Integrated Gate-Commutated Thyristor (IGCT)

1. Theory

Definition: An integrated gate-commutated thyristor (IGCT) is a GTO-derived press-pack thyristor that turns off by switching the entire anode current out through the gate, with its gate drive built into the package.
Same latching core: It shares the PNPN structure, junctions and two-transistor gains of the SCR and GTO; what is new is unity-gain “hard” turn-off.
Unity turn-off gain: The reverse gate current equals or exceeds the anode current ( $β_{o f f} \approx 1$ ), so the cathode current drops to zero before the anode voltage rises.
Integrated gate unit: Kiloampere gate currents must ramp in about 1 $μ$ s, which needs a gate-loop inductance in the nanohenry range; hence the gate driver is integrated into the package.
Buffer layer: An N-buffer gives a trapezoidal field, so the $N_{1}$ base can be about half as thick as a non-punch-through design for the same blocking voltage.
Transparent anode: A thin, lightly doped anode emitter keeps injection low so stored charge sweeps out quickly.

2. Working Principle

Turn-on: A gate pulse latches the device like an SCR, giving thyristor-like low on-state voltage.
Commutation: At turn-off the gate unit diverts the full anode current into the gate within about a microsecond.
Open-base PNP state: With zero cathode current the NPN half is fully off, so the device behaves as an open-base PNP that cannot re-latch.
Snubberless turn-off: Because regeneration cannot reform, the large turn-off snubber needed by a GTO is not required.
Low losses: The thin buffered base gives low on-state voltage and little stored charge, shortening the tail.
Unidirectional blocking: Buffered designs block in one direction only; reverse current is carried by an antiparallel or integrated diode.

3. Applications

Medium-voltage drives: Multi-MW industrial drives (mills, pumps, compressors) at several kV.
Grid interties and FACTS: STATCOMs, frequency converters and railway grid interties.
Wind and marine converters: High-power converters where low conduction loss and press-pack ruggedness matter.

Name	Formula	Related Formulae	Variables/Constants
Gate-commutation condition (unity turn-off gain)	$I_{GR} \geq I_{A}, β_{o f f} = \frac{I _{A}}{I _{GR}} \leq 1$	$I_{K} = I_{A} - I_{GR} \to 0, β_{o f f}^{GT O} = \frac{α _{2}}{α _{1} + α _{2} - 1} (see Gate Turn-Off Thyristor page)$	$I_{GR} : r e v er se g a t e c u r r e n t$ $I_{A}, I_{K} : an o d e an d c a t h o d e c u r r e n t s$ $β_{o f f} : t u r n - o f f g ain (a GT O t y p i c a l l y a c hi e v es ab o u t 3 - 5, an I GC T f or ces i t t o ab o u t 1)$
Open-base PNP state after commutation	$I_{A} = \frac{I _{C B O 1}}{1 - α _{1}}$	$I_{C 2} = α_{2} I_{K} \to 0$	$I_{C B O 1} : co l l ec t or - ba se l e ak a g e o f t h e P N P ha l f$ $(with the cathode current forced to zero the NPN half is off and the device is an open-base PNP, so no regenerative latch can reform)$
Required gate current slope	$\frac{d I _{G}}{d t} = \frac{I _{A}}{t _{c}}$	$t_{s} \approx \frac{Q _{pB}}{I _{GR}} (see Gate Turn-Off Thyristor page)$	$t_{c} : co mm u t a t i o n t im e in w hi c h t h e w h o l e an o d e c u r r e n t i s t r an s f er r e d t o t h e g a t e (t y p i c a l l y ab o u t 1 μ s)$ $Q_{pB} : s t or e d c ha r g e in t h e p_{2} ba se$
Gate-loop inductance limit	$\frac{d I _{G}}{d t} = \frac{V _{G U} - V _{G K}}{L _{G}}$	$L_{G, ma x} = \frac{( V _{G U} - V _{G K} ) t _{c}}{I _{A}}, V_{G K} ≲ B V_{J 3}$	$V_{G U} : g a t e u ni t s u ppl y v o l t a g e$ $V_{G K} : r e v er se g a t e - c a t h o d e j u n c t i o n v o l t a g e (l imi t e d b y t h e b r e ak d o w n o f J_{3})$ $L_{G} : t o t a l g a t e l oo p in d u c t an ce (p a c k a g e pl u s g a t e u ni t)$
Base thickness: buffered vs non-punch-through	$W_{b u f} \approx \frac{B V}{E _{cr i t}}, W_{N P T} \approx \frac{2 B V}{E _{cr i t}}$	$B V \approx E_{cr i t} W_{i} (see PIN Diode page)$	$W_{b u f} : n_{1} ba se t hi c k n ess w i t h an n - b u f f er (t r a p ez o i d a l f i e l d, i d e a l l imi t)$ $W_{N P T} : t hi c k n ess w i t h o u t a b u f f er (t r ian g u l a r f i e l d)$ $B V : b l oc k in g v o l t a g e r a t in g$
Blocking voltage	$V_{D R M} ≲ B V_{J 2} \approx \frac{ε _{s} E _{cr i t}^{2}}{2 q N _{D}}$	$(see Thyristor SCR / Avalanche Diode pages)$	$N_{D} : d o p in g o f t h e l i g h tl y d o p e d n_{1} ba se$

Important Questions

Ques: What does “gate-commutated” change compared with a GTO?

In a GTO the gate extracts only part of the anode current ( $β_{o f f}$ of roughly 3 to 5), so the NPN half switches off slowly while the anode voltage is already rising, and a snubber is needed to keep dV/dt in check. In an IGCT the gate drive is made strong and fast enough to carry the entire anode current ( $I_{GR} \geq I_{A}$ , so $β_{o f f} \approx 1$ ). The cathode current therefore falls to zero before the anode voltage rises, the NPN half is completely off, and the device becomes an open-base PNP that cannot regenerate. Without a regenerative loop to retrigger, a displacement current from a fast dV/dt no longer needs to be limited by a turn-off snubber in the same way, which removes the snubber losses that make GTO turn-off so costly. Reference: B.J. Baliga, Fundamentals of Power Semiconductor Devices, Springer, 2008; P.K. Steimer et al., “IGCT — a new emerging technology for high power, low cost inverters,” IEEE Industry Applications Magazine 5(4), 12–18 (1999).

Ques: Why must the gate loop have extremely low inductance, and what does “integrated” mean?

Moving the full anode current $I_{A}$ into the gate in about a microsecond requires $d I_{G} / d t = I_{A} / t_{c}$ , which for kiloampere currents is thousands of amperes per microsecond. The voltage available to drive that slope is only $V_{G U} - V_{G K}$ , and $V_{G K}$ is clamped by the low breakdown voltage of $J_{3}$ , so $L_{G, ma x} = (V_{G U} - V_{G K}) t_{c} / I_{A}$ comes out in the nanohenry range. A conventional GTO gate cable has far too much inductance for this. The IGCT therefore builds the gate drive unit directly into the device package with a multilayer low-inductance connection to the wafer, which is the “integrated” in its name. Reference: P.K. Steimer et al., IEEE Industry Applications Magazine 5(4), 12–18 (1999).

Ques: Why does the IGCT use a buffer layer and a transparent anode?

A buffer (punch-through) design lets the electric field in the $n_{1}$ base take a trapezoidal shape instead of a triangular one, so the same blocking voltage is supported by a base roughly half as thick ( $W_{b u f} \approx B V / E_{cr i t}$ versus $W_{N P T} \approx 2 B V / E_{cr i t}$ in the ideal limit; compare the PIN Diode page, where $B V \approx E_{cr i t} W_{i}$ for a uniform field). A thinner base lowers both the on-state voltage and the stored charge that must be removed during turn-off. The thin, lightly doped transparent anode keeps the anode emitter efficiency low so that carriers can sweep out quickly, shortening the tail. The design trade-off is that the buffered structure typically blocks voltage in one direction only, so reverse conduction is handled by an antiparallel (or integrated) diode. Reference: B.J. Baliga, Fundamentals of Power Semiconductor Devices, Springer, 2008.

Quartz 5

Explorer