Analog ElectronicsJuniorcommon

PIN Diode: Microwave Switch and Attenuator Applications

How PIN diodes work as microwave switches and attenuators: I-region physics, Rs vs. bias current, and a worked insertion-loss/isolation example.

7 min readAhmet Zahid ArıcanUpdated 11 Sept 2026
Contents & prerequisites

A PIN diode looks like an ordinary diode on a datasheet cover page, but its behavior at RF and microwave frequencies is fundamentally different from a signal or rectifier diode. Instead of acting as a rectifying junction, a forward-biased PIN diode behaves like a variable resistor controlled by DC current, with essentially no rectification at frequencies above a few MHz. That single property makes it the workhorse device for RF switches, variable attenuators, and phase shifters in radar, cellular front-ends, and test equipment — anywhere a signal path needs to be turned on/off or gained down electronically without mechanical relays.

Structure and Why the "I" Region Matters

A PIN diode is a p-n junction with a lightly doped (near-intrinsic) layer sandwiched between the p+ and n+ regions:

   p+          I (intrinsic)          n+
 ──────┼──────────────────────────┼──────
  heavily      lightly doped        heavily
   doped       high-resistivity      doped

In a normal diode, forward bias injects carriers that recombine within a thin depletion region, and the device still conducts efficiently even at high frequency because the depletion width is small. In a PIN diode, the I-region is much wider (tens to hundreds of µm), so under forward bias, holes and electrons are injected from the p+ and n+ regions and must cross this wide region by diffusion before recombining. The transit time through the I-layer is comparable to or longer than one RF cycle, so the injected carriers don't have time to follow the instantaneous RF waveform — they respond only to the average (DC) bias current. The result: at RF, the diode does not rectify; it simply presents a resistance set by the stored charge from the DC bias.

Forward Bias: The Diode as a Variable Resistor

Under forward DC bias, the I-region resistance Rs is inversely related to the total charge Q of injected carriers stored there:

Rs ≈ W² / [(μₑ + μₕ)·Q]

where W is the I-layer width, and μₑ, μₕ are electron and hole mobilities. Since the stored charge is proportional to forward bias current I_F and the carrier lifetime τ (Q ≈ I_F·τ), increasing I_F increases stored charge, which lowers Rs. A typical PIN diode's Rs vs. I_F relationship follows approximately:

Rs ≈ K / I_F        (K is a device constant, often specified as Rs at a reference I_F)

Practical numbers: many switching PIN diodes exhibit Rs from several hundred ohms near zero bias down to under 1 Ω at tens of mA. This continuous, monotonic resistance control — not a hard on/off transition — is what allows PIN diodes to be used both as switches (drive Rs very low or very high) and as continuously variable attenuators (bias current sets a specific Rs between the extremes).

Reverse Bias: The Diode as a Variable Capacitor

Under reverse or zero bias, the I-region is depleted of free carriers and the diode behaves as a low-loss capacitor Cj in parallel with a very high resistance (megohms), giving high isolation. Cj is small (often well under 1 pF for switching diodes) because the wide I-region acts as a thick dielectric:

Cj ≈ ε·A / W

Because W is fixed by the intrinsic layer thickness (not modulated by reverse voltage the way a varactor's junction width is), Cj stays nearly constant with reverse bias — a key difference from varactor diodes, where capacitance is deliberately voltage-dependent. This stability is desirable for switches (predictable off-state isolation) but is exactly why PIN diodes are not used for tuning applications.

Equivalent Circuit for RF Design

Bias stateDominant elementTypical magnitude
Forward (on)Series resistance Rs0.5 Ω – 10 Ω (varies with I_F)
Reverse/zero (off)Parallel capacitance Cj0.1 pF – 2 pF
Both statesPackage/lead inductance L0.5 nH – 2 nH

A complete small-signal model is a series R-L combined with a parallel C, switched between the "on" (R dominant) and "off" (C dominant) configurations by the bias network. At microwave frequencies, package inductance and pad capacitance are not negligible and must be included in matching network design.

Worked Example: Switch Insertion Loss and Isolation

Consider a series-mounted PIN diode switch in a 50 Ω line, with the diode biased to Rs = 1 Ω (on) or presenting Cj = 0.3 pF (off), at f = 2 GHz.

On-state insertion loss (diode in series, Rs=1 Ω, ignoring L):

Loss (dB) = 20·log₁₀[(Z₀ + Rs/2) / Z₀]  — for a series element, approximate as:
IL ≈ 20·log₁₀[1 + Rs/(2·Z₀)]
   = 20·log₁₀[1 + 1/(100)]
   = 20·log₁₀[1.01]
   ≈ 0.086 dB

Low insertion loss, as expected for a well-designed switch.

Off-state isolation (diode in series, now high impedance from Cj):

Xc = 1/(2π·f·Cj) = 1/(2π · 2×10⁹ · 0.3×10⁻¹²)
   = 1/(2π · 6×10⁻⁴)
   = 1/(3.77×10⁻³)
   ≈ 265 Ω

For a series element with impedance Xc in a 50 Ω system, isolation follows the same form as the insertion-loss formula above (keeping the '+1' term, since Xc/(2·Z₀) is not much greater than 1):

Isolation (dB) = 10·log₁₀[1 + (Xc/(2·Z₀))²]
             = 10·log₁₀[1 + (265/100)²]
             = 10·log₁₀[1 + 2.65²]
             = 10·log₁₀[8.02]
             ≈ 9.05 dB

Check: this modest isolation is realistic for a single series diode — it's why practical RF switches cascade two or more PIN diodes (or use series-shunt combinations) to reach 30–50 dB isolation. The on-state number (0.086 dB) matches typical datasheet insertion-loss specs for a single diode switch at low GHz frequencies, confirming the approximation is reasonable.

Attenuator Configurations

  • Series attenuator: one diode in the signal path; varying I_F sweeps Rs from near-zero (low loss) to high (large loss). Simple but limited dynamic range and poor impedance match at high attenuation (Rs alone doesn't absorb power symmetrically — VSWR degrades as attenuation increases).
  • Shunt attenuator: diode across the line to ground; low Rs shorts the line (high attenuation), high Rs (or reverse bias) leaves the line largely unaffected (low attenuation) — inverse behavior vs. series.
  • Bridged-T / matched attenuator: combines series and shunt PIN diodes with fixed resistors so that input/output impedance stays close to Z₀ across the whole attenuation range — the standard topology for voltage-variable attenuators (VVAs) used in AGC (automatic gain control) loops and receiver protection circuits, because it holds VSWR low even at deep attenuation settings.

Practical Design Considerations

  • Switching speed is limited by carrier lifetime τ and the bias circuit's ability to sweep stored charge in/out; typical PIN switching times range from tens of ns to a few µs depending on diode geometry and drive current — much slower than a Schottky diode but fast enough for most radar/comms switching needs.
  • Power handling: RF power dissipation in Rs, plus the diode's ability to avoid self-modulation ("RF-induced" rectification) at high RF voltage swing, sets the maximum handled power. Higher I_F bias raises the RF power level at which the diode starts to distort, at the cost of higher DC power consumption (and possibly increased self-heating) — note that Rs actually decreases (becomes more negligible) as I_F increases, so the tradeoff is DC power dissipation, not a rising Rs.
  • Bias networks: RF chokes and DC blocks isolate the DC bias path from the RF signal path without loading the RF impedance; layout parasitics at microwave frequencies (bond wire inductance, pad capacitance) must be co-designed with the matching network, not treated as an afterthought.
  • Temperature sensitivity: carrier lifetime and mobility vary with temperature, shifting the Rs vs. I_F curve — attenuator applications requiring precise gain control often need a closed-loop bias scheme or temperature compensation.

Key Takeaways

  • A PIN diode's wide, lightly-doped I-region gives it a transit time long enough that it stops rectifying at RF/microwave frequencies and instead behaves as a bias-controlled variable resistor (forward) or nearly-constant low-loss capacitor (reverse).
  • Forward bias current sets Rs continuously from sub-ohm to hundreds of ohms, enabling both switch (extreme states) and attenuator (intermediate states) operation from the same device.
  • Reverse/zero bias gives a small, roughly bias-independent junction capacitance Cj, providing high isolation in the off state — unlike a varactor, Cj is not meant to be tuned.
  • Single-diode switches typically give sub-0.1 dB insertion loss but limited (~9 dB) isolation; cascaded series-shunt topologies are used to reach 30–50 dB.
  • Bridged-T and similar matched topologies keep VSWR low across the attenuation range, which is essential for AGC and gain-control applications.
  • Switching speed (tens of ns to µs) is set by carrier lifetime and bias drive strength, and power handling trades off against DC bias current and self-distortion at high RF voltage swing.

Learning

Sign in to track your progress.

Evidence

Public projects engineers linked to PIN Diode: Microwave Switch and Attenuator Applications.

Add evidence

No engineer has linked a project to this topic yet. Built something that proves it? Add the project and tag it with analog-electronics-pin-diode-microwave-switch-and-attenuator-applicat — it then shows here and on your public profile.