Chuck D. Bones
Circuit Wizard
I had a request to give a basic primer on opamps. What does an opamp do? What's the difference between them? Why do builders use certain opamps? Which ones have the most mojo? Can I sub X for Y?
In part 1 I'll give a basic overview
First off, let's talk about what an opamp is. The most basic opamp has five terminals. The positive power rail, the negative power rail, the inverting input, the non-inverting input, and the output. The output voltage is the voltage difference between the inverting input (-) and the non-inverting input (+) multiplied by a very large gain. With an ideal opamp, that gain is infinite. In a real-world opamp, the gain is usually above 100,000. Many are around a million. Opamps were originally developed to be basic building blocks in an analog computer.
An ideal opamp has these characteristics:
infinite gain
infinite bandwidth - it can amplify all frequencies
infinite slew rate - the output can change as fast as the inputs
infinite input impedance
zero offset voltage - when the difference between the + and - inputs is zero, the output is exactly half-way between the power supply rails
zero bias current - no current flows in or out of the + and - pins
zero noise
infinite power supply rejection - hum or noise on the power rails does not affect the output
zero distortion
zero output impedance
infinite output current capability
the output can swing all the way to the power supply rails
the inputs will work correctly as long as they are greater than or equal to the negative power supply rail and less than or equal to the positive power supply rail
Most of these requirements cannot be met by a real opamp, but many opamps get pretty damned close. Close enough for building pedals.
The first opamps were built with vacuum tubes. They were not great and people started making them from silicon bipolar transistors instead. As soon as integrated circuit technology became commercially viable, opamps were made that way. With each technology advancement, opamps came closer to ideal.
The earliest use of an opamp in a pedal was the MXR Distortion+. It used a 741. The 741 was a cheap and plentiful general-purpose opamp. The biggest shortcoming of the 741 was noise. The 741 is a single opamp; there's one in a package. The 4558 is basically a dual 741. Pretty much the same design, same specs, except there are two on a die. The 4558 became the go-to dual opamp for low-cost consumer electronics, including guitar pedals.
Precision circuits required opamps with much better specs. Lower noise, lower offset voltage, higher input impedance, etc. One of the best solutions for this demand was the BiFET opamp. Chip manufacturers figured out how to put JFETs and BJTs on the same die. The input stage was JFETs and the rest was bipolar transistors. One of the earliest affordable JFET opamps was the LF351. Today, the JFET opamp of choice is the TL072.
The TL072 and 4558, appear in probably 90% of the opamp pedal circuits you'll come across. But there are a few other desirable opamps.
This brings us to the LM308. I've written about this guy on other pages in this forum. If you're going to run an opamp into saturation, this is one of the best choices out there. Unfortunately, they have been out of production for many years. You can find 'em on eBay and sometimes more reputable outlets, but they can be pricey. There are alternatives, but they don't always sound as good and you'll have to decide that for yourself. People like to talk about slew-rate and how the LM308's slow slew-rate is the source of the sweet tones. It may well be true, but I think there are other forces at play. The same mechanism that limits an opamp's slew rate also limits it's bandwidth - it's ability to amplify high freq signals. If we dime the DIST knob in the Rat, the LM308 runs out of gain just below 1KHz. Everything above that is diminished. There are far fewer harsh harmonics. Things that happen inside an opamp when it saturates are not well characterized (or in many cases, not at all) in the datasheet. Finding opamps that saturate in a way that sounds good is pretty much a matter of luck. Some people sub an OP-07 for am LM308. They don't sound the same to me in a Rat or similar circuit, but you'll have to decide for yourself. Opamps with external compensation (CA3130, LM301, LM308, NE5534, to name a few) allow the designer to dial-in the slew-rate & bandwidth with an external capacitor. That does not necessarily mean they can be made to sound the same. There are also "programmable" opamps that use an external resistor to control operating current, bandwidth & slew-rate. The MC3476 is one such beast. It sounds very good as a sub for the LM308 in a Roger Mayer Voodoo-1. It does not work well in a Rat.
Now let's talk about "audio-grade" opamps. These were designed specifically to meet the demands of high-end audio equipment. LM833, NE5532, OPAx134 are examples. They offer wide bandwidth and very low noise and distortion. This high performance comes at a price. These opamps typically consume more power (usually not a concern unless you're running on batteries) and they can be more expensive. Counterfeit OPA2134s are available on eBay, so beware. Putting a low-distortion opamp in a dirt pedal makes no sense to me, but some people do it.
Lastly, let's discuss the LM386. It's not exactly an opamp, although it shares some of an opamp's characteristics. The LM386 is a power amplifier designed to deliver up to 0.8W to a speaker load. They are also useful as a headphones amp. Somewhere along the line, somebody tried using one in a guitar pedal. Lo and behold, they sound pretty good when overdriven. EQD puts two of them in the Acapulco Gold. The SVDS wireless receiver uses one as the output amplifier. The SoloDalls Storm, which replicates the tone of the SVDS, also uses one. Brian Wampler put one in the front-end of the Velvet Fuzz. I put one in the Biggus Dickus. The LM386 is not interchangeable with regular opamps.
Coming up in part 2: How do opamps work in pedal circuits?
In part 1 I'll give a basic overview
First off, let's talk about what an opamp is. The most basic opamp has five terminals. The positive power rail, the negative power rail, the inverting input, the non-inverting input, and the output. The output voltage is the voltage difference between the inverting input (-) and the non-inverting input (+) multiplied by a very large gain. With an ideal opamp, that gain is infinite. In a real-world opamp, the gain is usually above 100,000. Many are around a million. Opamps were originally developed to be basic building blocks in an analog computer.
An ideal opamp has these characteristics:
infinite gain
infinite bandwidth - it can amplify all frequencies
infinite slew rate - the output can change as fast as the inputs
infinite input impedance
zero offset voltage - when the difference between the + and - inputs is zero, the output is exactly half-way between the power supply rails
zero bias current - no current flows in or out of the + and - pins
zero noise
infinite power supply rejection - hum or noise on the power rails does not affect the output
zero distortion
zero output impedance
infinite output current capability
the output can swing all the way to the power supply rails
the inputs will work correctly as long as they are greater than or equal to the negative power supply rail and less than or equal to the positive power supply rail
Most of these requirements cannot be met by a real opamp, but many opamps get pretty damned close. Close enough for building pedals.
The first opamps were built with vacuum tubes. They were not great and people started making them from silicon bipolar transistors instead. As soon as integrated circuit technology became commercially viable, opamps were made that way. With each technology advancement, opamps came closer to ideal.
The earliest use of an opamp in a pedal was the MXR Distortion+. It used a 741. The 741 was a cheap and plentiful general-purpose opamp. The biggest shortcoming of the 741 was noise. The 741 is a single opamp; there's one in a package. The 4558 is basically a dual 741. Pretty much the same design, same specs, except there are two on a die. The 4558 became the go-to dual opamp for low-cost consumer electronics, including guitar pedals.
Precision circuits required opamps with much better specs. Lower noise, lower offset voltage, higher input impedance, etc. One of the best solutions for this demand was the BiFET opamp. Chip manufacturers figured out how to put JFETs and BJTs on the same die. The input stage was JFETs and the rest was bipolar transistors. One of the earliest affordable JFET opamps was the LF351. Today, the JFET opamp of choice is the TL072.
The TL072 and 4558, appear in probably 90% of the opamp pedal circuits you'll come across. But there are a few other desirable opamps.
This brings us to the LM308. I've written about this guy on other pages in this forum. If you're going to run an opamp into saturation, this is one of the best choices out there. Unfortunately, they have been out of production for many years. You can find 'em on eBay and sometimes more reputable outlets, but they can be pricey. There are alternatives, but they don't always sound as good and you'll have to decide that for yourself. People like to talk about slew-rate and how the LM308's slow slew-rate is the source of the sweet tones. It may well be true, but I think there are other forces at play. The same mechanism that limits an opamp's slew rate also limits it's bandwidth - it's ability to amplify high freq signals. If we dime the DIST knob in the Rat, the LM308 runs out of gain just below 1KHz. Everything above that is diminished. There are far fewer harsh harmonics. Things that happen inside an opamp when it saturates are not well characterized (or in many cases, not at all) in the datasheet. Finding opamps that saturate in a way that sounds good is pretty much a matter of luck. Some people sub an OP-07 for am LM308. They don't sound the same to me in a Rat or similar circuit, but you'll have to decide for yourself. Opamps with external compensation (CA3130, LM301, LM308, NE5534, to name a few) allow the designer to dial-in the slew-rate & bandwidth with an external capacitor. That does not necessarily mean they can be made to sound the same. There are also "programmable" opamps that use an external resistor to control operating current, bandwidth & slew-rate. The MC3476 is one such beast. It sounds very good as a sub for the LM308 in a Roger Mayer Voodoo-1. It does not work well in a Rat.
Now let's talk about "audio-grade" opamps. These were designed specifically to meet the demands of high-end audio equipment. LM833, NE5532, OPAx134 are examples. They offer wide bandwidth and very low noise and distortion. This high performance comes at a price. These opamps typically consume more power (usually not a concern unless you're running on batteries) and they can be more expensive. Counterfeit OPA2134s are available on eBay, so beware. Putting a low-distortion opamp in a dirt pedal makes no sense to me, but some people do it.
Lastly, let's discuss the LM386. It's not exactly an opamp, although it shares some of an opamp's characteristics. The LM386 is a power amplifier designed to deliver up to 0.8W to a speaker load. They are also useful as a headphones amp. Somewhere along the line, somebody tried using one in a guitar pedal. Lo and behold, they sound pretty good when overdriven. EQD puts two of them in the Acapulco Gold. The SVDS wireless receiver uses one as the output amplifier. The SoloDalls Storm, which replicates the tone of the SVDS, also uses one. Brian Wampler put one in the front-end of the Velvet Fuzz. I put one in the Biggus Dickus. The LM386 is not interchangeable with regular opamps.
Coming up in part 2: How do opamps work in pedal circuits?
Last edited:
