• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 29 subscribers
  • Views 1016 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.

TLV320AIC34 noise

Mads Dyrmann Larsen
Prodigy 70 points
Other Parts Discussed in Thread: TLV320AIC34, TPA3100D2

Hi,

I'm building a amplifier based on the TLV320AIC34 codec. The problem is that there are a lot of crosstalk - even if the codec has been resat, and all of the input/outputs therefore are deactivated. Let me give an exampel:

A cd-player with a 200mVrms output signal is connected to the Line1 input through a 1uF capacitor. The digital output is shortet to the digital input, which means, that I don't use an audio processor. A Power Amplifier (TPA3100D2) is connected to the Left Line output; again through a 1uF capacitor. Even though there should be no sound, there is a crackling sound in the loud passages of the music (still less than 200mVrms). If I touch the input instead of connection a cd-player to it, a loud humming is heard.

The problem seems to be the same on the evaluation board.

In adition to this problem, white noise is generated in the codec. When I set the codec in "power down bypass mode", the sound is nice an clear, but when I instead send the signal through the PGA->ADC->DAC, whitenoise is added to the signal. The PGA is set to 0dB. When raising the sampling frequency to 96kHz, the noise is minimized, but still loud enough to be irritating.

The power amplifier curcuit is well tested, and has no problems.

 Any idea on how to solve the problems?

 

Thank you

  • 0
    TI__Mastermind 28585 points

    Several things can be going on. There is a cross talk spec in the data sheet. It will be different depending on which channels are compared. As with most low-power, portable CODEC's, some are better than others so in cases where you will be looking closely at cross talk, you should select channels that give the best performance or use external switching if needed.

     

    Hum from touching inputs is probably 60 Hz noise being introduced to the ground plane. Using differential signal path can help with this, but in general you need to keep 60 Hz out of the ground and signal path with shielding or filtering.

     

    There are several signal paths that can be selected through the AIC34. On the DAC side, L1, R1 through the mixer will be the lowest noise. The direct path L3, R3 will be lower power but can introduce more noise and should only be used if power is the primary design consideration.

     

    When using delta sigma DAC's with Class D amplifiers, special care must be taken to remove out-of-band-noise. This noise would not usually be an issue since it is outside of the range of human hearing (above 20 kHz) but will cause noise in the audio band when input into a class D amp. You will need to filter the out of band noise to stop it from coupling with the class D amp.

     

    There are quite a few app notes on the TI web site dealing with each of these issues.

  • 0 in reply to Converter Apps
    Prodigy 70 points

    That is bad news.

    The 50Hz hum from touching the inputs still appear even though all of the inputs are muted.

    I am using the TLV320AIC34, because it's the only ic that offer six stereo inputs, microphone amplifier (using the pga), several outputs and treble/bass control. But does your answer mean, that you can't connect two playing sources at the same time, but only activate one input on the codec without hearing sound from the other inputl?

    I'm glad you are taking your time answering my questions.

  • 0 in reply to Converter Apps
    Prodigy 70 points

    By the way, the low-pass filter solved the white noise issue. Thanks

  • 0 in reply to Converter Apps
    Prodigy 70 points

    The hum is minimized by unmuting the output, but still there, if I turn up the volume of the DAC. The white noise is removed by a simple low-pass filter.

    But I still have problems with the sounds from the muted inputs, that is still heard. It appears at the peaks, which is measured up to 1,5V, even though the Vrms is less than 300V. The datasheet says that the maximum input signal level to the ADC is 0,707V at 0dB. but as the input is muted, I would say that the signal should be decreased by at least 70dB. It sounds like the channel seperation is far from -80dB to -98dB as described in the datasheet.

    As the problem also appear on the evaluation board, it seems like I have to use an external circuit to prevent this problem. One way to solve it, would be a voltage divider in each input. -3dB seems to be enough, but as it will decrese the signal-to-noise ratio I would like to know if there might be other solutions. Maybe an external switching circuit?

    The problem with a external switching circuit is, that it need to have 16 inputs (2 x stereo balanced + 4 x stereo signle ended).

    Please give your input on how to solve this problem!