Characteristic study on an ESD suppressor by the FDTD method

Journal of Electrostatics 71 (2013) 625e634

Contents lists available at SciVerse ScienceDirect

Journal of Electrostatics journal homepage: www.elsevier.com/locate/elstat

Characteristic study on an ESD suppressor by the FDTD method Hsing-Yi Chen*, Pei-Kuen Li Department of Communications Engineering, Yuan Ze University, 135, Yuan-Tung Road, Nei-Li, Chung-Li, Taoyuan Shian 32003, Taiwan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 May 2012 Received in revised form 31 January 2013 Accepted 8 March 2013 Available online 26 March 2013

The FDTD method was used to study the characteristics of an ESD suppressor filled with air, neon, argon, and helium. Obtained capacitance of the ESD suppressor filled with air was validated by measurement data and TDMM simulations. No large differences are found among the obtained capacitances for the ESD suppressor filled with air, neon, argon, and helium. But the ESD suppressor filled with air has a much higher trigger and clamping voltage than the ESD suppressor filled with neon, argon, or helium. The calculated capacitances are presented for different conditions. The ESD currents, charges, and electric fields are also presented. Crown Copyright Ó 2013 Published by Elsevier B.V. All rights reserved.

Keywords: Capacitance Trigger voltage Clamping voltage ESD current ESD suppressor FDTD

1. Introduction Modern electronic products such as cellular phones, personal digital assistants (PDAs), digital cameras and camcorders, multimedia players, computers and peripherals, digital subscriber line (DSL) modems, universal serial bus (USB), high-speed Ethernet, and other portable electronics are becoming smaller, lighter, and simpler while having a higher speed data rate. For these products operating in radio-frequency (RF) and microwave bands, an ESD event can easily damage high sensitivity electronic components which make extensive use of digital technology and fast-speed memory integrated circuit (IC) chips with very high density of multilevel interconnects [1e8]. Therefore, circuit designers must take ESD problems into account in their quest for a functional and reliable product. Based on these requirements, ESD protection elements may be implemented in these high sensitive electronic circuits. ESD protection efforts may be divided into electrical breakdown, thermal melting, and charge shunt to ground. For thermal melting and charge shunt to ground ESD protection, there is a built-in protection circuit which is integrated in the interior of a high sensitivity electronic circuit [3e6]. During an ESD event, all high sensitivity electronic circuits with built-in protection circuits shunt the electrostatic charge to ground. For electrical breakdown ESD

* Corresponding author. Tel.: þ886 3 4636165; fax: þ886 3 4635319. E-mail address: [email protected] (H.-Y. Chen).

protections, a capacitor or an ESD suppressor is added outside the high sensitivity electronic circuit. Initially, the separated discharge plates of the ESD suppressor accumulate positive and negative charges as the power supply is plugged into the printed circuit board (PCB). During an ESD event, these accumulated charges can instantaneously be discharged through the ESD ionized channel which is located between these two charged electrodes. For protecting sensitive electronic circuits operating at RF and microwave frequencies, the ESD suppressors are connected in parallel with the signal lines. When an ESD event occurs, the ESD suppressors clamp the ESD voltage to a level where the sensitive electronic circuits can survive and shunt the ESD current away from the data line and the protected sensitive electronic circuits into the chassis ground [9]. The ESD suppressors are designed to have extremely low capacitance which can provide clamping functions for transient suppression and electromagnetic interference (EMI) filtering against unwanted high-frequency signals that couple into the protected sensitive electronic circuits. In this research work, the FDTD method is used to calculate an ESD suppressor’s charges, capacitances, ESD currents, and gap’s maximum electric fields under different conditions. The ESD suppressor contains two major structures of discharge electrodes, a substrate, an overcoat, two backside electrodes, two connect electrodes, and a solder layer. Originally, the ESD suppressor is filled with air in the space between two discharge electrodes. Alternatively, the space between two discharge electrodes is filled with argon, helium, or neon in order to investigate the performance

0304-3886/$ e see front matter Crown Copyright Ó 2013 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.elstat.2013.03.005

626

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

improvement of the ESD suppressor. The breakdown strengths of argon, helium, and neon have a relatively lower value than that of air. Lower breakdown strengths will result in lower trigger and clamping voltages of an ESD suppressor. In FDTD simulations, a double exponential electric field pulse [10,11] is proposed to model the ESD excitation. The charges, capacitances, ESD currents, gap’s maximum electric fields, trigger voltages, and clamping voltages of the ESD suppressor are presented in this report.

are placed at a distance of 8d on all sides of the scattering objective as shown in Fig. 1, where d is the cell size. In Fig. 1, the computational region of the FDTD model is used for the investigation of electric fields radiated from the ESD suppressor. The details of the FDTD method may be found in many publications and will therefore not be repeated here.

2. The FDTD method

An ESD suppressor is designed for high-speed data interfaces in computing, networking, and digital consumer applications [16]. The ESD suppressor is designed to have extremely low capacitance, superior electrostatic-noise, static-electricity suppression, and fast-response time features during an ESD event so that no electromagnetic interference (EMI) occurs in high sensitivity electronic circuits. The rating and characteristics of the ESD suppressor are listed in Table 1. The function of the ESD suppressor meets with the requirement of IEC 61000-4-2 [17] and the capacitance of the ESD is measured at 1 MHze1.8 GHz. Fig. 2 illustrates the structure of the ESD suppressor. It is composed of two copper discharge electrodes (suppressor element), an Al2O3 substrate, a protective epoxy resin coating (overcoat), two zinc backside electrodes, two nickel connect electrodes, and a solder layer. The copper discharge electrodes have a thumbnail structure on the opposite sides of the spark gap and a thickness of 50 mm. The geometric structure and dimensions of the ESD suppressor models used for FDTD simulations are shown in Fig. 3. Originally, the ESD is filled with air in the space between two discharge electrodes as shown in Fig. 3(a). Alternatively, the space between two discharge electrodes is filled with argon, helium, and neon in order to investigate the variation in capacitance, trigger voltage, clamping voltage, and the maximum electric field of the ESD suppressor. The FDTD model including the ESD suppressor, computational region, and the space between the absorbing boundary and the scattering objective for FDTD simulations is constructed with 10,434,000 cubic cells, where the cell size is taken to be d ¼ 5 mm. Since the operating frequency of 1 GHz is used for electrical breakdown ESD protections [16]. The cell size of 5 mm is capable of giving fairly accurate results for FDTD simulations at 1 GHz. The relative dielectric constants and electrical conductivities of various materials used in the ESD suppressor at frequency of 0.9e6 GHz are obtained from literature [18e22] listed in Table 2. The symbol f represents the frequency in GHz. In FDTD simulations, only nun-magnetic materials are taken into consideration. Therefore the relative permeability mr is assumed to be mr ¼ 1.0 for all materials of the ESD suppressor.

In 1966 Yee [12] first proposed the basic FDTD method for electromagnetic analysis. Due to its powerful and versatile features, in recent years the FDTD method has become a very popular technique for analyzing scattering and absorption problems, designing antennas, solving electromagnetic interference/compatibility (EMI/EMC) problems, manufacturing shielding materials, analyzing microwave engineering, studying bio-electromagnetics, and solving many electromagnetic problems. It is easily used to simulate the electric and magnetic field of arbitrarily shaped structures and materials with inhomogeneities, anisotropy, nonlinearities, and losses. In the FDTD solution procedure, the coupled Maxwell’s equations in differential form are solved for various points of the scatter as well as its surrounding in a time-stepping manner until convergent solutions are obtained. Following Yee’s notation and using centered difference approximation on both the time and space first-order partial differentiations, six finitedifference equations for six unique field components within a unit cell are obtained. In these six finite-difference equations, electric fields are assigned to half-integer (n þ 1/2) time steps and magnetic fields are assigned to integer (n) time steps for the temporal discretization of fields. To ensure numerical stability, the time step dt is set to d/(2Co), where dx ¼ dy ¼ dz ¼ d and Co are the cell size and the speed of light in vacuum, respectively. The center difference approximation ensures that the spatial and temporal discretizations have second-order accuracy, where errors are proportional to the square of the cell size and time increment [12]. An important problem encountered in solving the time-domain electromagneticfield equation by the FDTD method is the absorbing boundary conditions. Several absorbing boundary conditions (ABC) have been proposed in the FDTD method such as second-order Mur [13], and Liao et al. [14], and perfectly matched layer (PML) [15]. In our formulation, the second-order Mur approximation of absorbing boundary conditions [13] is used for the near-field irradiation problems. We use them because they do not require much memory and have a reasonable accuracy. The external absorbing boundaries

3. The ESD suppressor

Fig. 1. The FDTD model of an ESD suppressor. Where d is the cell size.

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

627

Table 1 The rating and characteristics of the ESD suppressor provided by TA-I Technology Co. Ltd. [16] Type (UMS) 04A03T1V1 06A03T1V1 04A03T2V2 06A03T2V2 04A05T1V1 06A05T1V1 04A05T2V2 06A05T2V2 04A12T2V2 06A12T2V2 04A24T2V2 06A24T2V2

Continuous operating voltage (Max.) (VDC)

ESD capability

Trigger voltage (V)

Clamping voltage (V)

Capacitance

Leakage current

Response time

ESD pulse withstand

Direct discharge: 8 KV Air discharge: 15 KV

150

17

<0.05 pF

<1 nA

<1 ns

>1000 pulses

250

25

150

17

250

25

12

250

25

24

250

25

3.3

5.5

4. Characteristic study on an ESD suppressor An important characteristic of an ESD suppressor is the capacitance which is used to protect the data lines of sensitive electronic circuits. The capacitance can provide clamping functions for transient suppression and be used as an EMI filter to avoid unwanted signals coupling into the sensitive electronic circuits. For highspeed data lines with data rates more than 100 megabits per second, the capacitance helps to eliminate unwanted signals. This may, however, also filter the data signals themselves and thus result in data waveform distortion and signal without integrity. Therefore, data errors can be introduced into the high-speed electronic circuits. The waveform distortion may take the form of rounded leading and trailing edges of high/low state transitions due to slower rise and fall times as illustrated in Fig. 4 [23]. Higher capacitances have slower response time features so that data waveform distortions may occur. In order to avoid signal distortion, it is necessary to make sure that the capacitance of the suppressor is not too high for the ESD protection. For extremely high-speed applications, the ESD suppressors should be designed to have extremely low capacitance values. It should be noted that the capacitance of an ESD suppressor is independent of ESD voltage. The capacitance depends on discharge electrode shapes, types of gas or materials filling in the gap, gap distances between two discharge electrodes, materials of the overcoat, and operating frequencies. The influence of discharge electrode shapes on the capacitance of an ESD suppressor is reported in previous research work [9] and will not be studied here. All ESD suppressors have a limited life-time span. The discharge electrodes must withstand up to 1000 times of ESD strikes (ESD pulses) regulated by the IEC 61000-4-2 [17]. The higher the discharge voltage is, the higher energy that an electron particle obtains from an ESD event. High electric field strength with high energy electron particles at the tip of an electrode leads to resistive heating via ESD current which will cause the electrode surface to melt. The ESD current results from an avalanche contribution when

Fig. 2. Illustration of the structure of the ESD suppressor.

the ionization occurs in the spark gap. After 10e100 discharges, the surface roughness (<2 mm) of an electrode is about 1 order below the gap distance (60 mm) [24,25]. The life-time span will decrease with increasing surface roughness. Therefore, how to lower the discharge voltage of an ESD suppressor becomes an important issue. In this study, the trigger and clamping (holding) voltages will be investigated for different gases filled in an ESD suppressor. For a pair of discharge electrodes with a thumbnail structure, the center of the spark gap has the shortest distance, while the edges of the spark gap have the longest distance. When the electric field at the edges of the spark gap is over the gas breakdown strength, the discharge voltage is defined as the trigger voltage. On the other

Fig. 3. Detail dimensions of the ESD suppressor model for FDTD simulations. The cell size d equals 5 mm.

628

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

Table 2 Relative dielectric constants and conductivities of various materials used in the ESD suppressor at frequencies of 0.9e6.0 GHz [18e22]. The unit of frequency f is in GHz. Materials

Relative dielectric constant (εr)

Conductivity (S/m)

Air Helium Neon Argon Al2O3 substrate Epoxy resin Copper Nickel Zinc Solder

1.0 1.000065 1.00013 1.0005172 10e0.035  f 4.2e0.069  f 1.0 1.0 1.0 1.0

0.295  1014 0.0 0.0 0.0 f  εr  0.5563  105 f  εr  0.9735  103 5.8  107 1.45  107 1.67  107 7.0  106

hand, when the electric field at the center of the spark gap decreases from the maximum value down to the gas breakdown strength, the discharge voltage is defined as the clamping (or holding) voltage. A discharge at lower discharge voltage occurs usually at shorter gap distances. A discharge will not occur when the discharge voltage is below the clamping voltage. The trigger and clamping voltages depend on the gas breakdown strength and the spark gap distance. In order to extend the suppressor’s life-time span, the trigger and clamping voltages should be designed as low as possible. The breakdown strength of air, argon, helium, and neon is listed in Table 3 [26e28]. Based on breakdown strengths and spark gap distances, the trigger and clamping voltages of the ESD suppressor filled with air, argon, helium, and neon for various spark gap distances are calculated and shown in Figs. 5 and 6. It is clear that the trigger or clamping voltage increases as the gap distance increases from 5 to 50 mm. The trigger or clamping voltage obtained for the ESD suppressor filled with air is much higher than that obtained for the ESD suppressor filled with argon, helium, and neon, respectively. In FDTD simulations, an ESD voltage V0(t) with a double exponential voltage pulse [10,11] similar to a standard ESD pulse current waveform, specified in International Standard IEC 801-2, is used to model the ESD excitation in the gap between the two discharge electrodes expressed by


 V0 ðtÞ ¼ Vesd eat  ebt ; for t  0

(1)

The double exponential voltage pulse is modeled with the sum of a fast wave and a slow wave, where constants a, and b are given below

Table 3 The breakdown strength of air, neon, argon, and helium. Gas type

Breakdown strength (V/m)

Air Neon Argon Helium

2.9  106 8.5  105 6.12  105 5.1  105

For the slow wave:

a ¼ 4:55  107 ; and b ¼ 5:00  107 for t  6:81 ns

The maximum voltage occurs at time tm ¼ ln(b/a)/(b  a), where constants a and b are equal to 4.55  108 and 5.00  108, respectively. From the calculation, it is found that the maximum voltage occurs at the time tm ¼ 2.09 ns. The waveform has a 1.20 ns rise time (the pulse increases from 10% to 90% of its peak value). It should be noted that the rise time of a standard ESD waveform is of order 0.7e1.0 ns according to the regulation required by the IEC 61000-4-2 [17]. The ratio of the voltage at 30 ns to the maximum voltage at 2.09 ns is 0.53 and the ratio of the voltage at 60 ns to the voltage at 30 ns is 0.48. Taking the trigger voltage as the maximum voltage, Vesd can be calculated for the ESD suppressor filled with air, argon, helium, and neon under the conditions of various spark gap distances as listed in Table 4. The ESD pulse voltage waveforms for the ESD suppressor filled with air, argon, helium, and neon under the condition of the spark gap distance of 5 mm are shown in Fig. 7. From Fig. 7, it is observed that the obtained ESD voltage for the ESD suppressor filled with air and helium has a larger and smaller value, respectively, while the obtained ESD voltages for the ESD suppressor filled with neon and argon are in between. Different trigger voltages for ESD suppressors filled with air, argon, helium, and neon are also shown in Fig. 7. The trigger voltages of 217.5, 45.9, 38.25, and 63.75 V for ESD suppressors filled with air, argon, helium, and neon are found in the rising pulses at 2.59, 2.55, 2.58, and 2.58 ns, respectively. For FDTD simulations, the ESD voltage V0(t) is transferred into an electric field over the gap between the two discharge electrodes by E(t) ¼ V0(t)/d, where d ¼ 1, 3, 5, 15, and 101d is the distance of the gap between the two discharge electrodes. Applying the excited

For the fast wave:

a ¼ 4:55  108 ; and b ¼ 5:00  108 for 6:81 ns > t  0

(2)

Fig. 4. Illustration of signal waveform distortion due to slow rise and fall times [23].

(3)

Fig. 5. Trigger voltage versus spark gap distance.

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

629

Fig. 6. Clamping voltage versus spark gap distance.

electric fields in FDTD simulations, electric fields at any location inside and outside the ESD suppressor can be calculated. From boundary conditions, electric fields obtained near the two copper electrodes can be used to calculate the surface charge density rs on their surfaces expressed by

V$D ¼ rs ¼ εr ε0 E2n  ε0 E1n ;

(4)

where E1n and E2n are the normal components of electric fields in and outside the discharge electrodes, D is the electric flux density, εr and ε0 are the relative dielectric constant of materials and the dielectric constant of air, respectively. The total charge on any one surface of the two discharge electrodes is determined by integrating the surface charge density over the surface expressed by

Z Q ðtÞ ¼

rs ðtÞds0

(5)

s

Fig. 7. ESD pulse voltage waveforms for the ESD suppressor filled with air, argon, helium, and neon under the condition of a spark gap distance of 5 mm.

C ¼

Q ðtÞ V0 ðtÞ

(6)

It should be noted that the capacitance depends only on the geometric structure, material used in the ESD suppressor, and the operating frequency; it does not depend on the applied voltage or current. Under the conditions of the ESD suppressor filled with air, a spark gap distance of 5 mm, the ESD suppressor with Al2O3 substrate and epoxy resin overcoat, and operating frequency of 1 GHz, the capacitance of 0.0242 pF calculated by the FDTD method as shown in Fig. 9 is in good agreement with the measurement data of 0.025 pF provided by TA-I Technology Co. [16] and the numerical result of 0.0264 pF obtained by the time-domain moment method (TDMM) [9]. Simulation results of capacitance for the ESD suppressor filled with helium, neon, and argon under the same conditions are also shown in Fig. 9. The relative dielectric constants

where s0 denotes the surface of the two discharge electrodes. Fig. 8 shows the total charges on the two discharge electrodes for the ESD suppressor filled with air, helium, neon, and argon under the condition of the spark gap distance of 5 mm. The operating frequency is at 1 GHz. It can be seen that the total charges on the two discharge electrodes are the same in magnitude but opposite on electricity. Once the total charges are obtained, the capacitance can also be calculated by

Table 4 Parameter Vesd for the ESD suppressor filled with air, argon, helium, and neon under the conditions of various spark gap distances. The unit is in V. Gap distance (mm)

Air

Helium

Neon

Argon

Vesd

5 10 15 20 25 30 35 40 45 50

Fast

Slow

Fast

Slow

Fast

Slow

Fast

Slow

6271 6689 7107 7525 7943 8361 8779 9197 9615 10,034

3574 3812 4050 4289 4527 4765 5004 5242 5480 5718

1102 1176 1249 1323 1396 1470 1544 1617 1691 1764

628 670 712 754 796 838 880 921 963 1005

1838 1960 2083 2205 2328 2450 2573 2695 2818 2941

1047 1117 1187 1257 1327 1396 1466 1536 1606 1676

1323 1411 1499 1588 1676 1764 1852 1941 2029 2117

754 804 854 905 955 1005 1056 1106 1156 1206

Fig. 8. Total charges on the two discharge electrodes for the ESD suppressor filled with air, helium, neon, and argon under the condition of a spark gap distance of 5 mm. The operating frequency is at 1 GHz.

630

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

Fig. 9. Simulation results of capacitance for the ESD suppressor filled with air, helium, neon, and argon under the condition of the spark gap distance of 5 mm. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat operates at 1 GHz.

Fig. 11. Capacitance versus relative dielectric constant of the overcoat. The relative dielectric constant of the overcoat has a variety of 1e6. The ESD suppressor is filled with air, helium, neon, and argon and has a spark gap distance of 25 mm. Al2O3 substrate is used for the ESD suppressor. The operating frequency is at 1 GHz.

only have a small variation among air, helium, neon, and argon. As expected, the obtained capacitance for the ESD suppressor filled with air has a little higher value than that obtained for the ESD suppressor filled with helium, neon, or argon. The obtained capacitances for the ESD suppressor filled with helium, neon, and argon are 0.0216, 0.02384, and 0.0227 pF, respectively. However, it should be noted that the ESD suppressor filled with air has a much higher trigger and clamping voltages than the ESD suppressor filled with neon, argon, or helium. It should also be noted that higher trigger and clamping voltages will shorten the suppressor’s lifetime span. After checking the validity of the FDTD method, it was used for further studies on the capacitance. In each study, one of the four parameters including relative dielectric constant of overcoat, relative dielectric constant of substrate, spark gap distance, and

operating frequency is changed and the other three parameters are kept unchanged. Fig. 10 shows the obtained capacitance versus spark gap distance for the ESD suppressor filled with air, helium, neon, and argon. It is clear that the capacitance decreases with an increase in the spark gap distance. The goal of design of an ESD suppressor is to reduce the capacitance value as small as possible, but the smaller capacitance with a larger spark gap distance will result in a higher trigger voltage. The higher manufacturing cost for an ESD suppressor with a larger spark gap distance may also be an important issue. Fig. 11 shows the obtained capacitance versus relative dielectric constant of the overcoat for the ESD suppressor filled with air, helium, neon, and argon under the condition of a spark gap distance of 25 mm. Obviously, the capacitance increases

Fig. 10. Simulation results of capacitance for the ESD suppressor filled with air, helium, neon, and argon under the conditions of the spark gap distance of 5e50 mm. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat operates at 1 GHz.

Fig. 12. Capacitance versus relative dielectric constant of the substrate. The ESD suppressor is filled with air, helium, neon, and argon and operates at 1 GHz. The spark gap distance is 25 mm and the relative dielectric constant of the substrate has a variety of 1e20. Epoxy resin overcoat is used for the ESD suppressor.

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

631

Fig. 15. Time response of the ESD current for the ESD suppressor filled with helium. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat has a spark gap distance of 5 mm and operates at 1 GHz. Fig. 13. Capacitance versus frequency. The ESD suppressor is filled with air, helium, neon, and argon and operates at frequencies of 0.9e6 GHz. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat has a spark gap distance of 5 mm.

with the increase of the relative dielectric constant of the overcoat. Therefore, dielectric materials with smaller relative dielectric constants should be adopted as the overcoat of an ESD suppressor. Fig. 12 shows the obtained capacitance versus relative dielectric constant of the substrate for the ESD suppressor filled with air, helium, neon, and argon under the condition of a spark gap distance of 25 mm. From Fig. 12, it is found that the capacitance increases with the increase of the relative dielectric constant of the substrate. Again, dielectric materials with smaller relative dielectric constants should be adopted as the substrate of an ESD suppressor. Fig. 13 shows the obtained capacitance versus the operating frequency for the ESD suppressor filled with air, helium, neon, and argon under the condition of a spark gap distance of 5 mm. From Fig

Fig. 14. Time response of the ESD current for the ESD suppressor filled with air. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat has a spark gap distance of 5 mm and operates at 1 GHz.

13, it should be noted that the capacitance decreases as the operating frequency increases from 1 to 6 GHz. This trend mainly results from the value decrease in relative dielectric constants of overcoat and substrate as the frequency increases from 1 to 6 GHz. The ESD current is calculated by the following equation:

IðtÞ ¼

vQ ðtÞ Q ðtÞ  Q ðt  DtÞ ; ¼ Dt vt

(7)

where Dt is the time increment and Q(t) is the total charge at time t in one of the two electrodes. Time responses of ESD currents for the ESD suppressor filled with air, helium, neon, and argon under the condition of a spark gap distance of 5 mm are shown in Figs. 14e17. It is found that ESD currents with a transient waveform reach a maximum value in a variety of 3.2e5.6 mA at 0.9 ns and decays sharply as the time increases. The duration of the transient

Fig. 16. Time response of the ESD current for the ESD suppressor filled with Neon. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat has a spark gap distance of 5 mm and operates at 1 GHz.

632

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

Fig. 19. Maximum electric fields distributed on the horizontal plane which consists of the center of the spark gap. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat is filled with air. The ESD suppressor has a spark gap distance of 5 mm and operates at 1 GHz. Fig. 17. Time response of the ESD current for the ESD suppressor filled with argon. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat has a spark gap distance of 5 mm and operates at 1 GHz.

phenomenon is very insignificant as compared with the operation time of an electronic system. Yet it is very important because the transient ESD current may cause a serious failure in high sensitive information circuits if the ESD current is not removed from the protected high sensitive information circuits. Maximum ESD currents versus spark gap distance for the ESD filled with air, helium, neon, and argon are shown in Fig. 18. It is found that the maximum ESD current decreases as the spark gap distance increases. In order to investigate the electric field near the ESD suppressor, electric field distributions on the horizontal plane which consist of the center of the spark gap are calculated and shown in Figs. 19e22. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat is filled with air, neon, argon, and helium, respectively. The ESD suppressor has a spark gap distance of 5 mm and operates at a frequency of 1 GHz. From Figs. 19e22, it is clear that the maximum electric fields decay from 4.338  107 to 1.436  104, from

Fig. 18. Maximum ESD current versus spark gap distance. The spark gap distance has a variety of 5e50 mm. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat is filled with air, helium, neon, and argon and operates at 1 GHz.

Fig. 20. Maximum electric fields distributed on the horizontal plane which consists of the center of the spark gap. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat is filled with neon. The ESD suppressor has a spark gap distance of 5 mm and operates at 1 GHz.

Fig. 21. Maximum electric fields distributed on the horizontal plane which consists of the center of the spark gap. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat is filled with argon. The ESD suppressor has a spark gap distance of 5 mm and operates at 1 GHz.

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

633

occurred. Outside the spark gap, the maximum electric field is below the breakdown strength, but the magnitude of the electric field resulting from ESD events is still very large as the distance from the ESD suppressor is less than 2 mm. This observation hints that the high sensitivity electronic circuits should be kept away from the ESD suppressor, because high sensitivity electronic circuits connected to external ports are susceptible to a damaging ESD pulse from the operating environment and from peripheral interference. The ESD suppressor should be placed at least 1 cm from high sensitivity electronic circuits. Maximum electric field versus spark gap distance for the ESD suppressor filled with air, helium, neon, and argon is shown in Fig. 23. As expected, it is found that the maximum electric field decreases sharply as spark gap distance increases from 5 to 50 mm. It should be noted that the trigger voltage will increase with an increase in the spark gap distance as shown in Fig. 5. 5. Conclusions Fig. 22. Maximum electric fields distributed on the horizontal plane which consists of the center of the spark gap. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat is filled with helium. The ESD suppressor has a spark gap distance of 5 mm and operates at 1 GHz.

1.266  107 to 8.136  103, from 9.169  106 to 6.171 103, and from 7.637  106 to 2.736  103 V/m for the ESD suppressor filled with air, neon, argon, and helium as the distance from the spark gap increases from 0.0 to 2.125 mm, respectively. Basically, the electric field produced by an ESD event has a near-field phenomenon which means the electric field decreases sharply as the distance from the radiation source increases. It is very clear that the electric field has a maximum value in the spark gap as shown in Figs. 19e22. Outside the spark gap, the electric field has a variation of exponential decay as the distance from the air gap increases. From the numerical results of maximum electric fields obtained for the ESD suppressor filled with air, neon, argon, and helium shown in Figs. 19e22, it can be seen that the maximum electric fields in the spark gap are over the air, neon, argon, and helium breakdown strengths of 2.9  106, 8.5  105, 6.12  105, and 5.1 105 V/m, respectively. This phenomenon indicates that an ESD event is confirmed to have

Fig. 23. Maximum electric field versus spark gap distance. The ESD suppressor is filled with air, helium, neon, and argon. The ESD suppressor with Al2O3 substrate and epoxy resin overcoat has a spark gap distance of 5 mm and operates at 1 GHz.

The FDTD method was used to calculate the charge, capacitance, trigger voltage, clamping voltage, ESD current, and electric field of an ESD suppressor filled with air, neon, argon, and helium under different conditions. The obtained capacitance of the ESD suppressor filled with air under the condition of a spark gap distance of 5 mm was validated by measurement data provided by TA-I Technology Co. and the numerical result obtained by the TDMM. After checking the validity of the FDTD method, the FDTD method was used for further studies on the capacitance, ESD current, and electric field for the ESD suppressor filled with air, neon, argon, and helium under different conditions. Only a tiny variation of obtained capacitances for the ESD suppressor filled with air, neon, argon, and helium under the same conditions was found. But it should be noted that the ESD suppressor filled with air has a much higher trigger and clamping voltage than the ESD suppressor filled with neon, argon, or helium. Higher trigger and clamping voltages may shorten the suppressor’s life-time span. As expected, the capacitance decreases with an increase in the spark gap distance for the ESD suppressor filled with air, neon, argon, and helium. However, the smaller capacitance with a larger spark gap distance will result in a higher trigger voltage. From obtained capacitances, it is shown that the substrate or the overcoat with a higher dielectric material used for an ESD suppressor will produce higher capacitances. Higher capacitances have slower response time features so that data waveform distortions and signals without integrity may occur. It is observed that the capacitance decreases as the operating frequency increases from 1 to 6 GHz. The transient ESD current also plays an important role in protecting high sensitive information circuits. It is observed that ESD currents with a transient waveform reach a maximum value in a variety of 3.2e5.6 mA at 0.9 ns and decays sharply as the time increases. The duration of the transient phenomenon is very insignificant as compared with the operation time of an electronic system. The maximum ESD currents affected by spark gap distance were also studied. It is found that the maximum ESD current decreases with an increase in the spark gap distance. The maximum electric fields near the ESD suppressor were also calculated. It is very clear that the electric field has a maximum value in the spark gap. Outside the spark gap, the electric field has a variation of exponential decay as the distance from the spark gap increases. The maximum electric fields of 4.338  107, 1.266  107, 9.169  106, and 7.637  106 V/m are obtained for the ESD suppressor filled with air, neon, argon, and helium under the condition of a spark gap of 5 mm, respectively. From obtained maximum electric fields, it can be seen that the maximum electric fields in the spark gap are over the air, neon, argon, and helium breakdown strengths respectively. This phenomenon indicates that

634

H.-Y. Chen, P.-K. Li / Journal of Electrostatics 71 (2013) 625e634

an ESD event is confirmed to have occurred. Outside the spark gap, the maximum electric field is below the breakdown strength, but the magnitude of the electric field resulting from ESD events is still very large as the distance from the ESD suppressor is less than 2 mm. Therefore, it is suggested that the ESD suppressor should be placed at least 1 cm away from high sensitivity electronic circuits. It is also found that the maximum electric field decreases sharply as spark gap distance increases from 5 to 50 mm. In this study, the research results may help to better understand the trade-off policy for performance improvement of ESD suppressors. References [1] C. Duvvury, R.N. Rountee, R.A. McPhee, ESD protection: design and layout issues for VLSI circuits, IEEE Trans. Ind. Appl. 25 (1) (1989) 41e47. [2] C. Grewing, K. Winterberg, S. Waasen, M. Friedrich, G. Puma, A. Wiesbaue, C. Sandner, Fully integrated distributed power amplifier in CMOS technology, optimized for UWB transmitters, in: Proc. IEEE Radio Frequency Integrated Circuits Symp., 2004, pp. 87e90. [3] C.Y. Lin, M.\D. Ker, G.X. Meng, Low-capacitance and fast turn-on SCR for RF ESD protection, IEICE Trans. Electron. 8 (2008) 1321e1330. [4] M.D. Ker, C.Y. Lin, Low-capacitance SCR with waffle layout structure for on-chip ESD protection in RF ICs, IEEE Trans. Microw. Theor. Tech. 56 (5) (2008) 1286e1294. [5] M.D. Ker, W.J. Chang, ESD protection design with on-chip ESD bus and high-voltage-tolerant ESD clamp circuit for mixed-voltage I/O buffers, IEEE Trans. Electron Dev. 55 (6) (2008) 1409e1416. [6] M.D. Ker, C.C. Yen, Transient-to-digital converter for system-level electrostatic discharge protection in CMOS ICs, IEEE Trans. Electromagn. Compatibility 51 (3) (2009) 620e630. [7] K.S. Im, J.H. Ko, S.J. Kim, C.H. Jeon, C.S. Kim, K.T. Lee, H.G. Kim, I.H. Son, Novel ESD strategy for high voltage non-volatile programming pin application, Microelectron. Reliab. 46 (2006) 1664e1668. [8] Y.B. Shi, W.Y. Yin, J.F. Mao, P. Liu, Q.H. Liu, Transient electrothermal analysis of multilevel interconnects in the presence of ESD pulses using the nonlinear time-domain finite-element method, IEEE Trans. Electromagn. Compatibility 51 (3) (2009) 774e783. [9] H.Y. Chen, Y. Suo, C.T. Kuo, J.H. Qiu, Design and analysis of electrostatic discharge suppressors with ultra-low capacitance used for IC protection, J. Electrostat. 69 (6) (2011) 604e610. [10] M. Camp, H. Garbe, Parameter estimation of double exponential pulses (EMP, UWB) with least squares and Nelder Mead algorithm, IEEE Trans. Electromagn. Compatibility 46 (4) (2004) 675e678.

[11] V. Prasad Kodali, Engineering Electromagnetic Compatibility: Principle, Measurement, and Technologies, IEEE Press, New York, 1996. [12] K.S. Yee, Numerical solution of initial boundary value problems involving Maxwell’s equations in isotropic media, IEEE Trans. Antenn. Propag. 14 (5) (1966) 302e307. [13] G. Mur, Absorbing boundary conditions for the finite-difference approximation of the time-domain electromagnetic field equation, IEEE Trans. Electromagn. Compatibility 23 (1981) 377e382. [14] Z.P. Liao, H.L. Wong, B.P. Yang, Y.F. Yuan, A transmitting boundary for transient wave analysis, Sci. Sin. A 27 (10) (1984) 1063e1076. [15] J.P. Berenger, A perfectly matched layer for the absorption of electromagnetic waves, J. Comp. Phys. 114 (1994) 185e200. [16] TA-I Technology Co., Ltd., Taoyuan, Taiwan [Online]. Available: http://www. tai.com.tw. [17] International Electrotechnical Commission, Electromagnetic Compatibility (EMC) e Part 4e2: Testing and Measurement Techniques e Electrostatic Discharge Immunity Test. IEC 61000-4-2, second ed., 2008. [18] Microwave Thin Film Circuits e Standard Materials Specifications, Properties and Conversion Charts, US MICROWAVES [Online]. Available: http://www. usmicrowaves.com/thinfilm/material_specifications.htm. [19] M. Xu, T.H. Hubing, J. Chen, T.P. Van Doren, J.L. Drewniak, R.E. DuBroff, Power-bus decoupling with embedded capacitance in printed circuit board design, IEEE Trans. Electromagn. Compatibility 45 (1) (2003) 22e30. [20] S.D. Pawar, P. Murugavel, D.M. Lal, Effect of relative humidity and sea level pressure on electrical conductivity of air over Indian ocean, J. Geophys. Res. 114 (D2) (2009). CiteID D02205. [21] A.A. Maryott, F. Buckley, Table of Dielectric Constants and Electric Dipole Moment of Substances in the Gaseous State, National Bureau of Standards Circular 537, Washington, DC, 1953. [22] A. von Hippel, Dielectric Materials and Application, Artech House, Boston, 1995. [23] TECH BRIEF, Capacitance and Signal Integrity, Littelfuse, Inc., January 2002. Available e-mail: http://dev.littelfuse.com/da. [24] S. Bonisch, W. Kalkner, D. Prommerenke, Modeling of short-gap ESD under consideration of different discharge mechanisms, IEEE Trans. Plasma Sci. 31 (4) (2003) 736e744. [25] S. Nitta, A. Mutoh, K. Miyajima, Generation mechanism of showering noise waveforms e effect of contact surface variations and moving velocity of contact, IEICE Trans. Commun. E79-B (4) (1996) 468e473. [26] L.I. Berger, Dielectric Strength of Insulating Materials, Table 1: Dielectric Strength of Gases [Online]. Available: http://www.mendeley.com/research/ dielectric-strength-insulating-materials/. [27] K.P. Brand, Dielectric strength, boiling point and toxicity of gases-different aspects of the same basic molecular properties, IEEE Trans. Electr. Insul. EI-17 (5) (1982) 451e456. [28] O.P. Gandhi, Microwave Engineering and Applications, Pergamon Press, New York, 1981, p. 24.