Impact Ionization Avalanche Transit Time (IMPATT) Diode

1. Theory

Definition: An IMPATT (IMPact ionization Avalanche Transit Time) diode is a reverse-biased junction diode that produces negative resistance at microwave frequencies by combining avalanche generation with a carrier transit-time delay.
Structure: A narrow high-field avalanche region is followed by a wider low-field drift region (e.g. the Read $P^{+}$ -N-I- $N^{+}$ structure, or simpler single- and double-drift profiles).
Avalanche phase delay: Avalanche current builds up as the time integral of the field, so it lags the AC voltage by about $9 0^{\circ}$ .
Transit-time delay: Generated carriers then cross the drift region at the saturation velocity, adding a further phase lag set by the drift-region length.
Negative resistance: With a transit angle near $18 0^{\circ}$ , the total lag is about $27 0^{\circ}$ , so the AC current is out of phase with the AC voltage and the diode delivers power to the circuit.

2. Working Principle

DC bias: The diode is reverse-biased just at avalanche breakdown, with a DC current set by the external supply.
Avalanche injection: On the positive half of the RF voltage, avalanche multiplication builds up a pulse of carriers that peaks a quarter-cycle later.
Drift: The carrier pulse crosses the drift region during the next half-cycle, inducing terminal current while the RF voltage is negative.
Frequency setting: The operating frequency is set mainly by the drift length, $f_{0} \approx v_{s a t} /2 W_{d}$ .
Efficiency limit: Power dissipated in the avalanche region does not produce RF, so efficiency is usually only 10–30%.
High noise: The random avalanche process makes IMPATTs much noisier than Gunn oscillators.

3. Applications

High-power mm-wave sources: Highest CW output power of any solid-state source at many mm-wave frequencies.
Radar transmitters: Compact pulsed and CW transmitters for military and missile-seeker radar.
Injection-locked amplifiers: Power stages for mm-wave transmitters where transistor amplifiers cannot reach the required power.

Name	Formula	Related Formulae	Variables/Constants
Avalanche breakdown condition (avalanche region)	$\int_{0}^{W_{A}} α_{n} d x = 1$	$M = \frac{1}{1 - ( V / V _{B} ) ^{n}} (see Avalanche Diode page)$	$W_{A} : w i d t h o f t h e hi g h - f i e l d a v a l an c h e (m u l t i pl i c a t i o n) r e g i o n$
Avalanche injection phase delay	$ϕ_{a v} = \frac{π}{2}$		$ϕ_{a v} : inh er e n t 9 0^{\circ} p ha se d e l a y b e tw ee n t h e A C a v a l an c h e c u r r e n t an d t h e A C f i e l d, a r i s in g b ec a u se a v a l an c h e m u l t i pl i c a t i o n b u i l d s u p a s an in t e g r a l o f t h e f i e l d o v er t im e$
Transit angle (drift region)	$θ = ω τ_{d} = \frac{ω W _{d}}{v _{s a t}}$	$τ_{d} = \frac{W _{d}}{v _{s a t}}$	$θ : p ha se an g l e cor r es p o n d in g t o t h e d r i f t - r e g i o n t r an s i t t im e$ $W_{d} : d r i f t r e g i o n l e n g t h$ $v_{s a t} : c a r r i er s a t u r a t i o n v e l oc i t y$
Optimum transit angle (max. negative resistance)	$θ_{o pt} \approx π$	$ϕ_{t o t a l} = ϕ_{a v} + θ \approx \frac{3 π}{2}$	$θ_{o pt} : d r i f t r e g i o n d es i g n e d so t r an s i t t im e \approx ha l f t h e R F p er i o d$
Transit-time-mode oscillation frequency	$f_{0} = \frac{v _{s a t}}{2 W _{d}}$	$(from θ_{o pt} = π)$	$f_{0} : d es i g n / o p er a t in g f r e q u e n cy$
Avalanche resonant frequency (Read / Gilden–Hines)	$ω_{a}^{2} = \frac{2 α ^{'} v _{s a t} J _{0}}{ε _{s}}$	$Re {Z (ω)} < 0 for ω > ω_{a}$	$ω_{a} : na t u r a l r eso nan t f r e q u e n cy o f t h e a v a l an c h e r e g i o n i t se l f$ $α^{'} = d α / d E evaluated at the breakdown field$ $J_{0} : D C bia s c u r r e n t d e n s i t y$
RF conversion efficiency (approximate upper bound)	$η \approx \frac{1}{π} \cdot \frac{V _{D}}{V _{D} + V _{A}}$	$η = \frac{P _{R F}}{P _{D C}}$ $\frac{V _{D}}{V _{D} + V _{A}} \to \frac{W _{d}}{W _{d} + W _{A}} (uniform field)$	$η : D C - t o - R F p o w er co n v er s i o n e f f i c i e n cy, t y p i c a l l y 10 - 30% in p r a c t i ce$ $V_{D}, V_{A} : v o l t a g e a cr oss t h e d r i f t an d a v a l an c h e r e g i o n s$ $W_{d}, W_{A} : d r i f t an d a v a l an c h e r e g i o n w i d t h s$

Important Questions

Ques: Why does an avalanche diode need a separate drift region to produce negative resistance at microwave frequencies, when the Avalanche Diode page already covers avalanche multiplication?

Avalanche multiplication alone — the subject of the Avalanche Diode page — is used there as a DC/low-frequency breakdown mechanism for voltage referencing, not as an AC negative-resistance mechanism. The avalanche process itself only provides an inherent $π /2$ phase delay between the AC current and AC voltage, which is not enough on its own to produce negative resistance (that requires the current to lag the voltage by more than $π /2$ ). An IMPATT diode adds a drift region specifically to supply a second, tunable phase delay: carriers generated by avalanche multiplication must then drift across this low-field region before reaching the output contact, and sizing that region so the added transit angle $θ$ is close to $π$ pushes the total phase lag to roughly $3 π /2$ — enough to make the real part of the diode’s small-signal impedance negative over a usable frequency band. This combination of avalanche generation plus transit-time delay is exactly what the name IMPATT (IMPact ionization Avalanche Transit Time) describes. Reference: W.T. Read, “A proposed high-frequency, negative-resistance diode,” Bell Syst. Tech. J. 37, 401 (1958); S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: Why is drift region length the primary design lever for IMPATT frequency, and how does this differ from the role device length plays in a Gunn diode?

In both devices, physical length sets the oscillation frequency through a transit-time relationship, but the underlying negative-resistance mechanism is completely different. In a Gunn diode, length sets how long a real-space charge domain takes to cross the device, with frequency set simply by $f_{0} = v_{d o m} / L$ . In an IMPATT diode, drift region length is chosen specifically to tune the phase delay of the AC current relative to the AC voltage to the optimum value near $3 π /2$ total — the transit time doesn’t carry a physical charge domain so much as it delays the arrival of avalanche-generated carriers at the terminal by the right fraction of an RF cycle. The two devices arrive at similar-looking $f_{0} \propto v_{s a t} / L$ design formulas for very different physical reasons: one from bulk transferred-electron domain propagation (see Gunn Diode page), the other from engineered AC phase relationships around a junction’s avalanche process. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Ques: Why do IMPATT diodes have relatively low efficiency (typically 10–30%) despite being capable of very high microwave output power?

A significant fraction of the DC input power is dissipated directly in the avalanche region through impact ionization and carrier heating without contributing to the time-delayed AC current that actually does useful work in the external circuit; only the voltage swing developed across the drift region effectively converts into RF output. This geometric partition between the (necessary but largely “wasted” for RF purposes) avalanche region and the (RF-productive) drift region is captured directly in the efficiency formula’s $W_{d} / (W_{d} + W_{A})$ ratio, and is the central physical tradeoff IMPATT designers balance: a longer avalanche region can support higher breakdown voltage and power handling, but at the direct cost of conversion efficiency. Reference: S.M. Sze and Kwok K. Ng, Physics of Semiconductor Devices, 3rd ed., Wiley, 2007.

Quartz 5

Explorer