I am making a VU meter, and I need to rectify the audio signal for the IC comparator chip. I found this circuit in the datasheet, but I am unsure of how it works.

enter image description here

I understand that the LF351 op amp is needed to overcome the threshold voltage of the diodes, but what is the purpose of the LM307 op amp? Is it just a buffer? Why not just use another LF351?

  • \$\begingroup\$ if this is feeding directly into a comparator circuit, have you considered a window comparator circuit instead of precision rectifier -> comparator? depending on your specific needs it might be small, cheaper, more accurate \$\endgroup\$ – JonRB Dec 16 '18 at 23:14
  • \$\begingroup\$ @JonRB Probably using an LM3915 or something like that. \$\endgroup\$ – Spehro Pefhany Dec 16 '18 at 23:32
  • \$\begingroup\$ ah, fair enough. \$\endgroup\$ – JonRB Dec 16 '18 at 23:34
  • \$\begingroup\$ There are better designs that are non-inverting single supply that work down to a few mV analog.com/en/design-center/reference-designs/… \$\endgroup\$ – Sunnyskyguy EE75 Dec 17 '18 at 4:31
  • 1
    \$\begingroup\$ @JonRG Yes, this is feeding into an LM3915. \$\endgroup\$ – Sal M Dec 17 '18 at 22:26

This is a rather tricky schematic; to understand it, you need to consider 2 case whether the input signal (after C1) is positive or negative.


simulate this circuit – Schematic created using CircuitLab

When it is positive, D2 is blocked and you can remove it and study the simplified schematic. OA1 behaves as an inverting amplifier with gain -1 whose output is \$V_2\$: \$V_2 = - V_{IN}\$. Then OA2 is an inverting summer: \$V_{OUT} = - {R_5 \over R_4} V_{IN} - {R_5 \over R_3} V_2 = - V_{IN} - 2 V_2 = V_{IN}\$.

When it is negative, D1 is blocked and you can remove it and study the simplified schematic. OA1 behaves as a follower with input to ground and output at \$V_1\$: \$V_1 = 0\$. Then OA2 is an inverting summer: \$V_{OUT} = - {R_5 \over R_4} V_{IN} - {R_5 \over R_2 + R_3} V_1 = - V_{IN} - V_1 = -V_{IN}\$.

As for the choice of OA1 and OA2, when $V_{IN}$ crosses zero, the output of OA1 needs to go quickly from -0.6V to 0.6V. You need an op amp with a hich slew rate. There is no such requirement for OA2.


In answer to your question, the most likely answer is that the LF351 was more expensive so they used a cheaper op-amp where high slew rate was not required. The LM307 is/was a 741 type of op-amp, slow and semi-precision. Slew rate is less than 1/20 that of the LF351.

It's important to have a high slew rate amplifier in the LF351 position because there the output of the input op-amp should move quickly from -0.7V to +0.7V (ideally instantly) since one diode or the other should always be conducting.

To see how it works, consider when the AC-coupled input is > 0, say V1. Current flows through R1 and R2 and D1 so the left side of R3 is at -V1, thus the current flowing into the node at pin 2 of the LM307 is Vin/R4 - Vin/R3 + Vout/R5, so the output voltage (ignoring C2 for the moment) is +Vin

enter image description here

When the AC-coupled input (right side of C1) is < 0, call it -Vin, D2 conducts (keeping the op-amp output from straying too far from ground, reducing the demands on slew rate) and the current flowing into the node at pin 2 of the LM307 is -Vin/R4 + Vout/R5, so Vout = -Vin (again ignoring C2).

enter image description here

C2 acts to average the output voltage with a time constant of C2*R5.

In the second case there is a resistance R2 + R3 between pin 3 of the LM307 and pin 2 of the LF351. Since both nodes are always at virtual ground there is (ideally) no effect from that 200K resistance.

  • \$\begingroup\$ When the AC-coupled input is positive (V1), you mentioned that the left side of R3 is at -V1. Why is that? Is it because the gain of the 351 is -1, making the voltage in the feedback loop through D1 negative? \$\endgroup\$ – Sal M Dec 17 '18 at 22:43
  • \$\begingroup\$ Yes, exactly. The gain is -1, measured at the right side of R2. The gain from the output of the LF351 will have a slightly different value and a relatively large offset, but we don't care much about that. To drive the inverting input of the LF351 to exactly 0V the voltage at that point must be -V1 (assuming no op-amp bias current and resistors exactly matched). \$\endgroup\$ – Spehro Pefhany Dec 17 '18 at 22:47
  • \$\begingroup\$ Awesome I understand now. I ordered the components, should be a while before they get here though. Thanks for the help! \$\endgroup\$ – Sal M Dec 18 '18 at 0:09
  • \$\begingroup\$ When wiring up the LF351 amplifier, do I need to use a dual supply voltage rail? Or can I just connect the VCC- pin to ground? I assume I can connect it right to ground because no signal is being amplified right? \$\endgroup\$ – Sal M Dec 21 '18 at 20:55
  • \$\begingroup\$ No, it has to swing negative by the maximum positive input voltage, so if the maximum peak input voltage is, say, 7V, you should have a negative rail of about -10V or better (the LF351 needs about 3V more for the output stage). \$\endgroup\$ – Spehro Pefhany Dec 21 '18 at 21:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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