• State TI Thinks Resolved
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 31 subscribers
  • Views 212 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TAS2563: Speaker Characterization

Rajesh Katikam
Prodigy 40 points
Part Number: TAS2563

Tool/software:

Hi,

 I am using the TAS2x63 EVM for speaker characterization. we are using the QFN package. I know that the speaker is working and the mic is working. But when I try speaker characterization, it always fails at the hardware checks. Surprisingly it says audio playback failed. Which is not true since in the next step of IV measurement, I can clearly hear the  sweep. Though hardware checks fail IV measurement and hence model fittings works fine. Pls see the attached images. The IV measurement plots a graph and the peak is observed at the resonant frequency given by the speaker manufacturer.

When I look at the PPC3 user guide, in the 2017 version(https://www.ti.com/lit/an/slaa757/slaa757.pdf) as well as the user guide video (https://www.ti.com/video/5595585518001) here, they talk about SPL measurement which is the acoustic characterization. A mic is being used to measure this. But the TAS2563 QFN package software doesnt show any SPL measurement. Also the user guide(https://www.ti.com/lit/an/slaa765/slaa765.pdf) no more talks about SPL.  Is that no more needed or is there now a different way to measure the SPL? What is the purpose of the mic that is connected to the learning board if not for the SPL measurement? Right now I am using an Iphone to capture the Sweep signal(20Hz to 20KHz) being played by the speaker. I tuned the equalizer to in such a way that the waveform shape is preserved  upto 4KHz. Now my suspicion is, the Iphone mic may not be good enough to capture the frequencies beyond 5-6 KHz and hence the need for the mic. 

My queries  are 

1. Is the hardware check failure causing the  SPL measurement not to show up?

2. Or SPL measurement not supported by the TAS2563 QFN? Then what is the alternate way to measure it and use it in the PPC3?

3. How do we make use of the Mic connected to the LB2 to capture the sweep being played out by the speaker. Will this capture beyond 5-6KHz properly

4. Also in my temperature measurement, I get a waveform as shown in the image, which is not correct. What could be the issue here?

5. In the bare metal code example provided, there are some sample configurations for speech, music etc. These are generated using PPC3 tools I guess. Hence  can these cfg files be loaded into the PPC3 and visualize the settings used to generate those cfgs?

Regards

Rajesh

 

  • 0
    TI__Guru** 108975 points

    Hi Rajesh,

    Addressing your questions below:

    1. No, the hardware checks won't prevent other sections to show. The issue you describe with the hardware checks can sometimes be due to some instability in the USB driver, if you're able to complete IV-sense and other tests you may ignore the error in hardware checks.
    2. SPL measurement is no longer used. The "smart EQ" feature has been discontinued for newer devices, and the microphone is not required for any of the characterization steps.
    3. Just for SPL capture you may connect the microphone directly to your PC or use any other USB microphone. Some microphone vendors provide frequency response of their products for reference.
    4. I assume after some time you get an error saying that the thermal characterization couldn't be completed and the speaker was not able to heat up. Is that correct? In that case you may have to use a lower frequency for the test tone, usually 3kHz works better.
    5. PPC3 won't update the GUI controls with cfg files, only ppc3 extension files can do that.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • 0 in reply to Ivan Salazar
    Prodigy 40 points

    Thanks Ivan for the quick reply.

    1. With 3KHz as the excitation frequency, I was able to get the thermal characterization. Is this inline with what is expected?

    2. I started tuning the equalizer for my speaker. I plugged in a calibrated mic (dayton) to my laptop to capture a linear sweep signal (20Hz to 20KHz with a rectangular shape) being played by the speaker. I used the sound recorder app on Windows11. What I see is there is no consistency in terms of the waveform captured. At times it shows lot of attenuation. I expect the recorded signal waveform to have  a flat shape (though not perfect) up to some high frequency like 16KHz. The  attached image is my expectation. Correct me If I am wrong.

    3. I feel there could be multiple limitations here  

    • The speaker itself may not be able to playout very high frequencies properly
    • The calibrated mic may not capture high frequencies properly
    • The sound card used to plugin the mic
    • The software used also might not do a good job

    How do I narrow down which could be a limitation?

  • 0 in reply to Rajesh Katikam
    TI__Guru** 108975 points

    Hi Rajesh,

    1. The results from thermal characterization seem fine.
    2. Do you have a data sheet for your speaker by any chance? Usually, speaker vendors provide a SPL vs frequency plot which would help to judge if your recording is correct. The frequency response of the speakers is not flat, and if you're playing the sweep at full volume there may be some additional non-linearities due to speaker protection. In this case I'd suggest playing the sweep at somewhat lower volume to avoid the protection from kicking in. Alternatively, you can configure the device in ROM mode to remove all the integrated processing. Note that speaker protection is preventing over-excursion and over-temperature on the speaker which may affect some frequencies more than others.
    3. Getting a speaker SPL response from data sheet would greatly help here.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • 0 in reply to Ivan Salazar
    Prodigy 40 points

    Hi Ivan,

    I have tuned the 10 band equalizer. I play a linear sweep through the PPC tool. And then capture the sweep with a calibrated mic. The calibrated mic has a pretty much  a flat response. I captured the sweep  multiple captures in succession. What  I find  odd is that the waveform doesnt look like a sweep after a few iterations. It degrades pretty badly after the 7KHz. The distance between the mic and speaker and all other settings remain the same. Its just that I play and capture the sweep multiple times. Any ideas on what could be happening here

  • 0 in reply to Rajesh Katikam
    TI__Guru** 108975 points

    Hi Rajesh,

    Can you share the wav files of the recording so I can analyze on my side?

    Do you mean that playing the same sweep several times with the same configuration, the audio output from the speaker changes as the sweep repeats?
    Is this something you catch only with the microphone or is it also perceivable by listening?

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • 0 in reply to Ivan Salazar
    TI__Guru** 108975 points

    Hi Rajesh,

    We'll follow up over email, we can come back to the forum if needed as long as the thread is not locked due to inactivity.

    Best regards,
    -Ivan Salazar
    Applications Engineer

  • 0 in reply to Rajesh Katikam
    TI__Intellectual 1010 points

    Hi Rajesh,

    If the sweep degrades pretty badly after the 7KHz(high frequency region), one possibility is that speaker is heated up and temperature protection is triggered to attenuate high frequency signals. Thanks.