• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 30 subscribers
  • Views 234 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.

OPA1632: The output THD+N is high

Cherry Zhou
Mastermind 22235 points
Part Number: OPA1632

Hi Team,

The audio DAC is designed using the OPA1632 to design the LPF circuit, which is placed after the IV circuit as shown in the following figure:

The circuit is simulated and has no problem, but practical testing found that one channel THD+N is significantly higher than the other(about 8 dB, which can be said to be very large.) when testing the THD+N with an AP analyzer.

The traces on the audio output on the PCB board are long and the degraded channel traces are longer than the quality channel. The strange thing is that the THD+N has improved by more than 10dB by touching the audio output line of a degraded channel with your hand. The THD+N will become degraded again when released.

The customer assumes that touching with hands changes the capacitance between audio transmission lines. But why is that? The filter limits the bandwidth of the OPA1632 to less than 100K, but is so sensitive to the inter-transmission line capacitance.

In addition, different filter parameters produce different results, such as between the head and end of the transmission line, where only touching a particular segment of the transmission line affects the THD. This position will change when the filter parameters are changed. Also, the input impedance of the AP analyzer also affects the effectiveness of the touch position.

Could you help check this case? Thanks.

Best Regards,

Cherry

  • 0
    TI__Genius 9170 points

    Hi Cherry,

    We do see similar responses in the lab when we touch and move around the cable going into equipment. The reason has mostly been due to an older damaged cable. When you change the cable do you still see the changes when touching the cable. Would you be able to also probe the output and see the signal and what it is doing when you do touch it versus what it does when it is not being touched? Is it also possible to shorten the cable from the board to the AP analyzer, you did mention the path to the analyzer is long so it would be good to see if the path length is playing a part in this.

    Best Regards,

    Ignacio

  • 0 in reply to Ignacio Vazquez Lam1
    TI__Mastermind 22235 points

    Hi Ignacio,

    Thank you for the support.

    Ignacio Vazquez Lam1 said:
    When you change the cable do you still see the changes when touching the cable.

    Yes, changing cables has no significant improvement for the issue.

    Ignacio Vazquez Lam1 said:
    Would you be able to also probe the output and see the signal and what it is doing when you do touch it versus what it does when it is not being touched?

    Could you help elaborate on what output and signal are you referring to?

    Ignacio Vazquez Lam1 said:
    Is it also possible to shorten the cable from the board to the AP analyzer,

    The customer also tried, but there is no obvious changes.

    Touching the PCB trace that runs from the OPA1632 output to the capacitive coupling to the XLR output has the greatest effect.

    Also, customers have tried to cover PCB traces with copper foil, drip several drops of water on the traces, and so on, all of which have a positive impact on their THD.

    These measures actually change the capacitance between P and N, and some of the relationship between this capacitance and the inductance of the PCB trace results in a change in the THD.

    Thanks and regards,

    Cherry

  • 0 in reply to Cherry Zhou
    TI__Genius 9170 points

    Hi Cherry,

    I apologize for the confusion I was curious if the customer was able to probe the signal right at the output of the device and see if the signal changes in anyway as the wire is touched. However, after seeing the length of the trace from the device to the output of the board, it is a good assumption that some form of capacitance is coupled on these long traces resulting in a change in the signal. The same method to ruling out this hypothesis would be to try to expose and cut the output traces and try to solder connectors from these exposed traces that are now closer to the device, effectively shortening the path from the original 20cm. This is not the best way since it will result in damaging the PCB, but the concern is the long traces and the impact it has on the signal chain. 

    Best Regards,

    Ignacio