• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 30 subscribers
  • Views 169 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPA6120A2: TPA6120A2 channel leakage when driving differential inputs with single-ended PGA2311 output

Falguni Sharma
Prodigy 190 points

Part Number: TPA6120A2
Other Parts Discussed in Thread: PGA2311

I am using TPA6120A2 as a headphone amplifier.
The audio source is PGA2311, which provides single-ended outputs.

The TPA6120A2 has fully differential inputs, but in my current design I am feeding it as follows:

  • PGA2311 ROUT → TPA6120A2 IN_R+

  • IN_R− → GND

  • PGA2311 LOUT → TPA6120A2 IN_L+

  • IN_L− → GND

So effectively, I am driving the TPA6120A2 in single-ended mode by grounding the negative inputs.

Problem:

When I give a command or signal to one channel (say Left), I observe audio leakage / crosstalk into the other channel (Right), especially noticeable at higher gain or volume levels.

My questions:

  1. Is it correct to connect IN± negative pins directly to GND when using a single-ended source with TPA6120A2?

  2. Can grounding IN− be the root cause of channel leakage / crosstalk?

  3. What is the recommended input configuration for using TPA6120A2 with a single-ended volume control IC like PGA2311?

Best regards

Falguni Sharma

  • 0
    TI__Mastermind 20330 points

    Hi Falguni,

    Can you share the schematic for your design? I am not certain I understand your circuit when you mention grounding the negative pins. Crosstalk performance is a somewhat complicated subject as it depends on the layout as well as the performance of the actual device. We do have some example circuits in the datasheet for how to configure the device. If you are driving each output single ended a standard non-inverting circuit configuration should work.

    Best Regards,

    Ignacio

  • 0 in reply to Ignacio Vazquez Lam1
    Prodigy 190 points



    This is the schematic from my TPA6120a2 section . Im feeding output signal LOUT and ROUT from PGA2311 to the Lin+ and Rin+ of TPA6120A2 . In the negative side input of TPA6120a2 section im connecting to GND . If this configuration is not correct will you please tell me how i have to connect single ended input from PGA2311 to the TPA6120A2 section.

  • 0 in reply to Falguni Sharma
    TI__Mastermind 20330 points

    Hi Falguni,

    I have looked at this design and did not find anything wrong with the circuit. Are you using the EVM for our device or is it your own PCB layout? Crosstalk performance depends on many variables. It does depend on PCB layout, but it is also inherently due to the fact that the amplifiers are in the same package. I am attaching an article discussing this topic that gives good information on some of the major sources of crosstalk.

    Reducing crosstalk of an op amp on a PCB

    Best Regards,

    Ignacio

  • 0 in reply to Ignacio Vazquez Lam1
    Prodigy 190 points

    We are actually sending audio signal to TPA6120  and then to  headphone output. Our application requires to have sound in only one channel when command is set to play tone in that respective channel only. (for ex left channel)
    There should be no  sound in the other channel.(Right channel)
    But the problem is that there is  small tone of around 2 to 20 mVrms. This becomes hearable when sound is above 95 db Spl.
    And it gets really audible around 35 db Spl to 45 db Spl at 120dB spl in other channel.How can we resolve this issue ?

    We are currently using our own PCB with schematic and layout guidelines from TPA6120A2 EVM user guide.

    Best regards 

    Falguni 

  • 0 in reply to Falguni Sharma
    TI__Mastermind 20330 points

    Hi Falguni,

    Could you share the layout for the PCB you are using? Could you also share what input level is being sent to the active channel that is resulting in 20mVrms in the other channel?

    Best regards,

    Ignacio

  • 0 in reply to Ignacio Vazquez Lam1
    Prodigy 190 points

    We are using same layout as given in EVM board.
    We have tried to separate  volume control IC PGA2311  for left and right channel along with two TPA6120A2 IC for both channels. Still we are facing the issue of leakage. What might be the problem for this.
    Will using daisy chaining method will be helpful to connect two different channels separately ?

    Is this possible that  any problem in the grounding side of the PCB  setups  can lead to this leakage problem?

    We are getting 20 mVrms at around 2 Vrms.

    Best regards 

    Falguni

  • 0 in reply to Falguni Sharma
    TI__Mastermind 20330 points

    Hi Falguni,

    Can you clarify the setup you used in your second statement. Did you drive one channel into one TPA6120A2 and you saw some effects on a different TPA6120A2 on the board? A drawing of what you did in this experiment will certainly be helpful. Can you also explain what you mean by daisy chaining the circuit. A circuit diagram will also be helpful to fully understand the setup you are considering. This information will be helpful to bring to our team to see if they can offer additional info on what could be causing this leakage.

    Best Regards,

    Ignacio