← Back to ElectroLab
Electronics & RFIEEE Std 100 • MIL-STD-348 Practice

RF Attenuator Pad Calculator

Pick a topology, target attenuation and impedance to get resistor values, standard-value loss and return loss, and power per resistor.

K = 10dB/20  •  T: R1 = Z0(K−1)/(K+1), R2 = 2Z0K/(K²−1)  •  Pi: Rsh = Z0(K+1)/(K−1), Rs = Z0(K²−1)/(2K)
Calculated Result
—

Step-by-step

  1. Enter valid values to begin.

Ideal lumped resistors: no parasitic inductance or capacitance, source and load both equal to Z0. Dissipation is computed with the ideal values at a matched input. Standard values are nearest in ratio within the chosen series; the actual loss and return loss come from a full two-port nodal simulation of the standard-value network, not an estimate. Above about 40 dB, or at UHF and above, resistor parasitics, leakage around the pad and layout dominate, so measure with a network analyzer.

IEEE Std 100 • IEC 60063 • IEC 61169 • RF Pads

RF Attenuator Pads: Resistors That Keep the Match

Core Engineering Principles

A pad is three or four resistors arranged so the network looks like Z0 from both ports while it burns a fixed number of decibels. The input stays Z0 only if the series and shunt values sit in exactly the right ratio,. We start from the voltage ratio K = 10dB/20. A T-pad has series arms of Z0(K−1)/(K+1) and a shunt of 2Z0K/(K²−1). The Pi-pad is its dual. The bridged-T uses two fixed Z0 series arms plus a shunt of Z0/(K−1) and a bridge of Z0(K−1).

Heat and parasitics are what bite in practice. In a T-pad the first series resistor takes the most power; in the 10 dB example below it dissipates more than half of the input. A pad also fixes bad loads: it improves the return loss you see by about twice its attenuation, so a 6 dB pad turns a nasty 6 dB return loss into roughly 18 dB. Past about 40 dB, though, a single pad stops behaving. Capacitance across the body and lead inductance in the shunt leg let signal leak around the resistors, and at 40 dB the T-pad shunt is only 1 Ω.

K = 10dB/20  •  T: R1 = Z0(K−1)/(K+1), R2 = 2Z0K/(K²−1)
Pi: Rsh = Z0(K+1)/(K−1), Rs = Z0(K²−1)/(2K)
Bridged-T: Rbridge = Z0(K−1), Rshunt = Z0/(K−1)

NEC & Standard References

IEEE Std 100 defines attenuator, insertion loss and return loss as used here. IEC 60063 defines the E24 and E96 preferred-value series that decide how close a real pad gets to the ideal numbers. IEC 61169 covers the RF connector families (SMA, N, BNC) fitted to fixed attenuators, and MIL-STD-348 gives connector interface dimensions for military-style hardware.
Worked Example: 10 dB T-Pad in a 50 Ω Bench Line
Given: T-pad, 10 dB, Z0 = 50 Ω, 1 W input, E96 resistors, 2× headroom.
1. K = 1010/20 = 3.162.
2. R1 = 50 × 2.162 / 4.162 = 25.97 Ω; R2 = 2 × 50 × 3.162 / 9.0 = 35.14 Ω.
3. Nearest E96 values 26.1 Ω and 34.8 Ω simulate to 10.065 dB with Zin = 49.98 Ω, a return loss of 73.9 dB.
4. At 1 W in: input arm 0.519 W, shunt 0.329 W, output arm 0.052 W, total 0.900 W. The remaining 0.1 W reaches the load.
5. Hottest part 0.519 W × 2 = 1.04 W, so fit a 2 W resistor in the input arm.
Safety & Installation Rules
  • Check the power before connecting. A 1 W pad on a 100 W transmitter vaporises in seconds. Use the peak envelope power for SSB and pulsed signals, not the average.
  • Derate for temperature. Resistor ratings are usually quoted at around 70 °C ambient and fall off above it, so check the datasheet before trusting a half-watt part in a warm enclosure.
  • Keep shunt leads short. Lead inductance in the shunt leg wrecks attenuation above a few hundred megahertz; use chip parts and a solid ground.
  • Don’t stack pads for the last 20 dB blindly. Leakage, ground loops and cable braid transfer limit what you measure, so shield each section.