The version almost everyone reaches for first: a "120 V" wall outlet supplies 120 volts, more or less constantly, with some gentle AC wiggle on top. That picture is wrong by about 40%. The outlet voltage is a pure sinusoid that swings between +169.7 V and -169.7 V every 1/60 of a second. The 120 V figure is the RMS value — a specific mathematical average designed to match the heating power of a 120 V DC source. Once you see what RMS actually computes, the 170 V peak stops being a surprise and becomes the only answer that could make sense.
Why the wrong reading feels right
The label "120 V" appears on the outlet, on the appliance, and in the circuit spec. Nothing on the wall plate says "RMS." The natural reading is that 120 V is the voltage at any given instant, the way "5 V" on a DC power rail means an actual steady 5 V.
A second pull toward confusion is the word "average." You might think: take the average of the sine wave over one cycle — that should give you the representative value. It does not give you 120 V. The average of a full sinusoidal cycle is exactly zero, because every positive half-cycle is canceled by the symmetric negative half-cycle. So "average" fails as a useful measure before it starts.
This leaves two candidates — peak and RMS — and they encode different things.
The correct mechanism
Peak voltage (V_peak) is the maximum instantaneous value the sine wave reaches. For North American household current it is approximately 169.7 V. It is the right number when you are sizing insulation, choosing capacitor voltage ratings, or checking whether a MOV clamp fires. It says nothing directly about how much power the signal delivers over time.
Peak-to-peak voltage (V_pp) is the span from the negative peak to the positive peak: 2 × V_peak, or about 339.4 V for a 120 V RMS circuit. Oscilloscope readings and some signal-processing contexts use peak-to-peak. It is not a power measure.
RMS voltage (V_rms) — root mean square — is the answer to the question: what constant DC voltage would dissipate the same average power in a resistor as this AC waveform? The computation follows the name exactly: 1. Square the instantaneous voltage at every point in the cycle (squaring makes all values positive, so the negative half-cycle contributes power rather than canceling). 2. Take the mean of those squared values over one full cycle. 3. Take the square root to return to voltage units.
For a pure sinusoid with peak V_peak, the mean of the squared values works out to V_peak² / 2, so:
V_rms = V_peak / √2 ≈ V_peak × 0.7071
This is why RMS is the value stamped on appliances and outlets: it is the number that correctly predicts power consumption (P = V_rms² / R), heat generated, and motor torque — all the things a user or engineer actually cares about. The R in that formula is whatever resistance the current actually flows through, which is its own source of error once a circuit gets more involved — tracing where the current genuinely goes in an op-amp's feedback network is a sharper version of the same discipline.
Worked example: the 120 V outlet
Start from the RMS value on the label and recover all three quantities.
Given: V_rms = 120 V, sinusoidal waveform, 60 Hz
Peak voltage:
V_peak = V_rms × √2 = 120 × 1.4142 ≈ 169.7 V
The sine wave swings to +169.7 V and -169.7 V. This is the voltage a component rated "150 V" would see exceeded every half-cycle — a fast path to failure.
Peak-to-peak voltage:
V_pp = 2 × V_peak = 2 × 169.7 ≈ 339.4 V
This is what an oscilloscope connected to the outlet would display as the span from trough to crest.
Average over a full cycle:
V_avg (full cycle) = 0 V
Confirmed: the simple average is useless for power calculations.
Power into a 144 Ω resistive load:
Using V_rms: P = V_rms² / R = 120² / 144 = 14400 / 144 = 100 W
Using V_peak: P_peak = V_peak² / R = 169.7² / 144 ≈ 200 W — double the actual average power
The factor-of-two error from using V_peak instead of V_rms is not a rounding issue; it is the definition. V_peak² / R gives the instantaneous power at the crest, not the average power over the cycle.
How to internalize it
-
Before picking a voltage value, ask what you are computing. Insulation and clamping: use V_peak. Average power, heat, or load current: use V_rms. Screen display or waveform span: use V_pp.
-
Derive the factor, don't memorize it. V_rms = V_peak / √2 because the mean of sin²(θ) over a full cycle is exactly 1/2. Squaring the sine, averaging it, and rooting gives you that 1/√2 factor every time — you can re-derive it in 30 seconds rather than recalling whether it is ÷√2 or ×√2.
-
The same signal, three valid numbers. 120 V RMS, 169.7 V peak, and 339.4 V peak-to-peak all describe the exact same North American outlet. None of them is wrong; they answer different questions.
The phase relationships that emerge when a capacitor or inductor is added to this AC signal — why current and voltage no longer rise and fall together — are a separate layer on top of this amplitude picture; that confusion has its own shape, covered in the ELI the ICE man breakdown.
Check yourself
A 240 V RMS circuit (European household standard) powers a 576 Ω resistive heating element. What is the average power dissipated?
A. 200 W — using V_peak = 240 × √2 and P = V_peak² / R
B. 100 W — using P = V_rms² / R = 240² / 576
C. 50 W — dividing by 2 because "AC is only positive half the time", giving P = V_rms² / (2R)
D. 400 W — using V_pp = 2 × 240 and P = V_pp² / R
Answer: B — 100 W.
P = V_rms² / R = 57600 / 576 = 100 W. V_rms is defined precisely so that this substitution gives correct average power. Distractor A uses V_peak² / R = (240√2)² / 576 = 115200 / 576 = 200 W — double the true average, because V_peak² / R gives instantaneous power at the crest, not the cycle average. Distractor C halves the result on the false premise that power flows only during positive half-cycles; both half-cycles deliver power to a resistor because power goes as V², which is always positive. Distractor D treats 240 V as a peak value and doubles it to get V_pp, then squares — stacking two factor-of-two errors to reach 400 W.
Close the gap
The RMS / peak confusion persists because the label "120 V" is never annotated in everyday life. You see the number, assume it is a steady value, and move on — the error only surfaces when a component fails at a voltage that "should" have been safe, or when a power calculation comes out wrong by a factor of two.
A tutor that watches you reason through a waveform problem can catch the exact moment you reach for V_peak where V_rms belongs — before the wrong factor propagates through the rest of your work. That catch, in the moment the reasoning forms rather than after the answer is wrong, is what changes the mental model permanently.