Fullscreen HTML5Med Res (??MB)HTML4

Circuit Surgery Regular clinic by Ian Bell Electrical Overstress Protection for Circuits R ecently, Jose Cordero posted a question on the EEWeb concerning protection for sensor circuit inputs. He wrote: ‘I am building with a sensor that will be capable of measuring voltages between 25-40V through the ADC of a microcontroller, and at the same time feeding on these voltages. A fundamental part of every sensor is to have an overvoltage protection circuit, to take care of the microcontroller and any extra components from any overvoltage that happens in the future’. He went on to mention crowbar circuits as potential protection and asked for help on a circuit design. We will not look at this specific circuit but will consider in more general terms the problem of protecting inputs of components such as op amps, analogue-to-digital converters (ADCs) and microcontrollers. This is a common requirement for sensing and measurement circuits, particularly where the inputs may be connected to a variety of sources during use, or where the environment may cause problems; for example, due to issues with power supply lines or the effects of induction or electromagnetic interference. Before looking at input overvoltage protection we will look briefly at circuit protection more generally and a range of devices which are used in protection circuits. We mentioned overvoltage, but in general the term ‘electrical overstress’ (EOS) is used to cover situations where a circuit or device is subjected to levels of voltage, current or power which cause damage. There are many ways in which EOS can occur, from simply connecting a circuit incorrectly to complex powerline noise issues in industrial installations. Devices do not even have to be in a circuit to suffer EOS – it is well known that many semiconductors (particularly Simulation files Most, but not every month, LTSpice is used to support descriptions and analysis in Circuit Surgery. The examples and files are available for download from the PE website. 54 Fig.1. Fuse circuit symbols. integrated circuits) can be damaged by electrostatic discharge (ESD) during handling. For example, the charge built up on a person can be discharged through a device when it is touched, delivering sufficient energy to destroy it. Overcurrent protection devices There are a number of components which are specifically used for protection against EOS. The most well-known of these is the fuse (see Fig.1 for circuit symbol). This is an overcurrent protection device comprising a wire or strip of conductor, designed so that heat generated by the current flowing in it will cause it to melt above a certain rated current. This causes an open circuit to occur, which stops the excessive current. Once a fuse has blown it must be replaced – but, crucially, the cause of the problem must be identified and fixed first. Fuses are typically used in protection and safety roles where EOS will cause damage or danger that cannot easily or cheaply be prevented by more sophisticated circuits that temporarily apply protection when needed. Fuses may be used along with other protection as a last resort, and for safety when other protection might be overwhelmed. Fuses respond to overcurrent but when used together with overvoltage protection may blow if activation of the overvoltage protection results in increased current flow (which it often does). For this to occur the fuse must be ‘upstream’ of the overvoltage protection (closer to the input or power source than circuit under protection). The most obvious parameter when selecting a fuse is the current at which it will blow, but it is also necessary to consider the response time. Fuses may respond too slowly to protect circuitry in some situations (although specifically fastreacting fuses are available) – on the other hand, some systems draw high currents for short periods under normal conditions (eg, motors at start-up) and may require time-delay (or ‘slow-blow’) fuses which can tolerate higher currents for short periods. Fuses also have maximum voltage and fault current levels at which they can be safely used. For situations where the overcurrent may be much larger than the operating current and there is a risk of damage from excessive energy then highrupture capacity (HRC) fuses can be used. These have a more advanced structure than simple straight-wire-in-glass-tube basic fuses. HRCs are tightly sealed and contain fillings (such as silica sand) which help absorb energy and prevent arcing. Care with protection device use As we will mention again later, inserting any component in a circuit to help protect it will change its electrical characteristics – adding some combination of parasitic resistance, capacitance and inductance, and possibly leakage current. For fuses, the resistance and inductance are likely to be most important. In some cases, these parasitics are too small to have any impact, but in situations such as the inputs of highperformance signal processing, protection components may have a significant effect. Polymeric devices An alternative to fuses in situations where it is inconvenient, or very difficult to P T C P T C Fig.2. (Top) A selection of radial-leaded resettable PPTC polyswitches from Littelfuse; (below) PTC circuit symbols. Practical Electronics | October | 2021 as transient voltage suppressors (TVS). They all exhibit changes in conduction with voltage. At relatively low voltages they are effectively open circuits T ransient U nprotect ed P rotect ed T V S cu rrent ci rcu it ci rcu it but will have some leakage current, typically in units or tens of microamps. At higher a) b) voltages they become much more conductive, effectively Fig.3. a) Circuit without suppression and b) Circuit clamping overvoltage spikes to with transient voltage suppression. a tolerable maximum voltage, thus preventing them from causing replace them is a positive temperature damage. To do this the TVS has to coefficient (PTC) device, also called a conduct a large transient current for resettable fuse or polyswitch. Polymeric the duration of the overvoltage. Time devices (PPTCs) are fabricated from a is an important factor – a TVS must combination of a conducting particles respond quickly enough to catch the and a non-conducting polymer. Heating voltage spike before it causes damage. caused by relatively high currents expands Capacitance is typically the key the polymer, separating the conducting parasitic characteristic of TVS devices. particles and so increasing resistance very Adding capacitance to signal paths may significantly above the switching (or trip) impact frequency response, reduce input temperature – note that some leakage impedance, or affect switching speed, current will still flow. The resistance of potentially degrading performance. TVS ceramic PTCs varies with temperature devices may exhibit capacitance which as the properties of the material’s grain varies with voltage – this will introduce boundaries change. distortion in signal processing circuits. The circuit symbol for a PTC is shown Leakage currents can also be problematic in Fig.2 – this is similar to that used for in precision circuits. thermistors in general, and varistors (see The basic TVS circuit concept is later). All these devices exhibit variable illustrated in Fig.3, which could apply resistance characteristics in response to to either a power supply connection or prevailing conditions. Note that sensor signal input. An EOS voltage transient thermistor symbols may not have the short is a rapid increase in voltage to a level line at both ends of the diagonal shown which will damage a circuit. Fig.3 in Fig.2 as this is not used consistently. shows a typical transient waveform PTC resettable fuse are specified in shape featuring a rapid rise followed terms of a hold current and trip current. by a slower exponential decay. Fig.3a The hold current is the maximum current shows the unprotected circuit which will under normal operating conditions – be damaged by the transient. The circuit this varies with temperature. The trip is protected by connecting a TVS across current is the maximum current at which the source, as shown in Fig.3b. The TVS it will trip (again at a given temperature). conducts above a certain voltage, typically Between the hold and trip currents the exhibiting a clamping behaviour, which device can be in either the high or low limits the voltage during the transient to resistance states. In the design process a a safe level – as shown in the waveform PTC device is selected with a hold current in Fig.3b. To do this, the TVS needs to above the maximum normal operating conduct a large transient current and current and a trip current at or below the absorb the excess energy. minimum current at which protection The combination of transient voltage is needed. If the device trips, then after and current, and the overvoltage duration removal of the triggering condition, it will determines the amount of energy that cool down and return to its conducting a TVS device must handle – it is the state after a certain amount of time. absorption of this energy by the TVS that protects the other circuitry. Poor Overvoltage protection choice of TVS for a given situation may There are several devices used for result in its energy capabilities being overvoltage protection that are classed exceeded, resulting in damage to the F use TVS. Protection may be lost, or in the worst case, the TVS may overheat or burn, P rotect ed resulting in further damage and potential T V S ci rcu it T ransient vo ltage C lam ped vo ltage a) Fig.4. Combined TVS and fuse protection (PTC can be used in place of fuse). Practical Electronics | October | 2021 T V S b) T V S Fig.5. TVS diode symbols: a) unidirectional and b) bidirectional. fire hazards. As mentioned previously, fuses can be used in some situations to prevent damage to TVS devices where conditions may result in failure of the TVS (see Fig.4). An alternative to voltage clamping is crowbar protection, in which the source is short circuited under EOS conditions. This is mainly applicable to power inputs, rather than signal inputs. Crowbars are typically implemented using a circuit (eg, thyristor plus trigger circuit) rather than an individual device. Crowbars are commonly used to cause fuses to blow, but alternatively may activate current-limiting behaviour in the source. Some TVS devices will fail as short circuits, causing a crowbar effect. If overcurrent protection is present this may be preferable to failing as an open circuit, which removes the protection. TVS devices TVS diodes are Zener or avalanche diodes designed for use in TVS applications and have similar voltagecurrent characteristics to standard Zener diodes. TVS diodes are optimised for TVS applications rather than voltage regulation; for example, they have a larger junction area to facilitate energy absorption. TVS diodes are used in reverse bias and undergo breakdown if a sufficient reverse voltage is applied, becoming significantly more conductive, absorbing energy and clamping the voltage. Zener or avalanche diodes have similar characteristics – the names reflect the different physical mechanisms involved in reverse breakdown. TVS diodes have a similar, or the same, symbol as a Zener diode (see Fig.5a). Two TVS diodes are often used in series to provide bidirectional protection, and these are available in single packages, with the symbol shown in Fig.5b. Varistor protection A varistor, or voltage-dependent resistor (VDR) is a device which has a resistance dependent on applied voltage. The most commonly used varistors are fabricated from ceramics and are known as metaloxide varistors (MOVs). Like the ceramic PTVs mentioned earlier, the electrical behaviour of MOVs is related to the grain boundaries of the ceramic, which behave as a network of many diodes. They have the same circuit symbols as PTCs but can be distinguished by appropriate labelling (see Fig.6). Their current-voltage characteristics are similar to Zener diodes, except that the breakdown occurs at the same voltage for either polarity in a single device (see Fig.7). MOV voltage is often specified as the voltage at which they conduct 1mA – considered the lowest current at which they are performing 55 during a high voltage spike. Spark gaps are cruder versions of the same things – in which the air ionises if a sufficiently high voltage occurs. See Fig.8. Protection circuits M OV M OV Fig.6. (above) Zinc metal-oxide varistors; (below) MOV circuit symbols. clamping. Below this level, they are viewed as leakage rather than clamping currents. At higher currents a given MOV will have a higher clamping voltage. Gas discharge tubes Gas discharge tubes (GDT) contain two electrodes in a sealed device containing a gas which will ionise and conduct C urrent 1 m A V oltage C lam ping vo ltage at 1 m A Fig.7. MOV basic characteristic curve. Fig.8. Gas discharge tube (GDT) overvoltage suppressors from Eurotronix. P T C R M OV P rotect ed ci rcu it IN P U T S M OV Fig.9. Example input protection circuit. 56 P rotect ed ci rcu it Fig.9 shows an example input protection circuit, similar to the circuits that might be used in test instruments such as multimeters. The MOVs provide voltage clamping. More than one MOV is used in series – two are shown, but more Fig.10. Using series diodes to clamp the could be used. Using multiple MOVs input voltage. like this reduces the risk of arcing across individual devices and reduces the D2. The resistor RL limits the current effective capacitance. During a voltage that can flow in the protection diodes, transient the MOVs will switch very preventing them from being damaged. quickly, but the PTC will take longer Standard diodes can be used in this type to respond and before it does a large of circuit, but Schottky diodes (as shown transient current will flow. This is in Fig.11) have a lower forward voltage. limited by the resistor, which needs to This clamps the inputs to a lower total be of sufficiently high rating to handle voltage and prevents the op amp’s internal the energy it will absorb during an ESD protection diodes from conducting, overvoltage transient. avoiding excessive current in these internal MOVs, TVSs and GDTs are typically diodes (important because the ESD diodes used for protection at relatively high are not designed to handle this situation). voltage levels (tens or hundreds of volts, With high EOS voltages the circuit in or more). The switch-on of forwardFig.11 will cause an increase in supply biased standard diodes can also be rail voltage. This could damage the op used for overvoltage protection where amp or the regulator providing the power protection is required at, or is appropriate supply. To prevent this, the supply can to occur at, much lower voltages. The be protected using TVS diodes, as shown clamping voltage can be set to multiples in Fig.12. Leakage currents in input of diode forward voltage (VF = 0.6 to diodes in the circuits in Fig.11 and Fig.12 can degrade the performance of some 0.7V for standard silicon diodes) by high-precision circuits (during normal using diodes in series. An example is operation). To overcome this, some shown in Fig.10, which uses sets of five precision op amps have internal clamping diodes in series to prevent the input circuits which provide protection with exceedingly around 3.5V (5 × 0.7V) in less impact on performance. either polarity. The circuit may also include a fuse or current limiting resistor. The same effect can be + V S achieved with fewer diodes using a bridge rectifier across the inputs, D 1 Output with additional diodes connected – I nput across the bridge output. + For signal inputs to circuits using, R L for example, op amps, ADCs or D 2 microcontrollers, single diodes can be used to clamp input voltages to – V S the supply rails (or more accurately one diode drop beyond the rail voltage). An example of such a Fig.11. Example circuit: overvoltage protection circuit – an op amp unity-gain buffer using diode clamping. with overvoltage protection is shown in Fig.11. Diode D1 + V S conducts if the input goes higher than the positive D 1 T V S 1 supply voltage (VP) by more Output – I nput than the turn-on voltage + (forward drop, VF). As the R L input voltage increases above VS, VF for D1 remains almost T V S 2 D 2 constant, clamping the voltage at the op amp input to VP + VF. – V S Similarly, for negative input voltages, the op amp input Fig.12. Adding TVS diodes to the circuit in Fig.10 is clamped at −V S − V F by provides additional protection. Practical Electronics | October | 2021