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.

TPA3131D2: Device Selection for a 4ohm 10W amplifier

Part Number: TPA3131D2
Other Parts Discussed in Thread: PCM1792, PCM1789, TPA3132D2, TPA6211A1

Hi All some questions,

  1. We have two speakers (left and right) and would like to drive these independently each at a max 10W per channel (with controllable gain) speaker impedance 4ohm can someone help with a device selection for the Audio amplifier, Class D will suffice for this application.
  2. The preceded Audio Chain I'm trying to decide on which DAC to drive the amplifier looking at both the PCM1789 or the PCM1792 some advice on this would also be appreciated with any link to reference designs.
  3. On the same system we also need an output into a 8ohm load at 5VPk-Pk which is required to be either single ended or differential (universal) with controllable gain

Many Thanks

  • Hello Nigel! Have you looked at the newer TPA3136 www.ti.com/.../tpa3136d2, or perhaps one of our digital input devices like the TAS5780, where you would not need the PCM device www.ti.com/.../tas5780m ? Otherwise, Shawn would be happy to help you with your questions. Thanks, Jeff
  • Hi Jeff,

    Have looked at these today I think that the TPA3132D2 will suffice with a max 16VDC swing that would make the Plimit ~4V with a gain the lowest I can specify, i need to use the DAC  for other reasons, my main concern really is how the PCM1789 data sheet takes the diff output of the DAC then filters and before going into a buffer to make the output single ended

    what I really want is a diff output for the DAC feeding into the Amplifier for left and right speakers, I could make the circuit above into single ended for VoutL+and VoutL- not sure though this would be the correct approach. So what I was contemplating is taking the PCM1789 outputs into a single ended LPF buffer then into the Amp for both the + and - do you think this is wise? Also bothe left and right speakers need to be independent of each other, so the question is if I run both speakers from the  same device i.e. the TPA3132D2 would this classify as independent?

    I have also looked at the line driver chips that have similar circuits in, like the DRV6x. Next question if I was to take the diff outputs into a single ended could I run the single ended outputs into the amplifier ?

    Thanks for the help !

    Nigel

  • Hi Nigel,

    I can't see the circuit which you shown in the above. You can click 'Use rich format' to insert picture. You can also upload it as attachment.

    TPA3132D2 can support 16V/4Ohm. Both differential-end and single-end mode input work for TPA3132D2, but as you know the differential-end mode is even a little better in noise, pop...

    From your description, two speakers are used as the output on TPA3132D2, and only one input for both L-channel and R-channel, right?  I will comment after I see your circuit.

  • Hi Shawn,

    Apologies for that , here are a few configurations appreciate your comments and advise on each, configuration 1 which shows Single ended buffered outputs of the PCM going into both channels of the TPA3132D2 with the one side of the AMP diff input pulled to ground - therefore PCM1789 VOUTL + and - feed the Left speaker and VOUTR + and - feed the right speaker.

    Configuration 2 shows that both sets of Diff outs on the PCM1789 go into a single input of the Amplifier which is connected in Mono Mode:

    Config 3 would buffer the diff outputs of the DAC through single buffers to retain the differential through to the Amplifier:

    Hope this helps you understand my dilemma...

    Thanks for the responses so far.

  • Actually one other configuration that could potentially be implemented is to rely on the input buffer of the amplifier and therefore take the diff outputs of the DAC through a Capacitor to the Amp input buffers:

    I would just need to check voltage levels from the DAC to the input of the AMP on first glance looks like I may have to do some attenuation as the DAC produces 8Vpp differential out and the TPA3132D2 as an input V of -0.3 to 6.3VDC, apologies for this just thinking out loud.

  • Hi Nigel,

    Thanks for your explanation. In Configuration 1, the audio amplifier is using single-ended input mode, and the output is stereo mode. In Configuration 2, the audio amplifier is working in PBTL mode. In this way, a lager output power on single speaker can be achieved. But two amplifiers are needed for stereo mode(two-channel outputs). In Configuration 3, differential-ended input mode is used on the audio amplifier. In this way, the common noise and pop noise are minimized. In Configuration 4, this also a common solution. But the DAC output is too large. You can try to decrease the DAC output range or attenuate the input signal for the audio amplifier.

    From the previous description, you want to achieve 2*10W output power on 4Ohm loads, right? I think 1 piece of TPA3132D2 can do that, no need to use 2 for this application. Please use higher than 10V power supply to avoid clipping.

    Best regards,

    Shawn Zheng

  • Hi Shawn,

    Thanks for confirmation of my configurations looking at the voltage levels A: from the DAC datasheet suggest that they are 4Vpp on each diff pair

    Then if I read the data sheet correct for the TPA3132D2 for INPL for instance it says max input voltage -0.3 to 6.3V therefore I believe I can assume that this should be fine,

    so in light of this can you recommend a solution from one of my previous configurations ?

    I think I would prefer to go with configuration 4 as this seems the easier approach, however because the DAC produces increased  out-of-band noise above the Nyquist frequency to improve in-band signal-to-noise ratio should I still filter before going into the Amplifier to achieve optimal converter performance or can I rely on the internal DAC low pass filtering for sufficient filtering ?

    Thanks again,

    Nigel

    Nigel

  • Hi Nigel,

    One thing needs to be noticed is cipping on the output of amplifier. The output range of DAC is -2V~2V, then the output of the amplifier is -20V~20V if 20dB gain is used. Then you need to use 22V power supply at least to maker sure the linearity on the output.  The maxmum output is 20Vp on the output of the audio amplifier, and the output power is very high in this case(50W for one channel for 4ohm load). The thermal is a challenge. The right way is to calculate the output range on DAC by the expected power on speaker and gain setting on the amplifier.  

    I think Configuration 4 is a good choice if the input range on the amplifier is appropriate. As you said, a low pass filter is necessary for the out-band noise supression. A RC low pass filter can work for this.

  • Hi Shawn,

    Thanks for the advice here is my final schematic:

    I will let you know how the project goes but really appreciate the help!!

    Nigel

  • Hi Nigel,

    Generally it looks OK. Please use 220uF+100nF+1nF for the PVCC(pin27,28), and use another group(220uF+100nF+1nF)for PVCC(pin13, 14). The two groups of power supply decoupling are for channel L and channel R.

    The rated output power is 10W on 4ohm load per channel. So the peak value of output voltage is around 9V. And from your SCH, 20dB gain is used. Then the input peak value should be less than 0.9V to avoid clipping on the output signal. Please use the appropriate input signal amplitude.

    Best regards,

    Shawn Zheng 

  • Thanks Shawn your update, I was also looking at using the TPA6211A1 for an output to a 3.125W amplifier into 8 ohm however this only appears to go to ~1.3W im looking for around 3.12W do you have any alternatives again its a diff input of the same Voltage ?

  • Hi Negel,
    I'm sorry I'm not an expert in TPA6211A1, becaues it's a lowe output power device. Could you please post a new question in E2E with TPA6211A1 in title? Then TI application expert in this device will give you professional answer. Thanks!
    Best regards,
    Shawn Zheng
  • Thanks Shawn,

    Done!