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.

OPA172: application note TIDU765 high distortion

Part Number: OPA172

Hello Team,

My customer uses the preamp circuit based on TI's application note TIDU765. This application details how to uses a TI OPA172 opamp as an audio preamp for a condenser microphone capsule. They implemented the design, and it works, but there is high distortion in the circuit. Can you give some guidance on minimizing distortion?

Regards,

Renan

  • Renan,

    What kind of distortion? [ Obvious visible on oscilloscope (such as clipping), THD, Noise, Oscillation, Power supply noise ]

    The most obvious improvement would be to filter the VCC to the MIC. Reducing bandwidth is another way.

    Can you show this distortion?

  • Hello Ron,

    Please see attached captures of distortion measured at the preamp output.

    distortion.zip

    Thank you in advance.

    Regards,

    Renan

  • Renan,

    The OPA172 portion will be the least contributor to the noise and THD. If you tested this section alone (see figure 15 in TIDU765), you should see low distortion like figure 17. Broadband noise down -120dB, and nearly indiscernible 3rd and 5th harmonics. Look very closely at figure 17 and you can see them.

    The MIC and R1 (pull up resistor) are another story.  The JFET gain introduces the 2nd harmonic. The resistor doesn't add distortion directly, but the voltage drop becomes a VDS for the MIC JFET so that could add to odd harmonics. JFET gain also adds to broadband noise.

    Electret Microphones are easy to use because of the large output signal. The OPA172 only amplifies by -5.5 V/V (15dB). However they are not the lowest distortion microphone option.  

    About the zip file data, am I correct that the CH1 waveform drove a speaker (sound waves) then the MIC picked up the sound?  Does the second harmonic percentage drop when the MIC input is reduced? (Just curious)

  • Hello Ron,

    Thank you for your feedback.

    I have attached the schematic snippet of our implementation of the preamp circuit referenced in TIDU765 for your review. We experimented with different values for R1 (in our design R224) and found that increasing the resistor value from 5.6K to 10K reduces the THD from 0.687% (5.6K) to 0.224% (with 10K). So, what is the best way to bias the capsule? See measurement captures in the attachment; also attached is the mic capsule datasheet.

    attachment(snippet).zip

    Regards,

    Renan

  • Hi Renan,

    Did the customer follow the TIDU765 guide and how to calculate the values of the external components? When I start calculating component values such as R2 I end up getting different values than what was present on the customer's circuit.

    For R2, If I assume they want a Vout of 1.228V, I get a value for R2 = 52.57kohms! Perhaps they are wanting a different output voltage? Still very different from the 6.65kohm value they went with! 

    Best Regards, 

    Robert Clifton 

  • Hello Robert,

    Thank you for this response.

    I have attached my customer's preamp calculation spreadsheet for your review; the component values calculated in the 9.7mm tab are used in the design

    attachemnts.zip

    Regards,

    Renan

  • Hi Renan,

    I'm still reviewing the data you have sent over but I have a few observations/questions: 

    • What exactly is the AP sending it's -8.9dBu (1.228 Vrms) signal to? 
    • There's also a few other assumptions that were made on the design calculator that don't appear to have happened in the system they tested. They calculated the output voltage of the LDO to be 4.42V but then measured 4.27V!
    • I saw that the MIC current seemed to be measuring only the DC component of the system. That current seems to be a bit lower than what the customer calculated, which is likely what's messing with the bias voltage.
    • What is the power supply used for the OPA172?  

    Best Regards,

    Robert Clifton

  • Hey Renan,

    Please let us know if your question has been resolved. I will be marking the above answer as TI thinks resolved and you may confirm the solution or reject and continue the thread.

    Best,
    Jerry