Potentiometer taper: audio vs. linear

Human hearing does not perceive loudness on a straight line — it responds roughly to the logarithm of sound power, a relationship known as the Weber–Fechner law. Wire a volume pot with a linear taper and the electrical output rises evenly with rotation, but the perceived loudness does not: because the ear's sensitivity itself falls off logarithmically, a linear pot crowds almost all of the audible change into roughly the last quarter of the shaft's travel. The first three-quarters of the knob feels like it is barely doing anything, and then the volume takes off all at once near the top.

An audio (logarithmic) taper pot pre-distorts the resistance curve to cancel that effect, so that a given amount of shaft rotation produces roughly the same felt change in loudness whether the knob starts at 1 or at 8. On a control marked 0–10, a setting of 5 is meant to sound subjectively about half as loud as a setting of 10 — the textbook definition of an audio-taper pot, per the "Audio control" section of Wikipedia's potentiometer article.

A true logarithmic curve is undefined at zero resistance, so it cannot actually be manufactured; ordinary "log" pots approximate it with two straight resistance sections of different resistivity, wound to meet near the middle of rotation — the dog-leg drawn in the chart above. It is a real approximation, not a cosmetic one: the seam where the two sections join is the well-known point where a cheap audio-taper pot can audibly "step," and it is one reason a worn or poor-quality log pot feels scratchy or uneven exactly around noon on the dial. See Wikipedia's "taper" section and Rod Elliott's Beginners' Guide to Potentiometers (Elliott Sound Products) for both points.

None of this applies to every knob on an amp, only to controls that stand directly between a signal and a listener's ear — which is exactly the split the corpus table below turns up.

Two patterns hold across every circuit here. Every Bass control (the 5F4, 5F6-A, and JTM45) is audio-taper, and the Treble controls beside them split: the 5F6-A's factory layout page marks its Treble 250K LIN., while the 5F4 and JTM45 drawings mark no taper on theirs — a genuine gap in those two amps' published documentation rather than an omission on this page's part. The Middle and Presence controls are linear throughout too — the 5F6-A's BOM notes its Middle pot as rheostat-wired (two terminals in circuit, not three), and every Presence control here sits in the negative-feedback path — at the phase-inverter tail, or on the 5F4 at its driver stage's cathode — rather than in the direct signal-to-ear path. Neither job is "how loud does this feel," so neither needs the perceived-loudness curve a volume or tone knob does.

Volume itself is audio-taper in every circuit but two: the 5F10 Harvard's single volume and both pots on the 5F2-A Princeton are marked linear on their published drawings. It is not a bias-topology split — the 5F10 is fixed-bias and the 5F2-A is cathode-biased, like most of the other small amps here — so treat it as exactly what each source drawing says rather than an inferred rule, and, as with any taper claim on a decades-old part, confirm it against the original drawing or a meter before assuming a replacement pot should match.

Potentiometers are often stamped with a single letter for their taper instead of the word — and that letter is not standardized. Manufacturers in the US and Asia generally follow one code; many European makers followed the opposite one for the same letters, and the two "standards" overlap instead of replacing each other cleanly:

So an "A" pot is the exact opposite of an "A" pot, depending on which convention its maker used: modern US/Asian practice reads A = log (audio), B = linear, while the older code some European manufacturers used reads A = linear, with log marked B or C instead — and the exact code still varies by manufacturer even within those groups. This is exactly the ambiguity every BOM entry in this corpus sidesteps by spelling out "Audio-taper" or "Linear" in English rather than reproducing a single stamped letter. It is also why a letter alone is not enough evidence when sourcing a replacement pot for a vintage circuit: confirm the taper against a manufacturer datasheet, or measure the resistance at the shaft's midpoint with a meter, before trusting the stamp.

Old/new code table per Rod Elliott, Beginners' Guide to Potentiometers (Elliott Sound Products); cross-checked against the "taper" section of Wikipedia's potentiometer article. The reference library's era-component shelf carries the period design texts — Terman's Radio Engineers' Handbook and Reference Data for Radio Engineers, both 1943 — that document potentiometer taper law from the vacuum-tube era firsthand.