The Map Is Not the Territory: The "Signal Path" Fallacy in Tube Amplifiers
"The map is not the territory" sounds almost too obvious. Of course a road map is not a highway, and a topographic map is not made of mountains. A map preserves the features that matter for a particular purpose and leaves out nearly everything else. That selectivity is not a defect. It is what makes a map useful. A road atlas that showed every tree, billboard, and buried cable would be a nearly useless guide for driving across the territory it so faithfully recorded.
Engineering depends on the same kind of simplification. A block diagram is a map. A schematic is a map. Even the familiar symbol for a capacitor is a map. Each represents certain properties of a physical system while deliberately omitting others. We could not think clearly about complicated circuits without these abstractions. Trouble begins when we forget that we are looking at an abstraction and start reasoning as though the picture were the physical circuit itself.
This confusion is the origin of the "signal path" fallacy in tube amplifiers. The phrase began as useful shorthand for describing how the major functional sections of an amplifier are connected. It has gradually hardened into something much more ambitious: a supposed rule declaring which components can and cannot affect the sound. A component that lies "in the signal path" is presumed important. A component outside it is dismissed as support circuitry, often with the confident declaration that "signal does not pass through it," therefore it cannot affect tone. That conclusion sounds plausible only because the signal path has been mistaken for something physically present in the amplifier.
The "Signal Path" Is an Artifact of the Block Diagram
At the level of a block diagram, the idea of a signal path is entirely sensible; in fact, the block diagram creates it. The guitar feeds the input stage, the input stage feeds the tone controls, the tone controls feed the phase inverter, the phase inverter feeds the output stage, and the output stage drives the speaker. Lines connecting those blocks tell us what feeds what. They give us a clear picture of the amplifier's overall architecture without forcing us to examine every resistor, capacitor, tube electrode, and power-supply node.
The blocks are useful precisely because they are black boxes. A rectangle labeled "gain stage" promises that some input will produce some output, but it says nothing about what determines the gain, how the stage clips, or why it responds differently at different frequencies. All of that has been hidden so that we can concentrate on the larger organization of the amplifier.
At this level, talk of flow is perfectly legitimate. "The signal goes from the preamp to the phase inverter" is a good sentence, and everyone understands it. It describes the transfer of information from one functional section to the next, which is exactly what the lines on the block diagram were drawn to show.
The mistake comes when we go on talking this way about the circuit itself, as if the line were a physical conduit inside the amplifier. The guitar signal is imagined as a substance flowing through it the way water flows through a pipe. Components lying directly in the pipe, we tell ourselves, can influence the signal because it passes through them; components connected elsewhere may provide power, establish bias, return current to ground, or control some other aspect of operation, but, we suppose, stand outside the real action.
The appeal of this picture is easy to understand. Humans are good at reasoning about transport. We instinctively ask where something enters, where it goes, and what it passes through on the way. A complicated electrical system becomes much easier to picture when we imagine audio entering one end, traveling through a chain of parts, and emerging from the other.
A tube amplifier, however, is not a transport system in that sense. The guitar does not supply the energy that moves the speaker cone. That energy comes from the amplifier's power supply. The guitar signal influences how the amplifier converts that energy into an output waveform. What matters inside a gain stage is not the passage of a substance through a privileged route, but the interaction of electrical conditions throughout the stage.
A Better Picture
A thermostat offers a simpler and more accurate way to think about this. Heat does not flow through the thermostat, and neither does the warm air produced by the furnace. In most modern heating systems, even the current that powers the furnace does not pass through the thermostat. The thermostat carries only a small control signal, yet its behavior determines whether the furnace starts, how long it runs, and how much heat is delivered to the house.
Suppose two thermostats appear identical from the outside and are set to the same temperature. Inside, they differ slightly. One switches the furnace on when the room drifts half a degree below the setting and off when it climbs half a degree above; the other allows a wider swing before acting. The first runs the furnace in short, frequent cycles, the second in longer, less frequent ones. The rhythm of heat delivery differs, and the house warms differently, even though no heat, no warm air, and no heating current ever passed through either thermostat. The thermostat matters because it sets a condition that governs the operation of the system.
At this point someone may object that the comparison is inapt because a furnace does not have a signal path.
Exactly! Neither does a tube amplifier!
A signal path is not a physical feature of an amplifier. It is a line that connects blocks on a block diagram. The line tells us how information moves between functional sections, but it does not exist inside the circuit any more than a pastel-pink stripe exists on the highway depicted by a road map. Open one of the black boxes and you will find no "signal path," because that term was never a description of what happens inside the box.
Part of the trouble is that electrical current does not travel from one place to another the way water moves through a pipe. Current flows in complete loops. Trace the signal currents in a real gain stage and you find them circulating not just through the coupling capacitor but through the cathode circuit, through the power supply's filter capacitors, through the ground returns — through components the block diagram never mentions. There is no single route from the input jack to the speaker. The line on the block diagram is a cartoon of the amplifier's organization, not a route that exists inside it.
And the thing imagined to be flowing along that route is not really a thing at all. When a string vibrates, voltages and currents throughout the amplifier change together, each according to its place in the circuit. No single current or physical route is the "signal." The "signal" is the coordinated pattern of changes, not a substance flowing from one end of the amplifier to the other.
A triode, then, is better understood as a control device than as a segment of pipe. A small change in the electrical conditions at its input controls a much larger transfer of energy from the power source to the load. The thermostat does not heat the house; it governs the machinery that does. The guitar does not drive the speaker directly; it governs the amplifier's conversion of power-supply energy into motion at the speaker cone. The output depends on the relationships among voltages, currents, resistances, capacitances, and power sources throughout the stage. Once the amplifier is understood in terms of control rather than transport, the division between "signal-path components" and everything else begins to collapse.
The cathode circuit provides an obvious example. If the voltage at the grid rises while the cathode remains fixed, the tube responds one way. If the cathode rises at the same time, the tube responds differently, although exactly the same voltage appeared at the grid. The tube does not respond to the arrival of something at the grid; it responds to the relationship between the grid and the cathode. The components that set the cathode's behavior are as much a part of the stage's operation as anything connected to the grid.
Here the pipe picture turns out to be not just misleading but backwards. The cathode bypass capacitor — the component so often dismissed as lying "outside the signal path" — sits in parallel with the cathode resistor, and at the frequencies where it does its work, nearly all of the stage's varying current returns to ground through it rather than through the resistor. That is what "bypassed" means. Its impedance, and how that impedance changes with frequency, determines how much the cathode voltage moves as the signal changes — and therefore how much the stage amplifies.
No one familiar with tube circuits disputes this. Change a cathode bypass capacitor from 25 µF to 0.68 µF and the gain-versus-frequency behavior of the stage changes. The smaller capacitor presents a higher impedance at low frequencies, so more of the varying current develops voltage across the resistor, the cathode moves more, and the stage responds accordingly. The capacitor plainly affects the output because its electrical behavior helps establish the conditions to which the tube responds. Once that much is accepted, the interesting question is no longer whether the bypass capacitor matters, but which aspects of its behavior matter and under what circumstances.
The Capacitor Symbol Is Another Map
The same mistake appears again at a finer level of detail. The capacitor symbol is a map too, and like the block diagram, it conceals more than it records. It represents an idealized capacitance described by a value such as 0.68 µF or 25 µF. That is ordinarily all the information needed to understand the broad function of the circuit, which is why the symbol is so useful.
It is easy to forget that the symbol is still only a picture, and that a real capacitor has more electrical properties than the schematic records. It has equivalent series resistance, leakage, dielectric absorption, voltage dependence, temperature dependence, and parasitic inductance. Its behavior may change with frequency, recent charge history, age, and construction. Different capacitor technologies are not merely different packages surrounding the same ideal mathematical object. They are different physical devices whose departures from the ideal model are not necessarily identical.
The schematic is not wrong for leaving these things out. A road map is not wrong for omitting the locations of potholes or the composition of the asphalt. The omission becomes a problem only when someone treats the simplified symbol as proof that the omitted properties cannot matter. If two capacitors bear the same nominal value, the schematic invites us to represent them with the same symbol. It does not follow that the physical components are identical in every behavior relevant to the circuit.
Granted, not every pair of capacitors of the same nominal value will produce an audible difference; many substitutions will make no practical difference at all. Some differences may be measurable but inaudible, while others may matter only in a particular circuit or near a particular operating condition. But those are questions for analysis, measurement, controlled experiment, and listening.
The distinction is important because skepticism about audio claims can easily turn into dogma of its own. It is reasonable to demand evidence that a claimed difference is audible. It is not reasonable to declare a difference impossible merely because an idealized model predicts none. A model is useful precisely because it leaves out effects assumed to be unimportant, but whether those effects are actually unimportant in a particular application is a question to be investigated, not an answer supplied by the model. A schematic establishes where a capacitor is connected; it does not describe how the physical component behaves there.
The lesson applies throughout a tube amplifier. Power-supply impedance affects sag, compression, recovery, and interaction between stages. Screen-supply behavior influences the operation and clipping of power tubes. Negative feedback alters gain, distortion, bandwidth, damping, and transient response. Ground impedance can produce hum, instability, and unintended coupling. These effects are readily understood when the amplifier is treated as an interacting electrical system, but not when it is imagined as a pipe carrying audio through a privileged chain of parts.
The point is not that every component matters equally, or that every tiny physical difference becomes audible. The point is that the familiar question — "Is it in the signal path?" — cannot tell us whether a component matters. It asks about a line on a block diagram when the answer depends on the behavior of the physical circuit.
A better question is whether the component changes a voltage, current, or impedance that influences the operation of the amplifier. The effect may prove substantial, negligible, or nonexistent in the particular case — but whatever the answer, it will have been reached not by consulting a map that was never designed to answer the question, but by examining the territory.