All electrical systems, including computers, entertainment systems, mobile phones, medical instrumentation and automotive electronics, need to be immune from damage due to electrical transients encountered on a day-to-day basis. Electrical transients can come from many sources. The best known is electrostatic discharge (ESD) which is created when charged objects such as people, furniture, and other pieces of electronics come into contact with an electronic product.

Other sources of stress can be from power line disturbances and electronic noise generated by the switching of large currents. Automotive electrical systems are particularly noisy electrical environments due to the high currents, switching of inductive loads and long wiring harnesses which combine many different types of signals.  One of the most effective ways to protect sensitive electrical components in a system is the placing of Transient Voltage Suppressors (TVS) on sensitive electrical nodes which are likely to be exposed to electrical transients. On many TVS datasheets the most prominent parameter is the peak power rating. While peak power is an important parameter it can also be misleading. This article will explain some of the important features of TVS devices and show the need to look beyond the peak power value.

How TVS devices protect
TVS devices can be used on inputs, outputs or power supply lines but we illustrate their use in Figure 1 on a system input.  TVS devices work by providing a low resistance path between a sensitive node and ground when the voltage on the node exceeds the intended voltage range for the circuit. This is illustrated in Figure 2 which shows the TVS device’s current-voltage properties in relation to the voltage and current properties of the circuit being protected.

Figure 2 illustrates that the most important properties of the TVS for protecting a sensitive node are the turn-on voltage and the dynamic resistance. If the TVS’s turn-on voltage is too high, or the dynamic resistance too high, the voltage on the sensitive node will exceed the range of safe overvoltage and damage will occur.

If threshold voltage and dynamic resistance are the most important parameters for providing protection, why is the peak power often the most prominently displayed parameter on many TVS datasheets? There are two reasons. Peak power is a measure of how much energy the TVS device can absorb without being damaged. Since TVS devices must not only provide protection but also survive the stress, peak power is a useful parameter when determining if a particular TVS can survive in a particular application. The second reason for the prominence of peak power is that TVS devices are often manufactured as a series of products with different breakdown voltages, but all in the same package format.

The amount of electrical stress that the TVS can survive is determined by physical properties such as the size of the silicon die and the thermal and mechanical properties of the package. For example, one series of TVS products has 31 different nominal breakdown voltages from 6.7 V to 87.7 V with working voltages from 5.0 V to 75 V, but all have a peak power dissipation of 400 W. Similar series of products are available in an SMB package with 600 W peak power and in an SMC package with 1500 W peak power. Peak power is therefore an important parameter that all members of the family have in common, while the breakdown voltage and dynamic resistance vary considerably within the series. To place peak power in better perspective, we need to understand how it is measured and how it relates to the important protection properties of threshold voltage and dynamic resistance.

Peak power dissipation is measured by forcing a specified current waveform through the TVS at progressively higher stress levels while monitoring the current through the device and the voltage across the device. The peak power is the product of the peak measured current and the peak measured voltage for the highest stress waveform which does not damage the TVS device. The most common waveforms used in peak power measurements are 8/20 µs and 10/1000 µs current waveforms. (In the electromagnetic compatibility (EMC) field stress waveforms are often described in an xx/yy fashion. The xx value describes the waveform’s rise time in ?s and the yy value describes the time in µs at which the waveform falls to half of its peak value). Higher peak power does not, however, automatically mean improved protection properties. To understand how peak power relates to protection properties we need to understand how peak power relates to breakdown voltage and dynamic resistance.

TVS parameters and protection
The relationship between breakdown voltage, dynamic resistance and peak power and their importance can be easily understood if we consider two hypothetical TVS devices as illustrated in Figure 3.  The two devices have the same reverse bias turn on voltage of 8.5 V and both have a peak power of 200 W. The dynamic resistance of the two devices are, however, quite different. At the peak power dissipation of 200 W, TVS A has 10 V across it at 20 A while TVS B at 200 W of peak power has 14 V across it at 14.2 A. Both devices have the same peak power but TVS A provides considerably better protection than TVS B.

Comparing the voltage and power versus time between two products for a similar current pulse is a dramatic way to demonstrate how power dissipation can be a misleading indicator of protection capability. Figure 4 and Figure 5 compare input to ground stress for two TVS protection products. Figure 4 shows the measured current and voltage pulses versus time for 8/20 µs pulses. The current pulses are almost identical in size, but the measured voltages are quite different. One device shows a peak voltage of about 10 V, while the other device has a peak voltage over 15 V. In this instance, the ESD1014 will provide better protection during a surge input resulting from a lightning strike or the switching of an inductive load, because of a lower transient voltage on the line being protected. Figure 5 shows the voltage as well as power versus time.

Since power is voltage multiplied by current, similar current levels with different voltages result in the one part dissipating considerably more power for the same current stress. The higher power dissipation for the same current stress provides no protection benefit, but only highlights the higher clamping voltage for that device. Datasheets often emphasize the power-dissipation capability of the protection device, but it should be clear that here is an example where higher power dissipation is hardly an advantage – it is the result of higher (worse) transient voltages for a given current surge.

There are two important properties of a TVS device during an electrical stress. The TVS must be able to self-protect and survive the stress current, and the TVS must clamp the protected line to a low voltage during the stress. Power ratings of TVS devices do not provide a good measure of clamping effectiveness because this metric is artificially enhanced by a higher, less desirable, clamping voltage. It is true that TVS devices with high peak power ratings are often larger diodes and may therefore also have low dynamic resistance and low clamping voltage but there are exceptions, and one such exception has been highlighted in this note. It is therefore important to always consider the dynamic resistance or the clamping voltage at specific current levels, rather than focusing solely on the peak power rating when selecting a TVS device.