as wideband, not in a single frequency. But I do not know about REW's internals, so it's only a guess.
I am sorry for the effort!
I'll be back in 2 days, it's late already in Germany...
Thanks again!
Cheers, Timo

 

[John Mulcahy]

REW doesn't need to know the test signal, it is just calculating the rms level of the input and the factor needed to convert that to the entered SPL figure.

When you run the calibrator after SPL calibration what level does the REW SPL meter show? Not 94 dB?

Note that it does not make sense to make a sweep measurement of a calibrator signal as the response is not related to the stimulus, to observe the calibrator signal use the RTA in spectrum mode.

 

[tiki]

Hi,
aha, the last remark points to the same direction. So only the sweep measurement is not suitable.
I will try the RTA mode next time.
Thank you very much!
Cheers, Timo

 

[tiki]

Hello,
The RTA measurement seems to show the correct levels. The Measurement was conducted using different gains in the RME input of 0 dB and +55dB, respectively.



A hint regarding REW capture function:
If the comment grows "too big", the dialog elements (especially the lower buttons) run out of the box.
If possible, I'd like the dialog box to be variable in size by dragging the frame or corner.



Thank you very much for the help!
And thanks for the great software!

Cheer, Timo

 

[John Mulcahy]

A hint regarding REW capture function:
If the comment grows "too big", the dialog elements (especially the lower buttons) run out of the box.

I have changed it so that a scroll bar appears if the comments are longer than the box can accommodate.

 

[tiki]

Wow!
Thank you for the quick update!
Meanwhile I did some measurements on the transducer, see the attached pdf (in German), if interested.

The software works fine for this application, except for some discontinuities from time to time. It can be observed by watching the "signal" LEDs at the soundcard's front. The result is a drop (or some of them) in the spectrum, to be seen in the picture below, above 80kHz. Sometimes it helps to change the input/output device buffer's sizes.
Sometimes also REW seems to output short signal bursts after the measurement has ended. But it changes from measurement to measurement, so it's not clearly reproducible.
And I do not know, if it's related to the combination of REW with the RME-soundcard Fireface UC since I tried no other.



Thank you!
Cheers, Timo

 

[John Mulcahy]

If you are using the ASIO drivers make sure the buffer size is set to something large.

 

[tiki]

Thanks!
For these measurements I used the Java drivers with 32k/32k buffers for the in-/outputs.
Mostly there will be one usable measurement, if repeated.
Cheers, Timo

 

[John Mulcahy]

I would try with ASIO. I have run without problems at 768 kHz with an RME ADI-2 Fs Pro using the ASIO drivers.