mmmerle

1. Fuzz

Fuzz is the payoff chapter for transistors and diodes: it’s what you get when a transistor’s small controlling current is pushed so far past its comfortable range that the output stops scaling and flattens at the top and bottom of its swing. Every fuzz circuit, from the earliest germanium designs to modern silicon clones, is some arrangement of that same overdriven-transistor mechanism — the differences between them are almost entirely about how many stages do it and how they’re biased.

Two circuits, not one genre

“Fuzz” covers a wide family of pedals, but nearly all of them trace back to one of two circuit topologies:

• Fuzz Face — a 2-transistor circuit built around a feedback bias arrangement: the second transistor’s collector feeds a bias resistor back to the first transistor’s base, which means the two transistors’ gain (hFE) values interact directly and the whole circuit’s bias point shifts if you swap either one. Originally built with germanium transistors, later reissued with silicon.
• Big Muff — a 4-transistor circuit built as two cascaded clipping stages (two transistors each) followed by a passive tone stack and an output stage. Each stage clips independently against silicon diodes, then hands the result to the next stage, producing a denser, more sustained, more “distortion-like” fuzz than a Fuzz Face’s single feedback-biased pair.

The mental model: an overdriven valve, twice or four times in a row

Go back to the valve mental model from transistors and diodes: a small current at the base controlling a much larger current between collector and emitter, until you push it hard enough that the output can’t scale anymore and flattens off. A Fuzz Face is that flattening happening across two transistors locked together by a shared feedback bias point, so the two stages behave as one interdependent clipping unit rather than two separate ones. A Big Muff is the same flattening happening four times, in two independent pairs, each pair handing an already-clipped signal to the next — which is exactly why a Big Muff sounds denser and more sustained than a Fuzz Face: the second stage is clipping a signal that’s already been clipped once, not a clean input.

Why a Fuzz Face reacts to your guitar’s volume knob and a Big Muff mostly doesn’t

This is the single most distinctive property of fuzz circuits, and it doesn’t show up anywhere else in this book: a Fuzz Face presents an unusually low, non-standard input impedance, low enough that it’s directly affected by the guitar’s own volume pot sitting upstream of it. Roll the guitar’s volume down, and you’re not just sending a quieter signal into the fuzz — you’re changing the load the fuzz’s first transistor sees, which shifts its bias point and measurably cleans up the clipping character, not just the volume. This is why “clean up the fuzz by rolling back the guitar’s volume knob” is a real, physically-grounded playing technique for a Fuzz Face specifically, not folklore. A Big Muff’s front end is designed with a more conventional input stage and largely doesn’t exhibit this behavior — its clipping character stays consistent regardless of where the guitar’s volume knob sits.

Common mistake: assuming any two “matching” transistors will bias a Fuzz Face correctly

Because a Fuzz Face’s two transistors share a feedback bias relationship, the circuit is unusually sensitive to the specific hFE (gain) values of the pair, and unusually intolerant of an arbitrary substitution — dropping in two transistors that are individually within spec but poorly matched to each other is a common reason a freshly-built or freshly-repaired Fuzz Face bias point reads far from what the schematic predicts, even though every individual part measures fine in isolation. Check the bias point at each transistor’s collector against the schematic’s predicted value using the debugging approach — stage by stage — before assuming a “bad part” when the real issue is an unmatched pair. Builders sourcing specific germanium transistors for a vintage-correct build lean on suppliers like Small Bear Electronics, who specifically stock obsolete and germanium semiconductors that general-purpose suppliers don’t carry.

A finished, board-level walkthrough of building one of these circuits is covered in Fuzz Face.

2. Overdrive and Distortion

“Overdrive” and “distortion” get used almost interchangeably by players, but as circuits they’re built differently, and the difference is exactly where the diode clipping sits relative to the gain stage — a distinction op-amps already flagged and promised would land here.

The Tube Screamer topology: clipping inside the feedback loop

The circuit behind the Tube Screamer and its many derivatives pairs an op-amp gain stage (see op-amps for the inverting-amplifier math) with a pair of diodes wired directly across that same feedback path — not after the gain stage, but inside it. The op-amp tries to keep amplifying according to the resistor ratio that sets its gain, but the moment the output swing exceeds the diodes’ forward voltage, the diodes start conducting and effectively clamp the feedback path, which caps the output right at that threshold. The result is compression baked into the clipping itself: quiet input stays clean and fully amplified, loud input gets progressively clamped, and the transition between the two is soft rather than abrupt.

Post-gain clipping: what makes a distortion circuit different

A distortion circuit — as opposed to an overdrive — typically runs a signal through a fully clean, wide-open gain stage first, and only clips it afterward, against diodes wired to ground rather than inside the gain stage’s own feedback path. Because the gain stage isn’t constrained by a clipping element sitting inside its feedback loop, the clipping tends to be harder and more abrupt rather than the Tube Screamer’s soft, compressed transition — which is the actual, circuit-level reason “overdrive” is consistently described as smoother and “distortion” as harsher, rather than the two words just being marketing.

The MXR Distortion+, first released in the late 1970s, is close to the simplest possible version of this topology: an op-amp gain stage followed by a pair of silicon diodes clipping to ground, with almost nothing else in the signal path. The ProCo Rat, released not long after, runs the same basic op-amp-plus-diodes-to-ground topology but builds in two deliberate twists: it uses an LM308 op-amp specifically for its unusually slow slew rate, which rolls off the circuit’s high-frequency response and gives the Rat its distinctive compressed, slightly dark clipping character even before the tone control does anything; and its single “Filter” knob is wired backwards from a conventional tone control — turning it clockwise cuts treble instead of adding it, the opposite of what a player used to a normal tone knob expects on first encounter.

The mental model, extended: the assistant now has a hard ceiling

Recall the “overzealous assistant” from op-amps: the feedback resistor tells the assistant to shout proportionally instead of maximally. Adding diodes to that same feedback path is telling the assistant “shout proportionally, but never louder than this specific volume” — the moment the proportional shout would exceed that ceiling, the diodes hold it there instead. That’s an overdrive circuit. A distortion circuit, by contrast, lets the assistant shout as loud as it wants with no ceiling built into its own instructions, and clips the shout only after it leaves the room.

Common mistake: assuming more gain turns an overdrive into a distortion

Turning up an overdrive’s gain control makes it clip more, sooner, and louder — but it doesn’t change where the clipping happens, so a maxed-out Tube Screamer still has the soft, compressed, in-loop clipping character its topology produces; it just has more of it. Getting the harder, more abrupt character associated with “distortion” requires a genuinely different circuit — post-gain clipping — not just more gain from an overdrive circuit. When a schematic or a forum thread specifies one over the other, that’s a topology decision baked into the design, not a knob setting you’re missing. A worked build of the in-loop-clipping topology is covered start to finish in Tube Screamer Clone.

3. Delay

A delay pedal is, at its core, a recording that plays back late: feed it a note, and some amount of time later it hands that note back to you, usually blended with whatever you’re playing in the meantime. Every delay circuit ever built comes down to one design question — how is that brief recording actually stored? — and the answer splits the whole category into three genuinely different mechanisms, not just three tone flavors of the same idea.

The mental model: recording the echo of the echo

Picture a delay as a short loop of tape that’s constantly being recorded onto and played back from, a fixed distance behind where the recording head currently sits — the “delay time” is just how far behind that playback head trails the recording head. Feedback is what happens when you route some of that playback signal back into the recording input: instead of recording only your guitar, the circuit is now recording your guitar and the echo of your guitar, so the echo itself gets echoed, and so on, decaying a little more with each pass. That’s the entire mechanism behind every repeat you hear trailing off after a single note.

Three ways to build the “tape loop,” in order of how they actually store the signal

The BBD row and the true-digital row are the two extremes most players actually mean when they say “analog delay” and “digital delay.” The PT2399-style middle row is the one that trips people up: it’s genuinely digital internally, but its low-resolution sampling gives it a lo-fi character closer to a BBD chip than to a full ADC/DAC digital delay — which is exactly why plenty of budget delay pedals marketed as “analog-voiced” or “warm digital” are quietly built around this same inexpensive chip rather than a true BBD chain.

Feedback and mix: the two controls that shape every delay regardless of mechanism

Every delay circuit, whatever it stores the signal in, exposes the same two controls because they’re the two knobs that actually define the repeats:

• Feedback (sometimes labeled “regeneration” or “repeats”) — how much of the delayed signal gets routed back into the circuit’s input, determining how many audible repeats you get before they decay into silence. Turned high enough, feedback approaches or exceeds unity gain, and the repeats stop decaying — they sustain indefinitely or build in volume, a deliberately-chased effect called self-oscillation.
• Mix (or “level”) — how much of the delayed signal is blended back in with the dry, unprocessed signal, independent of how many repeats there are or how long they last.

Common mistake: not distinguishing a runaway feedback control from a broken pedal

Turning the feedback knob past a certain point on almost any delay causes the repeats to stop decaying and instead build in volume or sustain indefinitely — a shrieking, cascading buildup that sounds like a malfunction to someone encountering it for the first time. This is the circuit doing exactly what feedback greater than or equal to unity gain does by definition, not a fault, and it’s a legitimate, deliberately-used technique (self-oscillating delay) rather than something to debug. If a build’s repeats run away at a much lower feedback setting than expected, or won’t reach self-oscillation at all even at maximum, that is worth checking against the schematic’s predicted gain at that stage using the debugging approach — but a delay that screams at 90% feedback and behaves normally below that is working correctly.

Push the repeat count high enough and close enough together, and the effect stops sounding like discrete echoes at all — that’s the exact boundary Reverb picks up from, simulating thousands of reflections rather than counting individual ones.

4. Modulation

Chorus, flanger, phaser, tremolo, and vibrato sound like five unrelated effects, but they’re built from one shared mechanism applied to five different targets: a low-frequency oscillator, or LFO, that slowly and silently steers some parameter of the circuit back and forth, the way a hand would turn a knob if it moved smoothly and automatically instead of in discrete adjustments.

The mental model: an invisible hand slowly turning one knob

An LFO is an oscillator too slow to hear directly — typically well under 20Hz, often under 5Hz for the effects in this chapter — running silently in the background and outputting a smoothly rising-and-falling control voltage instead of an audible tone. Every modulation effect wires that control voltage to a different point in the signal path: what changes between chorus, flanger, phaser, tremolo, and vibrato isn’t the LFO itself, it’s what it’s steering. Rate controls how fast the invisible hand moves; depth controls how far it turns the knob each cycle.

The five effects, sorted by what the LFO actually modulates

Why chorus, flanger, and vibrato get confused constantly, and phaser doesn’t need a delay line at all

Chorus, flanger, and vibrato are all built from the same underlying part — a short, LFO-modulated delay line — which is exactly why they’re so often mixed up by ear alone: the difference between chorus and flanger is a feedback path, and the difference between chorus and vibrato is whether the dry signal is blended in at all (chorus) or discarded entirely (100% wet vibrato). Phaser is the odd one out in this table: it produces a superficially similar sweeping, swooshing sound using all-pass filter stages instead of any delay line, and the number of filter stages (commonly four, six, eight, or ten in different commercial designs) directly determines how many notches sweep through the spectrum and how dense the effect sounds.

The historical mixup: Fender’s “tremolo” and “vibrato” are swapped from their technical definitions

This is worth naming explicitly because it’s a genuine, widely-documented historical naming error that still causes confusion today, not just imprecise slang: vintage Fender amplifiers famously labeled their amplitude-modulation circuit “vibrato” on some models and their pitch-bending guitar hardware “tremolo arm” (the whammy bar) — backwards from the technical definitions in the table above, where tremolo is amplitude and vibrato is pitch. The mislabeling stuck hard enough in guitar culture that “tremolo arm” for a pitch-bending bridge is still the near-universal term today, even though it’s modulating pitch, which is technically vibrato. When a schematic or spec sheet uses one of these two terms, check what it actually modulates rather than assuming the word matches its technical definition.

Common mistake: assuming rate and depth are the only two things that define a modulation effect’s character

Two modulation pedals with identical rate and depth settings can sound completely different because the underlying LFO waveform shape (triangle, sine, or square) changes how the modulated parameter moves between its extremes — a triangle-wave LFO moves at a constant rate and reverses direction sharply, a sine-wave LFO eases in and out of each extreme, and a square-wave LFO snaps between two settings with no gradual sweep at all, producing a much more abrupt, choppy effect (most commonly heard in hard-edged tremolo circuits). Rate and depth control the scale of the modulation; the LFO’s waveform shape controls its feel, and it’s just as responsible for a given pedal’s character as the two knobs a player actually sees.

5. Reverb

Reverb and delay are close relatives: a delay gives you back a handful of discrete, countable repeats, while reverb simulates thousands of reflections arriving so closely together that they blur into a single, continuous decaying wash instead of individual echoes. Where the two circuit families genuinely diverge is how that dense wash of reflections gets generated — and the two dominant answers, a physical spring and a digital delay network, don’t share a single component between them.

Spring reverb: sound converted to mechanical motion and back

A spring reverb tank does something no other circuit in this book does: it converts the electrical signal into physical motion, sends that motion down a real spring, and converts it back. A transducer at one end of the spring vibrates the spring itself in response to the incoming signal; a second transducer at the other end picks up whatever vibration arrives and turns it back into an electrical signal. Because a spring is a physical object with its own resonances, dispersion (different frequencies genuinely travel down the spring at different speeds), and multiple internal reflection paths bouncing between its ends, the recovered signal comes back smeared in time and colored in a specific, slightly metallic way that’s the entire reason spring reverb has its own distinct, recognizable identity rather than sounding like a generic “echo.”

Digital reverb: the same dense-reflection idea, built from delay lines instead of springs

A digital or algorithmic reverb has no physical spring anywhere in it — it recreates the same “thousands of overlapping reflections” idea using a network of short digital delay lines with feedback paths between them (commonly called a feedback delay network, or FDN), tuned so their combined output decays smoothly and densely rather than as a handful of countable repeats. This is a direct extension of the delay circuitry covered in Delay: the same circular-buffer mechanism that produces a single countable echo produces a convincing reverb wash when enough delay lines of different, carefully chosen lengths feed back into each other simultaneously. DaisySP, the DSP building-block library covered in the Electrosmith Daisy Guide, ships a ready-made reverb building block built exactly this way, so a Daisy-based pedal gets access to this circuit without hand-deriving the delay-network math from scratch.

Common mistake: treating a spring tank like an ordinary circuit board when handling it

A spring reverb tank isn’t just electrically sensitive the way any circuit is — it’s mechanically sensitive in a way nothing else in this book is, because the spring is a real, physically resonant object that converts a knock or a drop directly into an audible signal through the same transducers doing the reverb effect on purpose. This is a genuine, load-bearing part of spring reverb’s identity (the “crash” from kicking a tube amp with a built-in spring tank is a real, well-known effect players intentionally use), but it also means a spring tank has to be mounted and handled with actual physical care during a build or when transporting a finished pedal — something no other circuit in the Effects book requires thinking about at all.

6. Boost and Buffer

Boost and buffer are the simplest gain circuits in this book, and they’re deliberately the last chapter for that reason: everything needed to understand them — a controlled op-amp or transistor gain stage — was already covered in op-amps and transistors and diodes. What’s new here isn’t the circuit, it’s why something this simple is worth building at all — and its simplicity is exactly why breadboarding and prototyping singles it out as a good low-stakes circuit to practice the prototype-before-you-commit habit on: a handful of parts, one gain stage, and a result you can judge by ear immediately.

The mental model: a buffer changes who’s listening, not what’s said

A guitar’s passive pickups are a comparatively weak, high-impedance signal source, and what’s connected downstream of them — a cable, a pedal’s input, an amp — actually affects the tone that reaches your ears, because a long cable’s capacitance interacts with the pickup’s inductance and rolls off high frequencies the further the signal has to travel before being “picked up” by something with a low input impedance. A buffer is a circuit whose entire job is to present a high input impedance to the guitar (so it doesn’t load the pickups down) and a low output impedance to everything after it (so the rest of the chain, however long, doesn’t matter anymore). Nothing about the audio is amplified in a musically meaningful sense — a buffer’s gain is unity, exactly 1x — but the signal’s ability to survive a long cable and a pedalboard full of other pedals without losing treble is exactly what it exists to protect.

Buffer vs. boost: the same circuit, one resistor value apart

A boost is a buffer with gain added — literally the same non-inverting op-amp topology (or a comparable transistor gain stage), with the feedback resistor ratio set to something greater than 1 instead of exactly 1. Where a buffer’s job is purely preservation, a boost’s job is to push the signal louder before it hits the next pedal or the amp’s front end — useful for a volume lift going into a solo, or for driving an already-overdriven amp or fuzz harder into its own clipping.

True bypass vs. buffered bypass

A “true bypass” pedal, when switched off, disconnects itself from the signal path entirely — nothing about the pedal touches your tone when it’s off, but it also does nothing to counteract cable-length tone loss. A “buffered bypass” pedal keeps a buffer stage active in the path at all times, even when the pedal’s main effect is switched off, which is why a pedalboard with one well-placed buffered pedal near the start of a long chain of true-bypass pedals commonly sounds noticeably brighter and more solid than the same chain with no buffer anywhere in it. The actual footswitch wiring behind both — which poles of a 3PDT do what — is covered in Footswitch and True-Bypass Wiring.

Common mistake: putting a buffer in front of a fuzz

This is the one place a buffer actively makes things worse rather than better, and it directly connects back to fuzz’s input-loading quirk: a Fuzz Face-style circuit depends on seeing the guitar’s own passive pickups and volume pot directly, because that specific, non-standard input impedance is part of what makes its bias point (and its volume-knob cleanup behavior) work the way it does. A buffer sitting in front of it presents a low, stable output impedance instead of the guitar’s own — which is precisely what a buffer is designed to do everywhere else — and that changes the fuzz’s bias point and clipping character, usually described as the fuzz sounding thinner, more compressed, or “wrong” compared to plugging straight into it. The fix is ordering, not a different buffer: place any always-on buffer after a vintage-style fuzz in the signal chain, never before it, and if a pedalboard’s buffer placement is fixed, put the fuzz in a spot upstream of it or accept that the fuzz needs to run first.