• State TI Thinks Resolved
  • Locked Locked
  • Replies 20 replies
  • Answers 1 answer
  • Subscribers 29 subscribers
  • Views 3332 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.

LM4808: Popping Sound

Icy
Intellectual 265 points
Part Number: LM4808
Other Parts Discussed in Thread: TPA6132A2, TPA6205A1, OPA1612, LM2662

I'm using two LM4808 ICs in my audio mixer circuit. After I power the circuit (3V with 2 x AA batteries), I hear a popping sound when my headphones are connected to the outputs before about five seconds have elapsed. If I wait at least five seconds after power is applied, I don't hear a popping sound when I connect and disconnect my headphones. I followed the recommendations to reduce pop in the datasheet by adjusting the values of capacitors Ci and Cb, but it made no difference. I reduced the input capacitors (Ci) to 0.1 uF and it had no effect. It also had no effect to reduce my output decoupling capacitance by removing two 470 uF capacitors on each channel. I don't know why I have to wait so long after applying power before there's no pop sound. What could be causing this problem? Here is my circuit:

Note: the labels below the ICs should be LM4808...

  • 0
    TI__Guru** 104275 points
    Hi Icy,

    One thing I would recommend is to use a voltage divider (bias voltage) for each device. From the schematic above I see that one of the dividers is connected to non-inverting input B from both U2 and U9. The same goes for input A.
    What I would recommend is to use one divider for both input A and B from U2 and the other one for input A and B from U9. Or another thing to try is to use only one voltage divider.
    Perhaps these bias voltages are ramping different and it takes these ~5sec to be stable.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points
    I tried using only one voltage divider and I still hear a popping sound.
  • 0 in reply to Icy
    TI__Guru** 104275 points
    Have you tried by placing series capacitors at the outputs of the first stage? Perhaps the DC component is affecting the second stage input.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points

    I did place a 1 uF capacitor at the output of the first stage (for each op amp A and B) and it didn't make a difference.

  • 0 in reply to Icy
    TI__Guru** 104275 points
    Icy,

    The turn-on time should be related to some, or many, of these capacitors and resistors. For debugging I would follow some questions and steps:
    - What is the purpose of all those output coupling capacitors? Why four in parallel for each output instead of one?
    - If R32 and R33 are used to make the discharge time of output coupling capacitors faster, place them between device outputs and coupling capacitors, in parallel with internal 20kOhm resistors.
    - Both LM4808 stages seems to be unity gain. What is the purpose of having two stages if they're the same?
    - Is it possible to test pop performance at the first stage output?
    - Is it possible to test with a different supply voltage source for comparison?
    - Is it possible to use an oscilloscope to inspect several nodes in the circuit and try to identify which ones are taking long time to become stable.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points
    -The purpose of having four output coupling capacitors at each output was to get a larger capacitance (there wasn't a single suitably sized capacitor with this large of a capacitance), to satisfy a low frequency cut off. I removed two capacitors on each channel and it made no difference.
    - Ok, I'll try placing the 1k discharge resistors at the device outputs instead.
    - The purpose of the first stage is to mix the two audio input signals together (on each channel), and the purpose of the second stage is to de-invert the outputs from the first stage. It's just an application requirement to have the output signals not inverted from their original phase.
    - I would have to get board rework done to test the performance at the first stage output. I know I need both stages though.
    -I have tested with both two AA batteries and a 3.3V DC/DC step up converter
    -The supply voltage looks stable on the scope nearly instantly. I probed at the amplifier outputs and they also didn't look unstable after turn on.

    Thank you for the suggestions.
  • 0 in reply to Icy
    TI__Guru** 104275 points
    Icy,

    Have you tried by placing the discharge resistors at the device outputs?

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points
    Yes, I tried that and it didn't make a difference.
  • 0 in reply to Ivan Salazar
    Intellectual 265 points

    Can you recommend a substitute part that would have less potential for popping and still fit the application?

  • 0 in reply to Icy
    TI__Guru** 104275 points
    If this is a headphone application (LM4808 is listed as HP amp) you could take a look at our DirectPath devices, TPA61xxA2 family. These devices use an internal charge pump to generate a negative voltage rail, so that the output is centered at GND and there is no need to place output coupling capacitors.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points

    This is for an audio mixer so I need to use op amps in the summing amplifier configuration. 

  • 0 in reply to Icy
    Guru 140200 points
    Hi Icy,

    why not splitting the 3V from the two AA-batteries into a +1.5V and -1.5V supply?

    Kai
  • 0 in reply to kai klaas69
    Intellectual 265 points

    My application will require a single AA battery that is boosted to 3.3V with a step-up converter. For testing, it made no difference in regards to the popping sound whether I used the 3.3V booster or two AA batteries. 

  • 0 in reply to Icy
    Guru 140200 points
    Hi Icy,

    I didn't mean a 3V single supply but a +/-1.5V dual supply. This would allow to omit the voltage dividers arround R2, R6, R7 and R10. The dual supply must have an audible effect on the popping noise, because all the AC coupling caps must not longer charge to Vcc/2!

    Kai
  • 0 in reply to Ivan Salazar
    Intellectual 265 points

    Hi Ivan, 

    How can I use a device in the TPA61xxA2 family like the TPA6132A2 if I'm creating a mixer using the summing amplifier configuration? Is it true that you can't use a negative voltage rail with the LM4808?

  • 0 in reply to Icy
    TI__Guru** 104275 points
    Icy,

    You could use the same configuration as your original post. You can also see this summing configuration in TPA6205A1 data sheet Figure 30 (page 16); you would need to consider two channels as TPA6205A1 is a mono amp.
    LM4808 is single supply only. For dual supply you could use a device with charge-pump or an op amp (most op amps support dual supply operation).

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points

    Hi Ivan,

    Regarding the TPA6205A1, where does it indicate in the datasheet that it can use a dual supply? The application circuits are showing a single supply so I'm confused about that.

    Do you have a recommendation for a stereo amp so I don't have to use two TPA6205A1s? Also do you have a recommendation for an inverting charge-pump IC to use with it? 

    Are op amps normally not used to output directly to Class AB amplifiers? Why would you use "Class AB audio amplifiers" instead of op amps? 

    I did some searching and found the audio op amp OPA1612, which has an output drive capability of ±30 mA, which I think should be sufficient. I'm thinking of using the OPA1612 with the LM2662 for the negative supply. Does that make sense? 

    Thank you!

  • 0 in reply to Icy
    TI__Guru** 104275 points
    Icy,

    I used TPA6205A1 as a reference only, this is a mono, single supply differential amplifier. But is shows the summing input configuration in its data sheet.
    What I'm trying to say is that you can use TPA6132A2 with the summing input configuration. This device has the charge-pump integrated.
    Regarding the last comment, it seems to makes sense.

    Best regards,
    -Ivan Salazar
    Applications Engineer - Low Power Audio & Actuators
  • 0 in reply to Ivan Salazar
    Intellectual 265 points

    How can the TPA6132A2 be used with the summing input configuration, when it has four pre-determined selectable gains? The block diagram on Page 10 shows the amplifier feedback within the chip, unlike in the TPA6205A1 where the amplifier connections are accessible outside the chip. 

  • 0 in reply to Icy
    TI__Guru** 104275 points
    You can test by using the same connection as TPA6205A1. This would change the input impedance but it should work as summing inputs.