How this instrument works
A Pi attenuator is a simple three-resistor network — one resistor in series between input and output, and two shunt resistors, one to ground at each end — that reduces a signal's power by a known, fixed amount while keeping the impedance looking into and out of the pad matched to the rest of the system (commonly 50 ohm in RF work). Its shape, drawn out, resembles the Greek letter pi (π), which is where the name comes from; the T attenuator is its dual, arranged the opposite way.
The design boils down to one intermediate number: K, the voltage attenuation ratio, computed from the target attenuation in decibels as K = 10^(dB/20). Once you have K, both resistor values follow directly from Z0 and K alone — no iteration or lookup table needed. Higher attenuation (bigger dB, bigger K) produces a bigger series resistor and a smaller shunt resistor, which makes physical sense: more attenuation means blocking more of the signal power in the series path while shunting more of it to ground.
Pi pads like this are routine building blocks in RF test setups and signal chains — padding down a signal generator's output to protect a sensitive input, setting a known, repeatable loss for calibration, or improving impedance match between two mismatched stages by absorbing reflections in the pad's own resistors rather than passing them back and forth.
- Enter Reference impedance Z0 in ohms — the system impedance the pad is matched to (50 ohm is standard for most RF work, 75 ohm for video/cable systems).
- Enter Attenuation in dB — the amount of signal loss you want the pad to introduce.
- Read Series resistor — the single resistor placed in the signal path between input and output.
- Read Shunt resistors — the value used for both shunt-to-ground resistors, one at each end of the pad.
Worked example — a standard 50-ohm, 10dB pad
A common 50-ohm, 10dB pad: K = 10^(10/20) = 3.16228. Series resistor R_series = 50 x (3.16228^2 - 1) / (2 x 3.16228) = 50 x 9 / 6.32456 = 450 / 6.32456 = 71.15 ohm. Shunt resistors R_shunt = 50 x (3.16228 + 1) / (3.16228 - 1) = 50 x 4.16228 / 2.16228 = 96.25 ohm. These reproduce the well-known published reference values for a 50-ohm 10dB pad (96.2 ohm shunt, 71.2 ohm series) almost exactly, a useful sanity check that the series and shunt roles haven't been swapped.
A gentler 75-ohm, 3dB pad (a common video/RF impedance and a light attenuation level): K = 10^(3/20) = 1.41254. R_series = 75 x (0.99531) / (2 x 1.41254) = 26.42 ohm; R_shunt = 75 x 2.41254 / 0.41254 = 438.60 ohm. A light pad correctly produces a small series resistor and a large shunt resistor — most of the signal is meant to pass through with only a little loss, so the series path stays low-resistance and the shunt path stays high-resistance.
Questions
Why does higher attenuation need a bigger series resistor?
Because more attenuation means more of the signal's power has to be dissipated or diverted rather than passed through, and a bigger series resistor drops more voltage (and dissipates more power) across itself for a given current. At the same time, the shunt resistors shrink, diverting proportionally more signal to ground at each end — both changes work together to push more attenuation through the pad while keeping it impedance-matched to Z0 looking in from either direction.
What's the difference between a Pi pad and a T pad?
They're duals of each other, using the same underlying K = 10^(dB/20) relationship but arranged differently: a Pi pad has one series resistor and two shunt resistors (one at each end, to ground), while a T pad has two series resistors and one shunt resistor in the middle. Both achieve the same attenuation and impedance match for equivalent designs; the choice is mostly about which topology is more convenient to build or which one better suits the available resistor values.
Does this pad stay matched to Z0 looking in from both ends?
Yes — that's the defining property of a properly designed symmetrical resistive pad. As long as the resistor values are computed from Z0 and the target dB using these standard formulas, the pad presents Z0 looking into either port, regardless of what's connected on the other side, which is exactly why these pads are used to isolate and match mismatched stages in a signal chain.
Can I build a Pi attenuator with standard resistor values?
Usually only approximately — the exact calculated values rarely land on standard E24 or E96 resistor series values, so real builds pick the nearest available standard resistors, which introduces a small deviation from the exact target dB and from perfect impedance match. For most RF test and calibration uses that small deviation is negligible; for precision work, some designers use resistor combinations (series or parallel pairs) to land closer to the exact calculated value.
How much power can a resistive Pi pad handle?
That depends entirely on the resistors used, not on this formula — the calculator gives you resistance values, not power ratings. At higher signal power, especially in the series resistor which carries the full through-current, standard quarter-watt resistors can overheat quickly; higher-power applications need resistors (or purpose-built attenuator components) rated for the actual power the pad will dissipate.