Does there exist an Op-Amp network topology to amplify only the positive, or negative cycles, whilst keeping the corresponding (negative, or positive) cycles at unity gain?

I want to send a small signal into an op-amp, amplify only the positive or negative cycles, then send the output into another op-amp to amplify the 'other cycle'.


buffer, then amplify only positive cycles keeping negative at unity gain, then amplify only the negative cycles keeping positive cycles at unity gain, leading to the next stage/output


I understating the the super diode/ precision half wave rectifier (Abbreviation: PHRF) allows for splitting of positive or negative cycles, but I am hesitant to split the incoming signal into two parallel routes, use two PHWRs to separate the positive and negative cycles and then use further stages to amplify the two routes, and then sum them with another op-amp for fear of incurring phase issues when recombining the signals.

Split signal with two precision half wave rectifiers, amplify them separately and then sum the signals

This plan is to feed a clipping stage for asymmetric clipping. As long as the clipped (guitar) signal does not suffer too much phase difference upon recombination as per 2nd image, the audio signal will be recognisable to the user.

It is important to be able to vary the positive and negative cycle amplification independently as this will give the asymmetric clipping.



  • 1
    \$\begingroup\$ Ummm. Why? What's wrong with just amplifying it the same way everybody else does? It looks like you want the same result, but are taking a roundabout and complicated way. \$\endgroup\$ – JRE Jun 14 '17 at 16:08
  • \$\begingroup\$ I intend to a symmetrically clip the output \$\endgroup\$ – DWD Jun 14 '17 at 16:10
  • \$\begingroup\$ Although you say you want to avoid a PHRF: You now that there are specialized IC like AD630 built exactly for this purpose? It is also used in Lock-In-Amplifier applications where phase-noise from recombination would be bad. What is your exact requirement concerning the phase noise? \$\endgroup\$ – JLo Jun 14 '17 at 16:10
  • \$\begingroup\$ Then you need to look for a circuit that clips symmetrically. The usual way would be two diodes in antiparallel across the feedback resistor of an opamp. You follow that by a gain stage to bring the level back up to where you want it. \$\endgroup\$ – JRE Jun 14 '17 at 16:13
  • \$\begingroup\$ I do not have an exact specification, the application is an audio circuit which will be able to amplify positive and negative cycles independently. This will feed a clipping stage for asymmetric clipping. As long as the clipped (guitar) signal does not suffer too much phase tdiffetence upon recombination as per 2nd image, the audio signal will be recognisable to the user \$\endgroup\$ – DWD Jun 14 '17 at 16:15

Since you say you want asymmetric clipping, I'm going to present you with a circuit that does that.

It DOES NOT solve your stated problem of separate gains for the positive and negative halves of the signal - I think that is just your imagined way to get what you wanted.

I'm going to use a symmetric clipper, but make the signal asymmetric. The result is an asymmetrically clipped signal.


simulate this circuit – Schematic created using CircuitLab

R1 and OA1 are used to add a DC offset to the signal. Add more than half of V+, and the positive half gets clipped more. Add less than half of V+ and the negative half of the signal gets clipped more. Adjust the level of the signal going in to adjust how much clipping is done, adjust R1 for the asymmetry.

Much simpler than trying to split the signal into halves, fiddle with them, then add them back together.

| improve this answer | |
  • \$\begingroup\$ Thank you @JRE. Your comments and answer have given me much to think about and a way forward with the stated problem. \$\endgroup\$ – DWD Jun 14 '17 at 17:27
  • \$\begingroup\$ Basically a distorsion pedal for the guitar :D \$\endgroup\$ – Andrés Jun 14 '17 at 18:12

And another asymmetrical clipping circuit (but with more precision): -

enter image description here

U2 and U3 perform the clipping/clamping. U2 prevents the output signal rising above V3 (can be set with a pot) and U3 prevents the output signal falling below V4 (can also be set with a pot). You can check what the output looks like courtesy of this website.

| improve this answer | |

Your reticence to split the signal, process the two halves independently, and resassemble them pretty much rules out analog techniques. The alternative is to sample, process using DSP, and then output the new signal with a DAC.

I'd say try it with precision half wave rectifiers. More than likely, its good enough. I don't anticipate phase issues at audio frequencies.

| improve this answer | |
  • \$\begingroup\$ Thanks @Scott Seidman, I need to go down the analogue route, and am pleased to hear you think the PHWR will suffice for the application. \$\endgroup\$ – DWD Jun 14 '17 at 17:42

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.