• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 30 subscribers
  • Views 2330 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Shape of hp and sub pwm signals of TAS5706 and others

Hanns
Expert 1005 points
Other Parts Discussed in Thread: TAS5706, TAS5711, TAS5717, TAS5706B

Hello all,

I've been spending hours looking at the TAS5706 pwm outputs for headphone and subwoofer through a scope. I can't find anything that looks even remotely like audio in the same way as the speaker outputs do (which, I take it, are also pwm).

I have a couple of questions that I would be very happy if someone could answer for me:

1. What is the shape and levels of these pwm signals?
2. How do these signals differ in shape from the speaker outputs?
3. How can I convert these pwm signals to analogue audio?

Sorry for my ignorance in these matters. Thank you.

  • 0
    TI__Mastermind 33420 points

    Hello Hanns:

    1. What is the shape and levels of these pwm signals?

    3.3V @ 384 kHz carrier frequency with sampling rate of 48 kHz.
    2. How do these signals differ in shape from the speaker outputs?

    They are the same as the speaker outputs.
    3. How can I convert these pwm signals to analogue audio?

    We usually recommend RC LPF.  Please visit this web page for more information:  http://www.ti.com/lit/ug/slou220/slou220.pdf

    TAS5706 has been out for a while.  If you need a 2.1 ch, you can use TAS5711.  If you need a headphone or line-out you may want to consider TAS5717.  If you have to stay with TAS5706, please use TAS5706B.

    Hope this helps.

    Best regards,

    Tuan

  • 0 in reply to Tuan
    Expert 1005 points

    Hi Tuan and thank you for your reply.

    I am sorry for going on about this, but I need to make sure I understand you correcytly:

    1. I interpret this as: When feeding an i2s audio input with a sampling rate of 48 kHz, the output carrier will be 384 kHz - correct? On my scope, I found that signal coming out of sub_pwm+  (relative to pgnd), but I don't quite understand how the actual audio is represented upon the carrier - is it overloaded or what? In my scope the 384 kHz pulses were all equally high and wide and seemed perfectly periodic at 384 kHz.

    2. Great but when looking at the speker pwm outputs in my scope, I see a normal analog audio signal waveform. When compared to the above, I don't find them to be the same at all. What am I missing here?

    3. About the document, are you referring to an image like the one before the headphone amp on page 25 or do you mean something from schematics of the EVM? Could you please supply a more hands-on example of how to construct such a filter and where to put it with respect to both sub and headphone outputs of the TAS5706B?

    I'm not at all tied to the TAS5706B (sorry for not specifying the B), it's merely because I chose to evaluate that one since I need both headphone and sub and because its closed loop and accepts 192 kHz digital input. I will take a closer look at the TAS5711.

    Again thanks and best wishes,

    Hanns

  • 0 in reply to Hanns
    Expert 1005 points

    Sorry, maybe I should have mentioned that I'm working on a TAS5706B EVM board, so the speaker terminals mentioned are not chip bare bone.

  • 0 in reply to Hanns
    Expert 1005 points

    Hi all!

    I am interfacing directly with the EVM board so instead of going through the PSIA, I'm using a separate MCU to control the EVM board.

    I feed 16bit 192kHz audio via i2s.

    The chip on my EVM board is as TAS5706B.

    I have configured the EVM for 2.1 + headphones as can be seen below.

    I get sound through the regular speaker outputs, but nothing from headphones nor subwoofer outputs.

    When looking at the HPR, HPL, SUB- and SUB+ PWM outputs in my scope, I see nothing but a 50% dutycycle @ 384 kHz, i.e. dead silence:

    Scope probes are connected between HPR_PWM and PGND on the 2x8 header on the side of the EVM board.

    I have gone through the i2c-configuration several times and it all looks fine, here's the configuration i use:

    01	1B	00	Oscillator Trim
    01	03	A0	System Control Register 1
    01	04	05	Serial Data Interface Register
    01 05 24 System Control Register 2
    01	06	00	Soft Mute Register
    01	07	00	Master Volume Register (0xFF = Mute)
    01	08	30	Channel 1 Volume
    01	09	30	Channel 2 Volume
    01	0A	30	Channel 3 Volume
    01	0B	30	Channel 4 Volume
    01	0C	30	Channel 5 Volume
    01	0D	1C	Channel 6 Volume
    01	0E	91	Micro Register
    01	10	02	Modulation Limit
    01	18	0F	PWM Start Register
    01	19	00	Shutdown Group Resister
    01	1A	0A	Split Capacitor Charge Period	
    01	1C	02	Back-end Error Register
    00 01 (Below) Input Mux Register
    02 20 00 89 77 7A
    00 01 (Below) Downmix Register
    02 21 00 00 40 03
    00	01	(Below) AM Mode Register
    02	22	00	00	00	00
    00	01	(Below) Biquad1 Coeff
    02	23	00	80	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00
    00	01	(Below) Biquad2 Coeff
    02	24	00	80	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00	00
    00 01 (Below) PWM Output MUX Register (Note: Writes to this register affect Inter-Channel Delay)
    02 25 01 02 13 45
    00	01	(Below) 1/G
    02	26	00	80	00	00
    00	01	(Below) Scale = 1/(1-1/G)
    02	28	00	80	00	00
    01	11	4C	Inter-Channel Delay Channel 1
    01	12	34	Inter-Channel Delay Channel 2
    01	13	1C	Inter-Channel Delay Channel 3
    01	14	64	Inter-Channel Delay Channel 4
    01	15	D0	Inter-Channel Delay Channel 5
    01	16	90	Inter-Channel Delay Channel 6
    01	17	00	Offset Register (Absolute Delay)

    What am I doing wrong here?

    Thank you!

  • 0 in reply to Hanns
    Expert 1005 points

    I have now done a bit of reading and I now understand the nature of PWM audio signals. However, I'm still having trouble with the TAS5706 EVM board - please see this post instead for more details of the problem:

    http://e2e.ti.com/support/amplifiers/audio_amplifiers/f/6/t/166001.aspx

  • 0 in reply to Hanns
    TI__Mastermind 22965 points

    Hi Hans,

    I have merged your new post to the older one as they are both on the same device/topic. Please refrain from creating duplicate threads.

    To answer your question about whats needs to be done to play audio from the TAS5706 Head-phone outputs, you need to use a low-pass filter to filter the PWM modulated waveform and then pass this filtered signal along to an external head-phone amplifier (as shown in the application diagram on Page-3 of the Datasheet).

    This is in-fact no different from the speaker outputs, only that a passive-LC filter on the EVM is filtering the PWM outputs from the TAS5706 device. (you can try probing the pre-filter node to check this).

    -Ravi

    Audio Applications Engineer

  • 0 in reply to Ravi Singh
    Expert 1005 points

    Hello Ravi and thanks for your answer!

    The threads are not quite the same. This thread was about PWM signals, their shape and how to convert them to analogue. These questions have been answered.

    The second thread with a scope image attached (now merged into this thread) is about why I can't seem to get anything but 50% dutycycle pwm from headphone and subwoofer outputs. That question is yet unanswered. Given the I2C-configuration above, could you please help me figure out why I only get silence?

    Best wishes

  • 0 in reply to Hanns
    Expert 1005 points

    The great mystery has been solved!

    I finally discovered that volume had to be turned up quite a bit for anything noticeable to show on the non-amplified pwm outputs. I had to set volume in excess of 0dB in order to see any variations on the scope.

    Feel kind of stupid over here and sorry for wasting anybody's time on this.

    As for me however, I've learned something :-)