• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 756 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.

Problems using the PCM3052A

Alejandro Uriz1
Prodigy 30 points

Other Parts Discussed in Thread: PCM3052A

I am working with the codec PCM3052A and have some problems to operate the DAC.

First, the system is connected to a codec dsPIC33EP256MU806 through I2C and I2S buses.

In another application than previously realized, but using a dsPIC with more pins, the system operated correctly.

The system clock is a crystal (F = 4.194304 MHz) , which in the codec results in a sampling frequency of 16384Hz. The I2C bus is working correctly, because I can adjust the codec parameters .
The I2S bus also seems to be working properly, and that the waveform, amplitude and frequency of the signals are according to expectations .

The circuit used to implement the codec is suggested on page 41 of the datasheet it. The only change I made to him is that because it 's only going to use the line inputs , did not include R1, R2 C9 and C10 , and MINP minm leaving open. Note that the line input is selected by software .

The problem I have is that the DAC output is not working properly . Specifically , there is a 1 Vpp noise therein.

If  a send a 24 bits test signal generated within the DSP, it seems that since the bit 23 to bit 16 , the signal becomes well . From bit 15 to bit 8 noise is generated and is the 8LSB's are  well converted .

A test that I did was to send the zero signal level from the DSP codec and check the appearance of the noise , however if I turn off the codec ( / PWRDN = 0) the noise disappears . So in principle I have ruled out the problem was external to the circuit.

Notably mounted three units have this prototype and 3 have the same fault .

Also, I checked the settings , levels and frequencies of the most important signals with those of the previous prototype , which works properly and there are no differences .

To verify that no noise was due to a noise problem at the source, I checked the waveforms , and have added 10uF capacitors next to the power pins of the codec.

I would therefore like to know if possible if this issue can estarse owing to lack of components and minm MINP .

I would appreciate any information that allows me to determine the origins of the problem or the solution .

 Thank you very much.

Alejandro

  • 0
    TI__Guru**** 189401 points

    Hi, Alejandro,

    I can think of two areas to investigate:

    1. Since you have mic inputs floating, are they picking up noise?

    2. Do you have the i2c settings correct to not sum the mic inputs into the audio stream?

    -d2

  • 0 in reply to Don Dapkus
    TI__Guru**** 189401 points

    Hi, Alejandro,

    Sorry, I hit post too quickly...

    Also, check to make sure that you don't have a ground loop in your system which may be inserting the noise.

    -d2

  • 0 in reply to Don Dapkus
    Prodigy 30 points

    Hi Don,

    Thanks for the reply . I've done some testing based on them .

    First, to analyze whether the problem was caused by the disconnected input microphone, using the I2C buis I sent POWER SAVE command to ADC. ( ADPSV bit register 67 ) . This should turn off the ADC. Interference on a signal sent from the DSP is maintained .
    To verify that the I2C working correctly , I sent POWER SAVE command to DAC and as expected, the interference and the test signal disappeared.

    This led me to think that the problem is in the DAC , it tried to verify several configurations:
    - I Enable and disable the de- emphasis filter , the test signal and the interference continued to exist .
    - I Inverted phase output signal of the DAC , the test signal and the interference continued to exist.

    - I sent my MUTE SOFT ON command (bits MUT1 MUT2 registry and 68) and the test signal disappeared , but the interference was maintained.

    Regarding the PCB Layout , I've had the design is very similar to working prototype . I have divided the dough flat digital analog ( codec ) and ( DSP ) by a 2.2 Ohm resistor and have been careful not to generate induction loops . The only significant change is that the models have joined the AGND1 , AGND2 , AGND3 and DGND pin to plane below the codec mass plane in this model .

    Anyway, I will analyze the PCB with the EMI expert of the team .

    I hope to be close to solving the problem , hopefully with these measurements is possible to detect the problem. If you have news , the notify by this route .

  • 0 in reply to Don Dapkus
    Prodigy 30 points

    Dear Don,

    I "solved" the problem by using a more abrupt output LPF .

    The 1 Vpp noise previously described  was directly measured on the output of the codec. Therefore, after reading the following work: http://www.ti.com/lit/an/sbaa055/sbaa055.pdf , I observed the noisy signal in the spectrum analyzer and the measured noise , despite seem excessive in the time domain , conformed to the expected limits.

    Likewise , the power consumption despite being higher than in the previous prototype codec , also is within the allowable limit.

    Therefore, I used at the output of the codec a LPF with lower cutoff frequency and higher order , and the noise was removed in the output signal .

    Regarding the source of the interference , we found that BCK bit off , the noise disappears. Also,  the interference has the same frequency as the BCK signal , we noticed that we were both related effects .

    We contemplate that for this prototype the problem was solved . But for future prototypes of the system we want to know if we have any particular consideration in order to minimize this interference.

    Thank you very much .

    Best regards,

    Alejandro Uriz