Why are power transistors big

The transistor as an electronic switch

Transistors work in amplifier circuits with a defined operating point around which an input signal is amplified largely linearly and without distortion. In a second operating mode, there is a change between two operating points, the transistor having the function of an electronic switch. Its two operating states alternate between fully conductive, the switch is closed and non-conductive, the switch is open. Switching takes place via a small control signal on the base. Seen in this way, the transistor is a switching amplifier to switch low to medium power without contact.

As an electronic switching element, the transistor is inferior in some respects to a mechanical switch. The ohmic resistance of a closed mechanical switch is practically 0 Ω. No voltage can be measured across the switch contacts and there is no power loss at the switch. The maximum conductive collector-emitter path, on the other hand, has a very low resistance of around 3 ... 5 Ω. A small collector-emitter saturation voltage of 0.1 ... 0.5 V remains on the "closed" switch path. The transistor therefore does not switch without loss of power. The open mechanical switch has an infinitely high resistance because it completely interrupts the circuit. The non-conductive collector-emitter path has an extremely high resistance with a few 100 kΩ to 10 MΩ, but does not represent an absolute interruption.

The advantages of the transistor as a switch include its bounce-free switching behavior, there is no wear and tear because there are no mechanical contacts. The switching speeds that can be achieved are many times higher, with switching frequencies in the MHz range being possible. The control signal should be square-wave so that, as with a mechanical switch, the switching process can also take place momentarily with the transistor. The mode of operation of the switching transistor can be illustrated with the aid of the output characteristic field.

The following output characteristic field was created for the transistor BC 107 using a simulation program. The diagram is a combination of several individual oscilloscopic measurements for constant base currents. A triangular voltage with an offset of 0 to 2 V was selected as the operating voltage in order to examine the area with UCB = 0 V particularly well. The collector current was determined as a proportional voltage across a very small emitter measuring resistor of 1 mΩ. The measurement error due to the inclusion of the base current and the negative current feedback thus remains negligible. The input voltages for UCB = 0 V were determined by individual measurements.

Operating modes of a transistor

The restricted area

If the base-emitter voltage of an npn transistor is negative or 0 V, no base current flows and there is an intact junction between the base and emitter. Since there is also a barrier layer for the base-collector path, no collector current flows and the collector-emitter path has a high resistance. The transistor is blocked, its operating state corresponds to that of an open switch. The transistor is at operating point A1 in the diagram.

The not overdriven area

If the positive control voltage between the base and emitter is high enough, the junction of the base-emitter diode is broken down so that the base current flows. The emitter injects a large number of electrons into the thin base zone and reaches the base-collector barrier layer. The strong electric field there pulls the electrons through the blocking zone in the direction of the collector and causes a flow of electrons from the emitter via the base zone to the collector. The collector-emitter path becomes lower as the current flow increases. This process is shown in the video clip. The operating point moves from A1 in the direction of A2. Within this range the transistor acts as a signal amplifier. There is still an intact exclusion zone between the base and the collector. The working point is now just below A2 at the intersection of the working line with the output characteristic for IB = 500 µA.

The state of saturation

If the base current is increased further, the collector current increases and the voltage at the collector-emitter path decreases. At operating point A2, the blocking zone of the base-collector diode is so saturated with electrons that the voltage across this diode path is UCB = 0 V. The junction has just broken down and the transistor is in the saturation state. The associated basic saturation current can be specified in data sheets. At operating point A2, the transistor can be compared to the operating state of a closed switch. The collector-emitter path is now very low-resistance. The saturation voltage UCEsat at the beginning of the overdrive range is equal to the UBE. It is not a constant value. In data sheets it is shown in the diagram for certain B values ​​and crystal temperatures as UCEsat = f (IC).

The override area

A further increase in the base current shifts the operating point from A2 in the output characteristic field shown above into the yellow area to A3. The transistor works there in the overload range. The residual voltage at the collector-emitter path assumes the lowest UCEsat value in the knee area of ​​the characteristic curves. The override limit has been reached and can no longer be influenced by further increasing the base current. All characteristics converge on the left.

Transistors as electronic switches are almost always operated overdriven. An override factor is defined for better characterization. It is the ratio of the current base current to the base current value that flows at the start of saturation for UCB = 0 V.

For the working point A3 in the characteristic field above, the override factor is calculated as ü = 1.3 mA / 0.55 mA = 2.4. As the overdrive increases, the transistor switches to the conductive state more quickly, whereby its switch-on time is shortened. At the same time, the switch-off time is extended because the semiconductor crystal is flooded with charge carriers when it is overdriven. The blocked state can only be set after the load carriers have been cleared and the blocked zones have been set up.

Switching operations on the unloaded transistor switch

The following figure shows the transfer characteristic of the BC 107 dynamically created by the simulation program for the given dimensioning. A triangular voltage (ramp) was selected as the input signal. With input voltages Ue between 0 ... 0.6 V, the transistor is blocked and the operating voltage is present as a high signal at the collector. If the input voltage continues to increase, the output characteristic falls largely linearly. For Ue> 1 V the transistor is completely conductive and the collector shows a low signal. The steepness of the characteristic depends on the selected transistor and the dimensioning of the two resistors. The transistor in the basic emitter circuit inverts the control signal.

The resistance values ​​can also be calculated. The collector current should be 10 mA. According to the data sheet, the mean base current gain is B = 200. The override factor should be ü = 1 here. According to the diagram, the minimum input voltage for safe switching is 1.25 V.

Switching times of the transistor as a switch

An ideally rectangular control signal at the input does not generate a similar output signal. The switching transition is fluid and leads to different switch-on and switch-off times, which are influenced by the override factor. Depending on the transistor type, the switch-on times are between 5 ... 500 ns and the switch-off times 50 ... 1000 ns. The switch-on time decreases with increasing overdrive, but the switch-off time increases.

Switch-on process

The base current IB1 should flow from the point in time t1. After the depletion of the junction of the base-emitter diode, collector current begins to flow delayed by time td. The delay time is the time until the collector current has reached 10% of its maximum value. This is followed by the injection of charge carriers into the exclusion zone of the base collector diode with a rapid increase in the collector current. The time after which the collector current has reached 90% of its final value is referred to as the rise time tr. The switch-on time is calculated from the sum of both times: tein = td + tr. The manufacturer specifies the switching times of the semiconductors in the data sheets.

Switch-off process

At time t2, the control current is switched off at the base. The semiconductor crystal is flooded with charge carriers, so the blocking process is delayed. The graphic shows the case of overdriven operation. The collector-base section is conductive and a clearing current IB2 flows to the base. After this stored charge has been depleted, the collector current decreases. The storage time ts is the time until the collector current has fallen to 90% of its maximum value. For transistors that are operated in the saturation state and not overdriven, ts ≈ 0 can be set. The time after which the collector current has fallen to 10% of its maximum value is called the fall time tf. During this period, the barrier layer of the collector base diode regenerates and the electrons still injected there flow away. The switch-off time is the sum of both times: toff = ts + tf.

Optimizing the switching times

The switching times of the transistors are determined by their design. Switching transistors specially developed for switching functions are characterized by particularly short switching times. Switching on to the conductive state is accelerated by overdriving with an increased base current. The associated extension of the switch-off process into the locked state is disadvantageous. If, after switching on, the override is reduced so far that the operating point at the saturation point comes to lie on the line UCB = 0, the switch-off time is shortened by the storage time ts. The collector base line is no longer conductive at this point.

A dynamic shift in the operating point is achieved using a capacitor connected in parallel to the base series resistor. The value of the base series resistor is to be calculated for the saturation operating point. The capacitor is discharged before the control voltage is applied. When the capacitor is switched on and the capacitor is still low, a high charging current flows to the base and leads to the transistor being overdriven. After the capacitor is charged, a lower control current flows through the base series resistor and keeps the operating point at the saturation point. At the time of switch-off the capacitor is still charged, whereby the polarity directly at the base is now more negative. In addition, the discharge current of the capacitor flowing from the base accelerates the locking process and thus shortens the switch-off time.

The switch-off process can also be shortened statically from the overdrive state. At the time of switch-off, the control signal at the base-emitter path is not driven to zero, but rather with an oppositely polarized signal that accelerates the process of removing the charge carriers from the crystal zones.

Switching behavior for different output loads

Switching with ohmic load

There are only ohmic resistances in the load circuit. This applies to the working resistance and to other subsequent stages connected directly to the collector. If the collector is loaded with an effective resistance, the output voltage of the collector-emitter path is divided according to the resistance ratio, whereby only the steepness of the working straight line in the output characteristic curve field changes. The description applies to the unloaded output, where only the working resistance in the collector circuit determines the position of the straight line resistance.

In the non-activated blocking state, the operating point is at A1, the output voltage is equal to the operating voltage and no collector current flows. When switching to the switch-on state, the collector current and thus the voltage at the load resistor increase. The remaining voltage of the collector-emitter path is reduced to its lowest saturation value. The working point moves along the resistance line to the end point A2 or further to the left in the event of overdriving.

During the switchover, the range of maximum power dissipation above the blue power dissipation hyperbola is passed through. In normal amplifier operation, the working line must not cross this area. In switching mode, the operating points must lie outside the forbidden power loss zone and the range can be passed through quickly. As long as the switching times are short enough, the power loss and the associated heating of the transistor can be tolerated. The following equation can be used to calculate the average power dissipation. It neglects the already very small power losses in the blocked state with full control. It refers to the entire duration of the switching period and not just to the sum of the two switching times.

The maximum power loss for a purely ohmic load is the same for the switch-on and switch-off process. The course of the power loss can be calculated along the resistance line from the values ​​of UCE and IC. In the diagram above, the maximum value is in the middle at 10 V and is Pmax = 455 mW. The average power loss must not exceed the maximum power loss Ptot specified by the manufacturer for a longer period of time.

Switching with capacitive load

The working resistance is a parallel connection of active resistance and capacitor. The exit should remain unencumbered when viewed. Compared to the switching times of the transistor, the charging and discharging times of the capacitor should be long. The transistor is blocked and the operating point is at A1.

The transistor is controlled to be conductive and, due to its short switching times, reaches saturation, while the state of the capacitor in the load circuit can still be viewed as uncharged. At this point in time, the capacitor with its resistance of almost 0 Ω determines the working resistance in the RC parallel circuit. In the output characteristic field, the working line (green) therefore points vertically upwards. The working point moves upwards from A1 to point A11. The existing collector current charges the capacitor and as the capacitor voltage increases, the voltage at the collector-emitter path decreases. The operating point moves along the base current characteristic to the saturation point A2.

Switching to the blocking state is also done so quickly that the capacitor voltage is still present, but the transistor is already blocked. At this point in time, the voltage of the collector-emitter path cannot yet assume the value of the operating voltage, but remains at the low saturation value. If there is no collector current, the operating point moves to point A21. The capacitor discharges through the parallel collector resistor. The operating point moves back on the base current characteristic for IB = 0 mA to the starting point A1.

The smaller the capacitance of the load capacitor or the time constant of the RC parallel connection, the flatter (light blue) the curve that the operating point describes when switching over. The blue and green curve dashed in the picture is the simulation result for the two capacitor values. It is the oscillographic representation for UCE = f (UBE) with UBE = 0 V and 4 V switched alternately and is in accordance with the functional description given above.

The course of the switching curve shows that the switch-on process remains within the power loss zone for a long time. Every capacitive load is associated with a high power-on loss line. The expected turn-off power loss is significantly lower. The following equations can be used to calculate the two peak values.

Pe max = 0.8 U ICmax Pa max = 0.2 U 0.2 ICmax ⇒ Pa max = 0.04 U ICmax

Switching with inductive load

The working resistance is a series connection of active resistance and inductance. No other load resistor is connected to the collector. The inductance of the coil and thus the charging time to build up the complete magnetic field should be large. In comparison to this, the switching times of the transistor are then negligible. The transistor is blocked, the operating point is at A1.

The transistor is controlled to be conductive and, due to its short switching times, reaches saturation while the coil has not yet built up a magnetic field and thus offers a very high resistance to the change in current. The operating voltage is divided between the RL working resistance, which is still high, and the RCE, which is low in the connected state. The working point therefore moves horizontally to the left to point A11. The coil can build up its magnetic field via the conductive transistor and is then low-resistance for the operating voltage. The maximum collector current flows, the operating point A2 has been reached.

Switching to the locked state is also very quick. The collector-emitter path becomes high-resistance. The magnetically charged coil counteracts the sudden change in the collector current and in turn drives an almost constant collector current. Since RCE has a high resistance, UCE also increases up to the level of the operating voltage and at the same time the coil generates a self-induction voltage. During this time, the working point moves almost horizontally to the right to A21.

The more magnetic energy is stored in the coil and the faster the current flow is prevented, the higher the self-induction voltage. It is added to the operating voltage. By reducing the magnetic field, the collector current driven by it decreases until the working point has again reached the starting point A1 with the magnetic field energy reduced.On the way from A21 to A1, the permissible collector-emitter voltage can be exceeded by far and destroy the transistor due to the high self-induction voltage without protective measures. The smaller the ratio of L to RC, the closer the characteristic curves to be traversed to the real resistance line. Lower inductance values ​​produce lower self-induction voltages.

A so-called freewheeling diode is connected in parallel to the inductance as a suitable protective circuit. For the operating voltage, it is switched in the reverse direction and does not affect the switch-on process. If the self-induction voltage exceeds the operating voltage during the blocking process of the transistor, it becomes conductive and limits the collector-emitter voltage to the sum of the operating voltage and the diode forward voltage. The diode must have a sufficiently high pulse current resistance.

The average power loss is made up of a relatively small switch-on power loss and a high switch-off power loss. The following equations can be used to calculate the peak values:
Pa max = 0.9 * U * ICmax Pe max = 0.04 * U * ICmax

Estimation of the mean switching losses

An exact calculation of the average power loss is not easy and often not necessary. A good estimate can be made for all three load cases. The total loss line is the sum of the throughput, blocking and the two switching capacities. The transmission power can be measured well. It is the product of UCE and the maximum collector current ICmax. The switching losses can actually always be neglected if the switching times are short compared to the switching period. The switching times of the switching transistors are in the microsecond range. The blocking losses at operating point A1 are also negligible compared to the transmission losses at A2. As a good approximation, if the switching period is sufficiently large, the following remains for the total power loss:
Pges ≈ Pm ≈ PD = UCE sat · ICmax