13
\$\begingroup\$

I'd like to fully understand the circuit marked as "Minimum Parts" at page 5 in LM386's datasheet:

enter image description here

  • Is there any reason for 10k resistance at the input potentiometer or is it just an arbitrary value?
  • For an audio application, should that be a logarithmic pot?
  • There is a resistor and a capacitor connected in series to ground right after the output. It looks like a low-pass filter to me, but I've usually seen those with the load connected between R and C, not parallel to both. What's going on?
  • Why is there a 250uF polarized capacitor right before the speaker? Is it just a DC blocking cap? Why 250uF and not other value? Perhaps just a high value so that the resulting high-pass filter cutoff is low enough?
\$\endgroup\$
3
  • 10
    \$\begingroup\$ +1 for the well written question in the first try. There should be a badge for that. Easy to read and to follow, with convenient link to datasheet. An example for everyone. Well done! \$\endgroup\$
    – Ricardo
    Feb 27, 2014 at 17:22
  • 1
    \$\begingroup\$ "Zobel network", that is the name for the "resistor and a capacitor connected in series to ground right after the output". It is to prevent instability with the amplifier when no load is attached. \$\endgroup\$
    – jippie
    Feb 27, 2014 at 18:02
  • 3
    \$\begingroup\$ Spelling, punctuation, and capitalization are even correct! Tears of joy! \$\endgroup\$
    – JYelton
    Feb 27, 2014 at 18:14

1 Answer 1

6
\$\begingroup\$

Is there any reason for 10k resistance at the input potentiometer or is it just an arbitrary value?

It's more-or-less arbitrary. 20K or 5K would be fine too, 10M, not so much. They suggest less than 250K source resistance, which would imply a 1M pot driven from a low impedance source would be just okay.

For an audio application, should that be a logarithmic pot?

Yes.

There is a resistor and a capacitor connected in series to ground right after the output. It looks like a low-pass filter to me, but I've usually seen those with the load connected between R and C, not parallel to both. What's going on?

It's a type of compensation, to keep the amplifier from oscillating. You may notice that when you want it to be an oscillator, it's not required. Take note also of the minimum gain requirement. The part is only guaranteed stable for gain > 9. As Brian Drummond points out, it is a Zobel network that helps make the speaker look less inductive and more resistive.

Why is there a 250uF polarized capacitor right before the speaker? Is it just a DC blocking cap?

Yes.

Why 250uF and not other value? Perhaps just a high value so that the resulting high-pass filter cutoff is low enough?

It needs to be be low impedance compared to the speaker at the lowest frequency of interest. A 250uF capacitor has 8 ohms reactance at ~80Hz.

\$\endgroup\$
9
  • 1
    \$\begingroup\$ Thanks for the answer. About the RC keeping the amplifier from oscillating... where can I learn more or what would be good search terms for Google? I'm trying to understand the circuit as much as possible. \$\endgroup\$
    – DuckTyped
    Feb 27, 2014 at 17:12
  • 5
    \$\begingroup\$ Also search for "Zobel network". \$\endgroup\$
    – user16324
    Feb 27, 2014 at 17:20
  • 2
    \$\begingroup\$ Sorry for asking, but why a logarithmic pot should be used for an audio application? \$\endgroup\$
    – Ricardo
    Feb 27, 2014 at 17:26
  • 3
    \$\begingroup\$ @Ricardo Because human ears respond logarithmically. If you use a linear pot, the volume won't appear to change much until you get right down near the bottom, then it will suddenly drop off. \$\endgroup\$ Feb 27, 2014 at 17:28
  • 2
    \$\begingroup\$ @Ricardo: If the purpose of the knob is to adjust the relative loudness of the signal, much of the useful adjustment range would be toward the lower end of a linear pot; an audio-taper pot would shift that toward the middle. If the purpose of the knob is to reduce the volume enough to limit distortion to a tolerable level, more of the useful adjustment range may, depending upon input signal variations, be nearer the middle of a linear-taper pot, and thus a linear-taper pot may be preferred in such cases. \$\endgroup\$
    – supercat
    Feb 27, 2014 at 20:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.