Power MOSFET

from Wikipedia, the free encyclopedia
Two power MOSFETs in the D2PAK SMD housing. These FETs can switch a current of 30 A.

A power MOSFET ( English power MOSFET , power metal oxide semiconductor field-effect transistor ) is a specialized version of a metal oxide semiconductor field effect transistor (MOSFET) that is optimized for passing and blocking of large electric currents and voltages (up several hundred amperes and up to approx. 1000 volts , with a component volume of around one cubic centimeter).

Power MOSFETs differ from bipolar power transistors in terms of both functionality and efficiency. Some of the benefits of power MOSFETs are fast switching time, no second breakdown, and stable gain and response times. A MOSFET is assigned to the power MOSFETs from a current rating of around 1 A.

history

A drive to the development of power MOSFETs were the weaknesses of the hitherto dominant bipolar power transistors ( English transistor bipolar power junction , BJT). Bipolar transistors require z. B. high control currents up to approx. 1/5 of the load current, while power MOSFETs in principle do not require any control current when switched on or off.

Physical functionality

Power MOSFETs work on the same physical principle as the MOSFETs used in integrated circuits , but they differ in other geometrical shapes and dimensions. The high power density is achieved by a grid-like or honeycomb-like semiconductor structure, which corresponds to a parallel connection of thousands of individual MOSFETs.

The structure of the power MOSFET also corresponds to that of the MOSFET  - however, there are numerous special features. In contrast to the signal transistor in communications engineering, the source and drain are arranged vertically. In the semiconductor structure of the gate , drain and source in a MOSFET, numerous parasitic elements such. B. Resistances, capacitors and diodes. In power electronics , particular attention must be paid to these parasitic elements. The capacities have to be reloaded with each switching process, which leads to considerable switching losses , especially at high switching frequencies . The principle always included diode, which is switched in reverse direction during normal operation, can be used in terms of circuitry like a diode connected in parallel to the drain and source. Depending on the requirements, further effects must be taken into account.

There are mainly three types of power MOSFETs - the DMOS , UMOS , and the VMOS structure. They have a large drain-drift region, which protects the component against breakdown at high blocking voltages.

DMOS field effect transistor

Schematic structure (cross section) of two parallel-connected elements of an n-channel DMOSFET
Square structure of a DMOS cell

MOSFETs with this structure are produced by double implantation of the channel structure and are known as DMOSFETs ( double-diffused metal-oxide semiconductor field effect transistor ). In the NMOS type, the channel is located in the narrow p-area below the gate electrode, see picture.

For power MOSFETs, transistors with a large gain and low on-resistance are required. With DMOS, this can be achieved through a large channel width and parallel connection, and in a compact manner at the chip level. There are primarily square (e.g. SIPMOS ) and hexagonal (e.g. HEXFET ™ from International Rectifier, now Infineon Technologies) structures. The source electrode is at the top as a large-area layer, the gate electrode is buried underneath.

A special feature of the DMOSFET is u. a. that it has  no blocking capability in reverse operation ( V DS <0). The inverse diode of the pn junction between the base and collector of the parasitic npn bipolar transistor is then in forward polarity. This is used when switching inductive loads, here the inverse diode can act as a freewheeling diode and thus short-circuit the high voltage that occurs when the inductive load is switched off. However, it must be noted that the inverse diodes in MOSFETs for higher blocking voltages have significantly higher blocking delay charges and switching losses than conventional fast diodes. Even if special techniques (e.g. FREDFET ) are used to improve the behavior of the inverse diode, these diodes are still worse than separate fast diodes by a factor of about 3. The use of the inverse diode in MOSFETs with blocking voltages greater than 200 V in applications for switching inductive loads at high frequencies is therefore often not possible.

UMOS field effect transistor

The UMOS field-effect transistor (from U-shaped notch MOS field-effect transistor , UMOS-FET, also U-MOSFET) is a vertical power MOSFET structure in which the gate electrode is implemented in a trench etched in silicon from whose U-shape the name derives. Using the trench as a gate electrode creates a vertical channel.

VMOS field effect transistor

Technology section through a VMOS transistor

The VMOS field effect transistor (from English v-groved MOS field-effect transistor ) is a non-planar field effect transistor in which the channel length is reduced and the channel width increased with the help of a V-shaped gate area. The V-shaped trench for the gate area is often produced by anisotropic etching of silicon. The concept was introduced by TJ Rodgers in the mid-1970s. There are variants with both lateral and vertical current flow, with the vertical variant being more common. By making better use of the chip area, VMOS-FETs enable a higher current density and are therefore particularly suitable for use as (discrete) power MOSFETs. They are also characterized by a high input impedance.

Applications

Power MOSFETs are often used in amplifier circuits, as a currentless controllable switch and as a fast switch for pulse width modulation , e.g. B. in inverters , switched mode power supplies , DC / DC converters or motor controls.

The fast current rise time of MOSFETs is an advantage in switching applications. This can reduce the switching losses.

Inductive loads generate large voltage peaks during the switching process, against which the power MOSFET, like other semiconductor switches, must be protected. In the case of MOSFETs, however, this protection can be achieved in the component itself, in that the avalanche effect runs in a controlled manner during breakdown without partially overheating the component structures. Often, however, additional external circuitry ( snubber ) is necessary.

Power MOSFETs, like other power semiconductors, must be cooled at high powers. Since the channel resistance increases with increasing temperature, the power loss increases due to insufficient cooling in addition, it may be characterized to a so-called. Thermal runaway ( thermal runaway ) may occur. In addition, MOSFET structures only tolerate lower maximum temperatures than bipolar structures (approx. 125 to 150 ° C compared to 150 to 180 ° C). The rail resistance, which rises with increasing temperature, is an advantage when several MOSFETs are connected in parallel; it ensures an even, self-regulating current distribution to all components. Emitter resistors for current distribution, as required with bipolar transistors, can be omitted.

Parameters

As with bipolar power transistors is safe working area (Engl. Safe operating area , SOA) determined by three parameters at power MOSFETs:

  • Maximum drain current
  • Breakdown voltage (also ) and the maximum reverse voltage specified by it
  • Maximum power dissipation

Other important parameters:

  • Minimal contact resistance
  • Maximum energy allowed for avalanche breakthrough
  • Amount of charge Q g that is necessary for switching on and off (control losses)
  • Max. Rate of voltage change

Unlike bipolar transistors, MOSFETs can withstand very high values ​​for the rate of voltage change; they do not have to be protected against it by snubber elements. The so-called second breakdown (destruction by small currents at voltages below the reverse voltage) does not usually occur in MOSFETs with low power, provided the power loss is not exceeded. In the case of power MOSFETs, on the other hand, in saturation or linear operation (the drain current is largely controlled by the gate voltage) damage can occur due to local overheating in the semiconductor. The reason for this lies in the negative temperature coefficient of the threshold voltage, which is close to which one is in this operating state. The so-called SOA (Safe Operating Area) characteristic that the manufacturer specifies must be observed.

Further development

In most cases, the drain-source on-state resistance is ( English on-state resistance ) at a given breakdown voltage regarded as an essential feature of power MOSFETs. Basically, the following applies for a given chip area: the higher the maximum reverse voltage of the MOSFET, the higher its forward resistance. With a conventional layer structure, doubling the dielectric strength results in a five-fold increase in the . The chip area also increases with an exponent of 2.4 to 2.6. This relationship is referred to in English as the silicon limit .

In the late 1990s, semiconductor manufacturers made progress in reducing the forward resistance at high voltages, i.e. voltages from around 200 V upwards , which led to the widespread use of power MOSFETs. In this case, additional p-zones are introduced into the n - - epitaxial layer , which is normally thicker due to the higher dielectric strength, which modulate the course of the electric field strength in the blocking state and prevent local breakdowns in the silicon. The n - epitaxial layer can thus be made thinner in structure, which results in a lower sheet resistance even at high voltages.

The specific switch-on resistance varies with the breakdown voltage:

A common value in 2017 is 30 mΩ with a blocking voltage of 250 V in the TO-247 package. With maximum blocking voltages around 50 V, values ​​around a few milliohms (mΩ) are common.

In addition to the general improvement of the robustness against high current and voltage peaks and the reduction of the forward resistance, additional functions are increasingly being integrated into the component. These components are often referred to as "Smart Power Devices" and contain protective circuits (input protection, protection against thermal overload, current limitation, error signal generation), for example. B. level shifting (switching the positive load line with a ground-related signal, so-called high side switches ) or even complete PWM controllers.

Designs

Usual housing designs of power MOSFETs are for through-plated boards, for example TO-264, TO-247 and TO-220, for surface-mountable (SMD) components DPak, D²Pak and SO-8. There are also housings with screw connections for cables or busbars. These include the SOT-227 "Isotope" with four screw connections. The source connection is brought out twice in order to have a more precise reference potential available for gate control.

literature

  • Arendt Wintrich, Ulrich Nicolai, Werner Tursky, Tobias Reimann: Application manual for power semiconductors . ISLE Verlag, 2010, ISBN 978-3-938843-56-7 ( semikron.com [PDF]).

Web links

Individual evidence

  1. Patent US5055895 : Double-Diffused Metal-Oxide Semiconductor Field Effect Transistor Device. Published October 8, 1991 , Applicant: Matsushuta Electric Works LTD, Inventor: Sigeo Akiyama, Masahiko Suzumura, Takeshi Nobe.
  2. Thomas Tille: Microelectronics: Semiconductor components and their application in electronic circuits. Springer, Berlin 2005, ISBN 978-3-540-20422-0 .
  3. TJ Rodgers, JD Meindl: VMOS: high speed TTL compatible MOS logic . In: Solid-State Circuits, IEEE Journal of . tape 9 , no. 5 , 1974, p. 239-250 , doi : 10.1109 / JSSC.1974.1050509 .
  4. International Rectifier Application Note AN-1155
  5. ^ Joachim Specovius: Basic course in power electronics . 4th edition. Vieweg + Teubner, 2010, ISBN 978-3-8348-1307-7 .