Acoustic output of AO0 via USB-6216 BNC is 8.4 dB softer (compared to PCI 6221)

[Henrik_Volkers]

Repeat the measurements with the amplifier output.  (the AO output is harder to describe with its current limit etc.)

Your speaker ist not just a resistor 🙂  look up the complex impedance of speakers .. ask the manufactor of your device if they can provide some data...

this complex impedance interact with your power source (impedance) ...

one reason for the voltage peak is the inductivity of your speaker.

(in a simple model with an ideal source it also should show a negative peak at the falling slope I can't see in your figure)

or could be that the build in AO output amp shortly can provide some more current, or ....

and don't expect that the air pressure follow the voltage 🙂

In a simple model it will follow the current in the coil (in addition to a lot of other static and dynamic parameters).

I would try to measure the current and the voltage with the scope. put a 100 Ohm (more or less) in series (between AO, resp. the amp output and your speaker) and measure the voltage with the scope on both ends of the resistor. the voltage on the speaker side is your exitation voltage and the voltage difference together with the resistor value is your current . I think the scope can do the math for you if configured rigth.

test some sine frequencies ...  

[Fuh]

Thank you so much for the explanations. Greatly appreciated!!!

Even though I am unable to fully understand all the details, they make sense, due to the inductive and dynamic properties of the inset earphone and the entire system.

To determine the maximal voltage that can be used without too much distortion on the waveform, I did some measurements in my lab this afternoon. The setup is:

• A series of pulses (pulse width = 100 microseconds) generated by a custom-written LabView program >>
• Signal goes through channel Ao0 of PCI 6221 >>
• Signal goes to a 50-ohm ER-3A insert earphone

For clarity, I presented a series of pulses (pulse width = 100 microseconds). I observed that the waveform morphology of these single pulses were all perfect (or nearly perfect) until the voltage goes beyond 5 Vpk. For clarity, I took a picture here at 5 Vpk. The oscilloscope was connected, via a BNC T connector, to the Ao0 channel of PCI 6221 (when the 50-ohm insert earphone was also connected).

Does this mean that I can use this setup up to 5 Vpk (due to the conductive and dynamic properties of the earphone and the entire system), without adding a buffer circuit?

Certainly, if I need to use a buffer circuit, I am happy to learn and make that happen. However, given the measurements above, will I be lucky enough so that I can just use this setup (as long as that I limit the voltage to be within 5 Vpk for single pulses)?

Correct?

[santo_13]

When you're testing acoustic properties why use a pulse?

Typically all acoustic waves are sines (or a mix of sines), you need to test it by generating a sine wave.

[Fuh]

That is a good point. Thank you so much for the comments.

I did some additional measurements in my lab, by using this setup:

1. NI Measurement & Automation software >> Test Panel >> Analog Output tab >> this software generates continuous sine wave at a specific frequency to Ao0 channel of PCI 6221
2. Sine wave went through the Ao0 channel of PCI 6221
3. Sine wave went to the 50-ohm insert earphone

To determine the maximal voltage that can be used without too much distortion, I presented sine waves at different frequencies that are important to my research. The oscilloscope was connected, via a BNC T connector, to the Ao0 channel on PCI 6221 (when a 50-ohm insert earphone was also connected). The maximal voltage that can be used without visually identifiable distortions (judged by using my eyeballs) for each frequency is listed here.

10 Hz: 3.54 Vrms, 10 Vpk-pk

20 Hz: 4.13 Vrms, 11.8 Vpk-pk

50 Hz: 4.35 Vrms, 12.3 Vpk-pk

100 Hz: 4.39 Vrms, 12.4 Vpk-pk

200 Hz: 4.39 Vrms, 12.4 Vpk-pk

400 Hz: 4.39 Vrms, 12.4 Vpk-pk

500 Hz: 4.39 Vrms, 12.4 Vpk-pk

606 Hz: 4.39 Vrms, 12.4 Vpk-pk

For clarity, I took a picture of 100 Hz sine wave at 4.39 Vrms here.

Does that mean that I can use this setup (as long as that I limit the voltage output to be within the maximal voltages for those frequencies)?

Correct?

[santo_13]

If your oscilloscope supports THD measurement, please ensure that at the maximum drive amplitude that can ensure 0.1% THD.

[Henrik_Volkers]

Why do you need such high levels?

Again the headphone has specked

102.5 dB SPL in HA-2 coupler at 0.2 Vrms (50 Ohms)

at ~4Vrms this ends up in ~130 db SPL

How many dB do you want to send into that poor ear?

Can you measure the soundpressure at the end of your setup ?

If you provide that pulse I'm shure you end up with 'funny' pulse shape mainly determined by the volume and it's shape driven. But I'm not into acoustics, so a well designed setup migth end up in what you want (what we don't know).

[Fuh · Accepted solution]

I usually present acoustic sounds between 60 and 85 dB SPL, calibrated by using a 2 c.c. coupler and a sound level meter (SLM).

The purpose of all these measurements is to evaluate (1) whether the new setup in my lab is legit, and if so, (2) what is the maximal dB SPL that can be produced without significant distortion.

This afternoon, I did the following measurements, by using this setup:

1. NI Measurement & Automation software >> Test Panel >> Analog Output tab >> this software generates continuous sine wave at a specific frequency to Ao0 channel of PCI 6221
2. Sine wave went through the Ao0 channel of PCI 6221
3. Sine wave went to the 50-ohm insert earphone
4. Acoustic sound went into a 2 c.c. coupler and a SLM
5. Sound pressure level was measured via a SLM (Extech Instruments, dB C, fast)

To determine the maximal voltage (that can be used without too much distortion) and its associated air pressure of the acoustic sounds coming out from the 50-ohm insert earphone, I presented sine waves at different frequencies that are important to my research. The oscilloscope was connected, via a BNC T connector, to the Ao0 channel on PCI 6221 (when a 50-ohm insert earphone was also connected). The maximal voltage and its associated dB SPL values for each frequency are listed here.

10 Hz: 3.54 Vrms, 85.6 dB SPL 

20 Hz: 4.13 Vrms, 100.4 dB SPL

50 Hz: 4.35 Vrms, 109.2 dB SPL

100 Hz: 4.39 Vrms, 111.2 dB SPL

200 Hz: 4.39 Vrms, 111.5 dB SPL

400 Hz: 4.39 Vrms, 111.2 dB SPL

500 Hz: 4.39 Vrms, 111.2 dB SPL

606 Hz: 4.39 Vrms, 111.3 dB SPL

For clarity, I took a picture of a 100 Hz pure tone at 111.2 dB SPL here.

Please note that, this Extech SLM does not allow me to make frequency specific (i.e., 1/3 octave bandwidth) measurements. My industrial grade SLM (Larson Davis SLM system) was broken recently. I am currently in the process of acquiring a new Larson Davis SLM system. Once I have received the new SLM system (likely in 2 weeks), I will be able to make frequency specific measurements for the acoustic outputs.

As you folks have already pointed out, due to the conductive and dynamic properties of the insert earphone and the entire system, the voltage readings on the oscilloscope likely will not transfer (at least not on a linear scale) to the air pressure readings in dB SPL. I totally agree.

To determine the relationship between the voltage readings on the oscilloscope and dB SPL readings on the SLM, the above-mentioned setup was used (with the sine-wave frequency fixed at 100 Hz, i.e., the most important frequency in my research). The correspondence between the voltage and dB SPL readings are listed here.

100 Hz: 4.39 Vrms, 111.2 dB SPL

100 Hz: 3.50 Vrms, 110.8 dB SPL

100 Hz: 2.81 Vrms, 110.2 dB SPL

100 Hz: 1.40 Vrms, 100.5 dB SPL

100 Hz: 0.70 Vrms, 92.3 dB SPL

100 Hz: 0.351 Vrms, 85.5 dB SPL

100 Hz: 0.245 Vrms, 81.9 dB SPL

100 Hz: 0.200 Vrms, 80.1 dB SPL

100 Hz: 0.176 Vrms, 79.0 dB SPL

100 Hz: 0.105 Vrms, 73.8 dB SPL

100 Hz: 0.070 Vrms, 70.4 dB SPL

100 Hz: 0.050 Vrms, 67.2 dB SPL

100 Hz: 0.035 Vrms, 64.3 dB SPL

100 Hz: 0.010 Vrms, 60.0 dB SPL

According to the specifications of this 50-ohm ER-3A insert earphone, it should produce 102.5 dB SPL in a 2 c.c. coupler at 0.2 Vrms. However, by using this new setup in my lab, it produced 80.1 dB SPL. There is a difference of 22.4 dB. This difference could be due to (1) the Extech SLM is unable to measure dB SPL at a 1/3 octave bandwidth and (2) the Extech SLM does not provide a "dB flat" weighting network, and its closet weighting network is "dB C". At 100 Hz, there is a dB drop created by the "dB C" weighting network. Nevertheless, for these two possible issues, I should be able to address them when my new Larson Davis SLM system arrives in about 2 weeks.

It now circles back to my original question: If I limit the maximal voltage to be within these limits for those frequencies, is this new setup in my lab legit?

Any comments and suggestions will be greatly appreciated!!!