7
\$\begingroup\$

I have been trying to implement a basic hard-clipping circuit for a distortion guitar pedal, based on two simple diodes, but somehow I can't get it to work as it is supposed to be. The basic design is the following:

enter image description here

However, my out signal is something like this, which doesn't seem very "hard clipped".

enter image description here

In this basic circuit the signal gets clipped, but i don't know what could be the difference between the two: enter image description here

Am I doing something wrong? The signal in OUT_INPUT_STAGE is a sine wave with an amplitude of 2 V, so it should be enough to be clipped by the diodes.

\$\endgroup\$
4
  • \$\begingroup\$ Try eliminating C5 and change R5 to 100 ohms. \$\endgroup\$
    – MOSFET
    Apr 23 at 20:08
  • \$\begingroup\$ A diode clamp is usually to protect whatever input it's connected to, not to "hard clip". \$\endgroup\$ Apr 23 at 20:24
  • 2
    \$\begingroup\$ @MOSFET C5 serves as the decoupling from DC to AC on the output and cannot be removed while still achieving the (probably) desired decoupled ground centered AC signal. Note that the opamp has its negative supply tied to ground. \$\endgroup\$
    – Mels
    Apr 24 at 11:30
  • 1
    \$\begingroup\$ @ScottSeidman special effects circuits for electric guitars (like the OP's 'distortion guitar pedal') often break the usual circuit design rules and introduce deliberate distortions/nonlinearities into the audio signal. I have seen a number of schematics for such circuits that use diodes in this kind of "hard clip" configuration. \$\endgroup\$
    – Sotto Voce
    Apr 24 at 16:01

6 Answers 6

6
\$\begingroup\$

There are diodes better than everyday diodes :)

This is a Q1 working "backwards" (emitter and collector flipped) as an improved diode that clips very nicely:

schematic

simulate this circuit – Schematic created using CircuitLab

Compare the characteristic curve of the improved diode to an ideal diode and a regular Si diode:

enter image description here

We can imagine using these improved diodes in a clipper.

schematic

simulate this circuit

A complete clipper will look as follows:

schematic

simulate this circuit

The circuit is designed for a single supply, e.g. 9V. The buffer between CLIP and OUTPUT has not been shown.

OA1 provides gain to the input signal. The gain can be variable. What matters here is that the signal is buffered before it enters the clipper.

Here's how the clipper reacts to amplitude adjustment:

enter image description here

And here's what happens when the stiffness is increased - there's punch-through, which may be a desirable effect in itself:

enter image description here

\$\endgroup\$
5
\$\begingroup\$

You won't get hard-clipping with a basic diode because its forward conduction point is not on/off at 0.7 volts but, gradually increases from about 0.5 volts to heavy conduction at 0.8 volts (typically). No basic diode will do what you are trying to do.

Your scope image is precisely what I'd expect: -

enter image description here

Image from Diodes.

\$\endgroup\$
5
\$\begingroup\$

The diodes don't suddenly begin to conduct when some voltage across them (700mV) is exceeded. Rather they begin to conduct as the voltage nears 500mV, conducting better and better as the voltage continues to rise, until about 700mV where they have negligible resistance.

This gives rise to the somewhat sluggish curves you see, instead of the nice sharp corners you want.

You can obtain much sharper clipping using so-called "diode-connected transistors", which behave much more like "ideal" diodes, and you can further improve things by placing them in the op-amp's feedback path, instead of relying on clamping the output of a follower via those 470Ω resistances.

Here's the simplest implementation I can think of right now:

schematic

simulate this circuit – Schematic created using CircuitLab

enter image description here

\$\endgroup\$
4
\$\begingroup\$

To get the clipping in the diagram you'll need to use ideal diodes in spice that have a better roll off.

If you want to build a more realistic circuit, try bringing down the resistor below 400Ω (lower the better until but still limit current from the opamp) and ditch the capacitor.

If you want hard clipping,try an opamp clipping circuit like this:

schematic

simulate this circuit – Schematic created using CircuitLab

Determining noise performance of clipper circuit

\$\endgroup\$
2
\$\begingroup\$

Put a small current through the diodes to make them limit harder. The output is limited to 200mVpp in this example:

Diode limiter

Another version with adjustable limit that needs four diodes. Supplied with 1.5V DC you get 600mVpp at the output:

Adjustable limiter

\$\endgroup\$
1
\$\begingroup\$

The difference in your two circuits is the following:

The op-amp controls its output to be an exact replica of the input (amplified) until it reaches its output current limit.

In the first circuit you have another 470 Ω resistor inserted between the clipping diodes and the stiffly controlled op-amp output. So the point after that resistor fails to maintain the sharply clipped signal from the op-amp output because the rising diode current also creates an additional voltage drop across the resistor, which smooths the signal.

In practise you always do want some resistance in series with the diodes though, as otherwise you create a short circuit condition. That might not immediately fry your op-amp or diode but is not a graceful operation condition regardless. So if you still want an analog hard clipping, you have to invoke some of the suggestions from other answers.

\$\endgroup\$

Your Answer

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

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