Home > 9 Semiconductor junction surge-protective devices This clause presents a description, theory of operation, test characteristics, and an applica

9 Semiconductor junction surge-protective devices This clause presents a description, theory of operation, test characteristics, and an applica

  1. Semiconductor junction surge-protective devices

This clause presents a description, theory of operation, test characteristics, and an application guide for semiconductor junction surge-protection devices. These products are often referred to as Transient Voltage Suppressors (TVS). Most junction surge-protection devices are diodes consisting of a single P-N junction. Diodes with a well defined breakdown voltage are often called Zener diodes, although as will be discussed below, this is often not technically correct, but the use of this terminology is widespread. Another term often used for TVS diodes is an avalanche diode and in most cases this is a more correct term than Zener diode. There are also TVS devices which use a pair of junctions, either P-N-P or N-P-N. These TVS are special cases and will be discussed in Clause 1.5.3. The most basic TVS products are two terminal devices but multi terminal products are also available for specific applications. Single diode TVS devices are intrinsically asymmetric (unidirectional) but products with two back to back TVS diodes, providing symmetric (bidirectional) operation are also available. Applications to ac or dc power and analog or digital communication transmission lines are referenced in this guide. Three junction, PNPN, protection devices are called Thyristors or Silicon Controlled Rectifiers (SCR) but will not be covered in this document since they are discussed in IEEE Std. C62.37-1996 and IEEE Std. C62.37.1-2000.

TVS devices are designed to provide protection to voltage-sensitive components or circuits. With the guide, the reader will be able to compare the standard electrical characteristics of the TVS devices and be able to select a protector to meet most applications.

Although many TVS devices are two terminal protectors, some applications may require these devices to be configured in series or parallel combination to increase the voltage or power handling capability. In other applications, the TVS may require some additional components to limit or divert a specified level of transient threat or to lower the capacitance.

    1. Diode Description

The diode, in its basic form, is a single P-N junction consisting of an anode (P) and a cathode (N). See Figure 1a. In DC circuit applications, the protector is reverse biased such that a positive potential is applied to the cathode (N) side of the device. See Figure 1b. 

Figure 1 Basic form of a diode

    1. Theory of Operation

Diodes have there operating regions, 1) forward bias (low impedance), 2) off state (high impedance), and reverse bias breakdown (relatively low impedance). These regions can be seen in Figure 2. In the forward bias mode with a positive voltage on the P region, the diode has very low impedance once the voltage exceeds the forward bias diode voltage, VFS. VFS is usually less than 1 V and will be defined below. The off state extends from 0 V to just below a positive VBR on the N region. In this region the only currents that flow are temperature dependent leakage currents and Zener tunneling currents for low breakdown voltage diodes. Zener tunneling will be discussed in Clause 1.2.1. The reverse bias breakdown region begins with a positive VBR on the N region. At VBR electrons crossing the junction are accelerated enough by the high field in the junction region that electron collisions result in a cascade, or avalanche, of electrons and holes being created. The result is a sharp drop in the resistance of the diode. Both the forward bias and reverse bias breakdown regions can be used for protection. 

Figure 2 Avalanche diode component I-V characteristics

The electrical characteristics of an avalanche diode are intrinsically asymmetric. Symmetric avalanche diode protection products consisting of back to back junctions are also manufactured. This will be discussed in Clause

There are some additional parameters, which are not shown in Figure 2, such as capacitance, and characteristics that vary with temperature, which may affect the application of the avalanche diode. These can be found in manufacturers’ data sheets and will be discussed here when they apply to specific applications for products selection. The avalanche diode performance characteristics are defined in IEEE Std C62.35-1987, which will be used as a source document for this guide.

      1. Zener Diodes

Diodes used for protection are often called Zener diodes. Zener conduction is direct band to band tunneling between the valance and conduction bands of the semiconductor. It occurs in low breakdown voltage diodes because the high doping levels needed for low breakdown voltage creates a very narrow transition region around the junction. Zener conduction does not have as sharp a turn on as avalanche breakdown. In fact Zener conduction creates high leakage levels in low breakdown diodes. This is shown in Figure 3 for diodes with 5.1, 7.1 and 9.1 V breakdowns. The Zener and Avalanche regions are shown. It is clear that the Zener regions conducts more current as the breakdown voltage decreases, but that the high current conduction is in fact avalanche conduction. 

Figure 3 I-V curves for diodes with breakdown voltages of 5.1, 7.1 and 9.1 V showing Avalanche and Zener breakdown voltages

    1. Diode test characteristics

Basic parametric considerations of TVS diodes (see 1.3.1, 1.3.2, 1.3.3 and 1.3.4) are detailed explanations of the ratings and characteristics that are the minimum necessary to use the devices in most applications. 1.3.5 through 1.3.19 of this guide define additional parameters, which might be useful in other applications. Current impulses for these tests use a 10/1000 waveform.

The 10/1000 waveform is being used in a different manor than it is used in other documents such as C62.41-1991. In other documents the 10/1000 waveform is used to determine the withstand capability of a product in a given application. The product may be a sub-assembly or a completed system. In this portion of this guide the 10/1000 waveform is being used to characterize the properties of an avalanche diode. In most cases the 10/1000 waveform will be used to determine the diodes properties below the destruct level.

In an application test, it is expected that the protection device will modify the applied waveform. Application waveforms are supplied by circuits that have fixed source impedance. As a result, application waveforms may only be specified for specific load impedance (for example, an open circuit, or a short circuit). Typically, a 10/1000 application waveform is a short circuit current (SCI) waveform. When such a waveform is applied to an actual device, its amplitude and waveshape will differ from that of a SCI waveform. An example of this is given in Clause 1.3.1. Clause 1 of the guide does not discuss the application waveforms.

      1. Clamping Voltage (Vc)

Clamping voltage is the measured, peak (crest) voltage across the device assuming that the external lead terminations do not adversely add to this voltage. For this definition, a waveform of 10/1000 is used as the standard test current impulse to minimize the effects of the external lead lengths. The user is cautioned, however, that an impulse with a fast rising front will add overshoot voltage (VOS), due to lead inductance, to the clamping voltage. Overshoot voltage is discussed in 1.3.14.

Clamping voltage will vary as a function of the magnitude of the applied peak impulse current for a specified waveform. This voltage is a result of the device’s breakdown voltage, VBR, device series resistance and thermal effects. Figure 4 shows the voltage across and current through a transient suppressor diode, showing how the peak voltage and peak current do not necessarily coincide. The device is a 400 W Peak Power TVS with a nominal breakdown voltage of 82 V stressed with a 120 V, 10/1000 waveform. The data clearly shows that the current peak, at about 10 us, occurs well before the voltage peak. This data also shows how the device under test significantly changes the waveform properties as discussed in Clause 1.3. The voltage peak is clamped at about 96 V. The clamping of the voltage makes the decay time look much longer than expected for a 10/1000 waveform. The current through the TVS drops much faster than expected for a 10/1000 waveform. Early in the pulse device heating increases the resistance, causing a rapid decrease in current. Later, as the voltage across the TVS drops, the nonlinear I-V properties of the TVS result in a very rapid drop in current. When the voltage reaches the TVS’s breakdown voltage current through the TVS stops. Once the current through the diode is zero the voltage across the TVS will drop at a rate determined by the properties of the pulse source. 

Figure 4 Impulse voltage and current waveforms (10/1000us)

Users are cautioned to review the individual manufacturer’s data sheet to confirm the level of peak pulse current applied to the device for each impulse waveform. For example, the minimum clamping voltage is at the point of junction avalanche, which is measured at peak current level of l mA to 10 mA for less than 10 ms. Although this voltage is often expressed as a clamping voltage by some manufacturers, it is defined here as the breakdown voltage (see 1.3.7).

Temperature will also cause the clamping voltage to vary due to the temperature coefficient of the device. Most devices have a positive temperature coefficient that will cause the voltage to increase in temperature as defined by a manufacturer’s data sheet under temperature coefficient of breakdown voltage expressed as mV/�C or %/�C (see 9.3.19).

Typical values of clamping voltage range from 7 V to 540 V.

      1. Rated peak impulse current (IPPM)

This is also referred to as rated multiple peak impulse current, reference 9.3.5 of IEEE Std C62.35-1987. Rated peak impulse current is the maximum value of a peak impulse current that is applied to a device for a minimum of 10 pulses using a 10/1000 waveform and maximum duty cycle of 0.01% without causing device failure. The clamping voltage measured during this test is considered the “maximum clamping voltage” for a given current and pulse waveform for any device.

Due to the number of waveform combinations and the possible peak current ratings for a wide voltage range of devices, most manufacturers will include a peak pulse power vs. impulse time curve, similar to Figure 5, on their data sheets to define the operating characteristics of a device over time. Using this curve and the maximum clamping voltage indicated in the previous paragraph, it is possible to define the peak impulse current of a given device for any impulse duration. In Figure 5, the Pulse Time (td) is considered the time duration of an impulse waveform. For example for a 10/1000 waveform the 1000 is the impulse duration or td. Divide the peak pulse power by the “maximum” clamping voltage to determine the peak impulse current for a specific impulse duration. 

Figure 5 Peak pulse power vs. pulse time

It is this peak impulse current that is used to select a device for a specified transient waveform. The transient environment is usually defined in terms of its peak transient current over a specified impulse, duration. It is this transient current that the device is designed to divert away from the load. The user is advised to understand the manufacturer’s current ratings on a data sheet and apply them to the application. In some applications, it may be necessary to insert a series impedance in the line ahead of the protector to assure its performance to the designed specifications.

The peak pulse current specified by the manufacturer is considered repetitive with a specified duty cycle. Refer to the individual manufacturer’s data sheet for this specific information, which relates to the average power handling capability of the device. For duty cycles greater than specified amount, contact the manufacture for details on the alternatives. For impulse durations greater than 1.0 ms, or a duty cycle exceeding the manufacturer’s rating, an alternate device might be required.

Although IPPM is determined by conducting multiple pulses, the purpose of conducting multiple pulses is to establish that the device can reliably handle a single pulse and is not intended to be a measure of the lifetime rated pulse current (see 1.3.6).

      1. Rated stand-off voltage (VWM) / Rated working RMS voltage (VWM(RMS))

Rated stand-off or working RMS voltage is that voltage whose peak value is the manufacturer’s recommended maximum continuous operating voltage of the avalanche diode that is determined by multiplying the minimum breakdown voltage by .9 or .95.

Datasheets typically specify these values at room temperature. Consideration should be given to variation of these values with temperature. Unfortunately most datasheets do not include temperature coefficients. Selection of the working voltage should therefore be done conservatively or tests done at temperature to insure proper operation over the full temperature range. The TVS manufacture may also be contacted to provide temperature coefficient information.

These voltage ratings are to be considered the maximum circuit, system, or equipment operating line voltage when operating over the device’s full operating temperature range. Exceeding this value will cause clipping of the signal or an increase in the steady state power operating condition of the device.

The blue is original wording from document. This does not reflect the reality of most datasheets which specify only 25C values and no temperature coefficients.

        1. Rated stand-off voltage (VWM)

For dc applications, the rated stand-off voltage applies to the maximum working dc or peak voltage application. For avalanche diodes, this rating is related to the minimum avalanche breakdown of the P-N junction. This voltage rating is typically 10% below the breakdown voltage. Refer to the individual manufacturer’s data sheet for actual values. For operation away from room temperature care must be used. All silicon avalanche junction diodes have a positive temperature coefficient of breakdown voltage. At low temperatures the breakdown voltage will get closer to VWM and it may be necessary to compensate for this or choose a higher breakdown voltage diode. At high temperatures there will be more margin between VWM and the breakdown voltage. For a temperature value other than 25 �C, the temperature coefficient parameter can be used to calculate a specified stand-off voltage below the breakdown voltage at that temperature. For example, if the low temperature operation is –10 �C, the difference from room temperature (25 �C) is 35 �C. The equations below can be used to calculate voltage differences at temperature other than the datasheet temperature.



    VdiffT0 = voltage difference between the Stand-off voltage and the Breakdown voltage at reference temperature

    VdiffT = voltage difference between the Stand-off voltage and the Breakdown voltage at temperature of interest

    α = temperature coefficient

    T = temperature difference between temperature of interest and the reference temperature at which the diode properties are specified

Note: The equations in the original document while very simple were to my surprise incorrect.

The following paragraph in blue contains much of the original wording. As I looked into this more this seemed to contradict common practice in which all parameters are specified at room temperature. Some datasheets include derating for power capability at other than room temperature values of breakdown voltage and working voltage are only specified at room temperature without temperature coefficients. For that reason I used the above wording.

For dc applications, the rated stand-off voltage applies to the maximum working dc or peak voltage application. For avalanche diodes, this rating is determined by the minimum avalanche breakdown of the P-N junction. This voltage rating is typically 10% below the minimum breakdown voltage due to the temperature coefficient of the breakdown voltage. Refer to the individual manufacturer’s data sheet for actual values. Differences between stand-off and breakdown will also be determined by the maximum operating temperature range of the system. For operating temperature ranges less than the maximum, the voltage difference can be reduced. All silicon avalanche junction diodes have a positive temperature coefficient of breakdown voltage. The stand-off voltage is based upon the minimum temperature value, which requires that the voltage to be specified at 10% below the minimum breakdown voltage. For a temperature value between minimum and 25 �C, the temperature coefficient parameter can be used to calculate a specified stand-off voltage below the minimum breakdown voltage. For example, if the low temperature operation is –10 �C, the difference from room temperature (25 �C) is 35 �C. The equations below can be used to calculate voltage differences at temperature other than the datasheet temperature.



    VdiffT0 = voltage difference between the Stand-off voltage and the Breakdown voltage at reference temperature

    VdiffT = voltage difference between the Stand-off voltage and the Breakdown voltage at temperature of interest

    α = temperature coefficient

    T = temperature difference between temperature of interest and the reference temperature at which the diode properties are specified

Note: The equations in the original document while very simple were to my surprise incorrect.

        1. Rated working RMS voltage (VWM(RMS))

For ac applications, the peak (crest) voltage value must be used when determining the stand-off voltage. The rated working RMS voltage VWM(RMS) applies to bidirectional devices with symmetrical voltage characteristics around zero volts. Since most electrical characteristics are specified in dc or peak values, all RMS operating circuit values must be converted to their peak values. For example, a sinusoidal RMS voltage of 110 V (ac) must be multiplied by 1.414 to obtain the peak voltage.

Voltage ratings are subject to the same cautions and instructions for temperature as the stand-off voltage considerations stated in Users are to review the manufacturer’s data sheet in reference to this parameter.

      1. Stand-by current (ID)

Stand-by current is the dc or steady state reverse leakage current through the diode under a reverse bias voltage condition. This parameter is always measured at room ambient temperature 25 �C. In general, the reverse leakage current will double with every 10 �C to 15 �C rise injunction temperature. The stand-by current is the level of current that is diverted by the presence of the diode being placed across the line. Its specific value is not always an indication of the device reliability, but must be considered for circuit operation in some applications. Stand-by current is measured at the rated stand-off voltage for dc or steady state applications and at the RMS voltage for ac applications.

Typical values of stand-by current range from 0.01 μA to 1000 μA.

      1. Rated peak single surge transient current (ISM)

This is the maximum value of peak impulse current, 10/1000 waveform, which a device must withstand without causing failure. This rating is a means of establishing the fail short condition of a specific voltage type before failure that might be expected in an application environment.

A specific device type may withstand a rated peak single impulse transient current. However, due to the variations in the cross sectional area of an active junction, not all manufacturers products are alike. This value might vary from one manufacturer to another. Each manufacturer must identify this parameter for each device type and voltage (VWM) rating. A TVS diode might withstand several exposures at this surge value, but it should not be considered a peak rating nor a design limit for an application. For lack of an industry standard, it is suggested that this value be derated by 5% in any application.

      1. Lifetime rated pulse current

This is a multiple impulse current rating for a defined waveform that may be applied over the lifetime of the device without causing failure. The ratings are based upon both an 8/20 and 10/1000 waveform, refer to Table I in IEEE Std C62.35-1987. These ratings are not usually a part of a manufacturer’s specification for TVS diodes. Generally, data of this kind must be performed on an individual product bases. This is due to the different methods of device construction, between the various product families. Individual manufacturer’s data is a meaningful measure of the life expectancy of a TVS diode under actual in-service conditions.

      1.   Breakdown voltage (avalanche) V(BR)

The breakdown voltage is a characteristic that is measured at a specified current, usually l mA, which may be stated as a minimum or nominal value. This voltage measurement is characteristic of the avalanche point for a P-N junction device.

This characteristic is usually tested with automatic test equipment using a pulse width of less than 10 ms. For dc current measurements, the user must take into consideration the temperature coefficient of the device and the increase in voltage due to self heating. This same consideration must be given to the room ambient temperature conditions.

Typical values of breakdown voltage range from 3 V for single diodes to 850 V or higher for multi-stacked diodes (diodes in series)

      1. Rated multiple peak pulse power dissipation (PPPM)

The multiple peak pulse power is a rating that is derived from multiplying the maximum clamping voltage (VC) times the rated peak impulse current (IPPM). Due to self heating of a P/N junction device for long pulses (greater than 300 μs), the voltage is not coincident in time with the peak pulse current. For lack of an industry standard, an impulse of 10/1000 is used to classify most manufacturers’ peak pulse power ratings, such as 500 W, 1500 W, and 5000 W. This parameter is commonly used by a manufacturer as the product identifier for a series or family of device components. Contact the individual manufacturer for the derating factor for this parameter within a device family.

      1. Clamping factor (CF)

The clamping factor of a specific device is the ratio of its clamping voltage (VC) to its breakdown voltage (VBR), where the clamping voltage is measured at a pulse current of specified peak value and waveshape, and the breakdown voltage is measured at the specified value of direct current.

A lower clamping factor indicates a lower clamping voltage, for a given device at specified current. Devices with higher power ratings, but of the same breakdown voltage value, usually have lower clamping factors than devices with lower power ratings.

      1. Voltage clamping ratio (VC/VWM)

This ratio is a value obtained by dividing clamping voltage by the maximum rated standoff voltage. The value will depend upon the test current at which clamping voltage is measured for a given device type. Both voltage measurements must use the same units of measurement, such as peak volts.

In making comparisons between products the user should take care to see that similar definitions and test conditions are used. The surest course is to define the stand-off voltage relative to the minimum breakdown voltage and the clamping voltage conditions of the devices under question. Comparing the data in a manufacturer’s data sheet is important but may not be adequate for all products.

If this parameter is important for a specific application it is best to contact the individual manufacturer of interest. They can provide the data or suggest a product that can meet the application needs. As with other parameters the voltage clamping ratio is dependent upon the manufacturing process and procedures as well as other device characteristics, such as junction area.

      1. Incremental surge resistance (RS)

An avalanche junction semiconductor, in its on-state, presents a resistance to the flow of surge current. When the applied surge pulse is of short duration, causing very little heating in the device, the value of RS is constant. Pulses of longer duration, which generate heating, will result in an increase of RS.

The incremental surge resistance of an avalanche junction semiconductor is directly related to its clamping voltage (VC). Clamping voltage is a much more useful parameter than RS in determining the protection characteristics of a device. Knowledge of the Incremental surge resistance may be useful when a surge generator is designed to deliver a precise peak current of a precise waveshape through a specific device.

      1. Capacitance (C)

All avalanche junction semiconductor surge-protective devices have a value of capacitance that is directly proportional to their area and approximately inversely proportional to their breakdown voltage as shown in Figure 6. Capacitance will decrease with an increase in an applied dc bias voltage across the reverse direction of a P-N junction, as in Figure 6. Capacitance will also decrease with increasing frequency from 5% to 50% depending upon the breakdown voltage. 

Figure 6 Typical capacitance vs. breakdown voltage (unidirectional only)

Due to the large cross sectional area of the avalanche junction device, the capacitance can range from several picofarads up to several thousand picofarads depending upon the device breakdown voltage and power rating. Whereas these values of capacitance may not be significant for dc applications, for high-frequency (fast data rates) data lines the capacitance may have to be compensated. A high-voltage rectifier diode may be connected in series combination with the avalanche diode of opposite polarity for capacitance compensation. The capacitance of a bidirectional device to a unidirectional device of the same breakdown voltage will be 30% to 50% less.

      1. Voltage overshoot (VOS)

Overshoot is the voltage that occurs above the P-N junction clamping voltage when subject to a fast rise time impulse. Lead inductance and the metallic interface will cause an incremental voltage increase due to the L di/dt effect. The value of overshoot will depend on the rate of change of the impulse front wave. In typical applications, lead length refers to the length of leads inherent to the device (i.e., device package), and any additional leads provided as part of the application circuitry. This can cause voltage overshoot when testing products in a circuit. From a practical standpoint, users should try to minimize the length of wires connecting surge-protective devices to the power bus lines and the data line I/O ports. It is also important to consider the board layout of the printed circuit board when designing surge-protective devices in their circuits.

      1. Response time, overshoot duration

As defined in IEEE Std C62.35-1987, response time is the time between the point at which VC is exceeded and the peak overshoot voltage is reached. Due to the high frequencies involved in fast rise time waveforms, response measurements require special fixtures and extremely fast-responding instrumentation. Reference to response time in device specification is discouraged. It is not characteristic of its clamping effectiveness and is dependent upon defined test methods and circuit configuration. Overshoot duration is the time between the same point when VC is exceeded and the 50% point on the overshoot wave tail.

      1. Rated forward surge current (IFSM)

The forward surge current applies to unidirectional TVS diodes. It is the maximum forward bias single peak current for an 8.3 ms, half sine wave, without causing device failure, see Figure 2. Although IFSM is defined with an 8.3 ms half sign wave, due to the availability of standard test equipment, this characteristic is also being tested at the same waveform as for the reverse polarity such as 10/1000 and 8/20. The voltage measurement will always be less than the clamping voltage of the protection device which is the voltage measurement of a single forward diode, P-N junction. {I don’t follow this last sentence which is from the original document.}

      1. Forward voltage (VFS)

Forward voltage is the peak voltage measured across a forward biased unidirectional avalanche diode at a specified pulse current and waveform. See Figure 2. This is usually provided for by the manufacturer to help the user determine the voltage that will be experienced on the back side of the circuit element being protected. It is determined by the manufacturer through a series of tests on a population of devices and is usually expressed as a maximum value.

      1. Temperature derating

Due to the dependency of some electrical characteristics to temperature, temperature derating is applied to both the peak pulse power and peak pulse current of a device, see Figure 7. Both power and current are usually rated from 100% at 25 �C to 0% at some elevated temperature defined by the individual manufacturer. Temperatures will range from l00 �C to 175 �C depending upon the method of manufacturing the product and the materials used in the device construction. The derating is normally given as a percentage of the peak pulse power or peak pulse current above the 25 �C temperature. 

Figure 7 Derating curve

      1. Thermal resistance (R8JA, R8JC, R8JL)

Thermal resistance is the effective temperature rise per unit power dissipation of a designed junction, above the temperature of a stated external reference point under conditions of thermal equilibrium. The thermal resistance is an indication of the device’s ability to dissipate heat away from the junction of the semiconductor chip. Most manufacturers will show this parameter as a maximum rating under specific conditions such as may be defined at a lead, surface, or room ambient temperature. When the lead or surface temperature is given, it is defined at a distance from the package for plastic devices or the case for throughhole, metal packages.

      1. Temperature coefficient of breakdown voltage (V(BR))/( V(BR))

All silicon avalanche junction surge suppressors will have a temperature coefficient of breakdown voltage that is positive. TVS devices which are not avalanche diodes may have negative temperature coefficients. This can be expressed as a percentage of the breakdown voltage or defined as a millivolt change with temperature (see IEEE Std C62.35-1987). The clamping voltage will also have the same voltage change with temperature independent of the value of peak pulse test current. The pulse duration of the peak current does not affect the temperature coefficient. There may be an increase in the voltage due to variations in the pulse duration. This will affect the junction heating of the diode, if not properly mounted to a heat dissipating surface, causing a slight increase in voltage.

    1. Failure modes
      1. Degradation failure mode

When damaged a TVS diode will experience an increase in the stand-by current. This is usually a result of degradation in the junction surface area of the device. It can be caused by an imperfection in the junction coating or cracking of the coating material. This can be observed with a curve tracer, in which case the knee (the transition between off condition and the full on condition) will appear to be unstable.

      1. Short circuit failure modes

An avalanche diode will have two types of short circuit failure modes. The first is a shorted junction in which the surface or the bulk material will have failed. Evidence of this type of failure mode is a remelting of the silicon material. This is usually due to the device being subjected to a peak transient current value in excess of the device rating. The second failure mode is melting of internal materials that arc across the semiconductor junction as a result of a long duration transient. Both failure modes will cause the device to appear as a low value resistor, usually less than l Ω. In some applications an over current device may be required to prevent continuous current flow if this failure occurs.

      1. Open circuit failure mode

A device will be considered open circuit when the breakdown voltage exceeds l50% of pre-tested value at a current used to obtain V(BR) or at a lower current. This condition can exist when the unit has been exposed to a very high level of transient current of a short duration, usually less than l μs. Another condition may exist when the internal leads have fused open. This condition is a result of the device first shorting and then drawing enough continuous current to melt open the lead of the device.

      1. High clamping voltage failure mode

A device is considered a failure when the voltage exceeds 120% of the pre-tested clamping voltage. An avalanche junction surge suppressor that has been improperly processed through production may experience this phenomenon.

    1. Specialized Avalanche Diode Products

The two terminal avalanche diode is a very useful product but a wide variety of specialized products have been developed for specific applications.

      1. Unidirectional and Bidirectional Products

The basic avalanche diode has an asymmetric I-V curve and is well suited to protecting circuit nodes that are unidirectional, having voltage that is always positive or always negative. This is illustrated in Figure 8 a and b. To protect a node with positive only voltages the avalanche diode’s cathode is connected to the positive only node and the anode is typically connected to ground. To protect a node with a negative only signal the anode is connected to the negative signal node and the cathode is connected typically to ground. 

Figure 8 Unidirectional and Bidirectional avalanche diode product usage

To protect bidirectional signal nodes which have both positive and negative voltages on them requires protection with symmetric properties as shown in Figure 8c. This can be done in a variety of ways as shown in Figure 9. Back to back diodes can be used as shown in Figure 9a, with either the cathodes tied together, as in the top figure, or with the anodes tied together, as in the bottom figure. Many manufacturers provide bidirectional products. This can be done by assembling two die in one package as illustrated in Figure 9b. It is very common to use a one silicon die solution as shown in Figure 9c where the two junctions share a common cathode, top picture, or common anode, bottom picture. Such an arrangement looks like a bipolar transistor and some bidirectional products may show some snapback behavior characteristic of a bipolar transistor. The central shared region is commonly made very long, minimizing the bipolar transistor properties. 

Figure 9 Providing bidirectional or symmetric protection using avalanche diodes

Bidirectional avalanche diode products are characterized similarly to unidirectional devices as illustrated in Figure 2. Only the reverse bias characteristics are meaningful for a bidirectional product. Note that the properties of a bidirectional product may not be perfectly symmetrical.

      1. Avalanche Diode Arrays

Avalanche diodes are typically made using integrated circuit technology, making it possible to manufacture arrays of diodes to meet specific circuit applications. Several common configurations are shown in Figure 10. 

Figure 10 Avalanche diode arrays

Figure 10a shows a dual common anode product and Figure 10b shows a dual common cathode product. Both of these products can be used as two unidirectional protectors or as single bidirectional protectors with the common connection left floating during use. Figure 10c shows a 4 line common anode protection product.

Figure 10d is an integrated solution that can provide protection to two data lines as well as provide protection between power and ground. It consists of four high breakdown voltage diodes and a single avalanche diode with a breakdown voltage appropriate for the specific application. The 4 high breakdown voltage diodes are commonly called steering diodes. They are only intended to carry current in forward bias mode. The low breakdown voltage diode between power and ground provides protection for power with respect to ground. Protection for the two I/O lines is provided by forward bias of the steering diodes directly to power or ground or with a combination of forward bias of one of the steering diodes and reverse breakdown of the low voltage avalanche breakdown diode.

      1. Foldback and Punch Through Diodes

Foldback and punch through devices have similar two junction constructions but their modes of operation are different and their regions of application are different.

        1. Foldback Diodes

I am still gathering information for this section.

        1. Punch Through Diodes

Punch through diodes are primarily intended for low voltage applications. Standard diodes with low breakdown voltage have high leakage and high capacitance. Punch through diodes are designed as a substitute for low breakdown voltage diodes. Punch through diodes have an npn structure. The doping of the base region is low, such that punch through occurs before avalanche breakdown can occur. The edges of the junctions are tailored to improve their high current behavior.

Publication on punch through diodes: Journal of Electrostatics, Volume 61, Issues 3-4, July 2004, Pages 149-169

      1. Avalanche Diodes for ESD Protection

Electrostatic discharge (ESD) is a special class of stress characterized by relatively low total energy but very fast rise time. ESD occurs when people or objects become charged to high voltage, several to tens of thousands of volts, and discharge to a surface at a different potential. Current levels can be up to 30 or 40 A, which is low compared to other stress waveforms. The very fast rise times, from less then a ns to a few ns, make some protection products such as gas discharge tubes and Thyristors ineffective due to slow turn on speed. Avalanche diodes have very fast turn on speeds, making them effective for ESD protection.

Circuit nodes requiring ESD protection often do not need the high energy protection needed for telephone circuits where lightning and power cross can occur. High speed data lines such as USB 2.0 (Universal Serial Bus) and HDMI (High-Definition Multimedia Interface) however, require low capacitance protection. Avalanche diodes intended for ESD protection therefore do not have the energy absorption capability to be adequately characterized with the traditional 10/1000 surge currents discussed in Clause xx. Special tests are needed for ESD protection products. Unfortunately there are currently inadequate standards for ESD protection products. There are some common practices in the industry and those will be discussed. Effort is underway by the Electrostatic Discharge Association (ESDA) to address some of these issues.

The parameters and test methods used to characterize avalanche diodes for normal, non-protection, operation discussed in Clause 1.3 work perfectly for avalanche diodes intended for ESD protection. It is only the diode properties in the protection region that need special characterization for ESD. Characterization of avalanche diodes for ESD protection falls into two categories, the ability of the protection element to survive the ESD threat, and the effectiveness of the protection element at providing protection.

        1. Survival of ESD Avalanche Diodes

Electrical systems are typically ESD tested to the IEC 61000-4-2 standard. IEC 61000-4-2 defines a current stress as shown in Figure 11 with parameters as specified in Table 1. Manufacturers typically define the ability of their ESD protection products to survive a specific value of the IEC 61000-4-2 current waveform. 8000 V is a typical value for protection products to pass, since that is the highest voltage specified in the IEC 61000-4-2 standard. Many products are specified at much higher voltage levels such as 16,000 V.  

Figure 11 IEC 61000-4-2 current waveform for contact discharge

Parameter Value Tolerance
Rise Time (10 to 90% of Ip) 0.8 ns �25 %
Peak Current 3.75 A/kV �10 %
Current at 30 ns 2.0 A/kV �30 %
Current at 60 ns 1.0 A/kV �30 %

Table 1 Parameters for IEC 61000-4-2 for contact discharge

Specifying that protection products pass the IEC 61000-4-2 standard at a specified voltage is not strictly speaking a valid procedure. IEC 61000-4-2 is a system level standard and does not define how to test components. It is therefore not clear if all manufacturers are testing their products in the same way. The ESDA is currently working on a test procedure for testing electrical components using the IEC 61000-4-2 waveform. The test method is being called Human Metal Model (HMM). This is because the IEC 61000-4-2 waveform is supposed to represent a person holding a metal object such as a key touching a grounded surface. The initial current spike is the discharge of the low resistance key and the remainder of the pulse is discharge of the higher resistance person.

        1. Effectiveness of ESD TVS Diodes

The effectiveness of any protection element depends on the resistance of the protection element in its on state. There is no standard method for measuring this and a variety of procedures have been used. A popular method is the voltage screen shot during an IEC 62000-4-2 current stress. Screen shots are shown on many data sheets for ESD protection products. A schematic of a typical test setup is shown in Figure 12and a sample measurement is shown in Figure 13. The screen shot gives a good “at a glance” impression of the clamping ability of the protection product. There are no agreed upon parameters that can be extracted from the measurement. The measurement is therefore most useful when comparing different protection products using the same measurement technique.

In doing voltage screen shot measurements it is important to use the best high quality RF measurement techniques in connecting the device under test to the oscilloscope. It is also usually necessary to use a 50 ohm attenuator at the input of the oscilloscope to protect the input circuitry. When viewing voltage screen shots it is important to remember that even with the best of measurement technique the first several ns of the measurement are of questionable value. The initial spike and oscillations is a combination of the initial current spike from the ESD gun, induced voltage due to the inductance in the test fixture coupled with the fast current rise, and reflections due to impedance mismatch between the ESD gun and the 50 ohm measurement system. 

Figure 12 Test setup for screenshot during IEC stress 

Figure 13 Sample IEC 61000-4-2 screen shot

Transmission line pulse (TLP) measurements have also been used to characterize ESD protection products and appear on some datasheets. TLP is a technique for measuring the current versus voltage characteristics of circuit elements at the time scales and current levels characteristic of ESD events using square pulses. Figure 14 is an example of a TLP measurement of the reverse bias characteristics of an avalanche diode intended for ESD protection. In this example each of the data points was measured with a 120 ns long pulse. The technique allows for the extraction of breakdown voltages and dynamic resistance. Commercial TLP systems with 100 ns long pulses typically have maximum current limits of 10 or 20 A. This is often not enough to measure the damage threshold of ESD protection products, but the measurement of the on state properties below the damage point is extremely useful for comparing protection products and predicting their effectiveness. 

Figure 14 TLP measurement of an avalanche diode intended for ESD protection

        1. Ultra Low Capacitance Avalanche Diodes for ESD Protection

Modern high speed data interfaces such as USB 2.0 and HDMI require ESD protection with very low capacitance, typically under 1 pF. High speed interfaces are also running at lower and lower voltages as more advanced technologies are used. This presents a problem for avalanche diodes intended for ESD protection. As the voltage levels decrease protection elements must turn on at lower voltages. Unfortunately the capacitance per unit junction area of avalanche diodes increases as the breakdown voltage is decreased. Providing low capacitance requires a reduction in junction area which compromises the ability of the diode to provide protection. Fortunately a combination of circuit and technology techniques has been used to provide sub pF protection without giving up adequate ESD protection.

Figure 15a shows the effective circuit diagram for a unidirectional ultra low capacitance avalanche protection diode. Figure 15b shows the actual schematic. The device has one low breakdown, and therefore high capacitance diode, and two high breakdown voltage, low capacitance, diodes. Two capacitors in series always have less than the capacitance of the smaller of the two capacitors. Therefore the high capacitance of the low breakdown voltage diode is masked by the high breakdown diode in series with it. The I-V characteristics of the circuit in Figure 15a is indistinguishable from the I-V curve for Figure 15b. In forward bias the single high breakdown voltage diode on the right of the circuit conducts in the forward bias mode. In the reverse breakdown mode the high breakdown voltage diode on the left of the circuit is forward biased and only adds a forward bias diode drop to the breakdown voltage of the low breakdown voltage diode. 

Figure 15 Sample of ultra low capacitance avalanche diodes are created

    1. Application of Junction Surge Protection Devices

Junction surge protection devices can be used in a number of different types of applications, ranging from ac and dc power circuits to both analog and digital signal lines. The key for applying these devices is the location in the circuit being protected, the severity of the transient stress and the duration of the stress. These factors are the determining parameters that must be used in the design of protection circuits. It is unrealistic to equate the energy level of the transient to the energy handling capability of the avalanche diode. These two values are not equatable due to the low clamping voltage of the device and the fact that some of the energy will be dissipated in the wire or cable connecting the surge suppressor to the point injection of the transient threat. 

Annex D 

Set Home | Add to Favorites

All Rights Reserved Powered by Free Document Search and Download

Copyright © 2011
This site does not host pdf,doc,ppt,xls,rtf,txt files all document are the property of their respective owners. complaint#downhi.com