Scribd
Upload
119 views

Littelfuse Selecting An Appropriate Esd Device Application Note

ESD protection devices

Uploaded by

luroguita
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
119 views

Littelfuse Selecting An Appropriate Esd Device Application Note

ESD protection devices

Uploaded by

luroguita
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Selecting an Appropriate ESD Device

Application Note:

Selecting an Appropriate ESD Device

Background

Board designers today are confronted with a multitude


of choices for ESD protection. Often the designer is
constrained by certain limits such as the amount of
parasitic capacitance his/her application can withstand or
the required ESD level the board must pass without fail.
More often than not, the constraints do not narrow
down the available number of ESD devices to a
manageable list. This white paper will provide guidance
to the designer that will help him/her in choosing an
ESD device that will give the best chance of a
successful first pass design.

There are several types of ESD suppression devices on


the market today. Examples include MLVs (Multi-Layer
Varistors), polymer ESD suppressors, and silicon diodes
or arrays. For simplicity this paper will only focus on
silicon devices as their performance tends to be superior
in terms of ESD clamping capability. Nevertheless, the
points discussed hereafter can be applied generically to
any ESD device regardless of its technology.
The two most common silicon protection devices are
TVS/Zener diodes (Figure 2) and Diode/Rail Clamps
(Figure 3) which are both intended to present a low
resistance shunt path to GND during an ESD event.
(The main difference between the two is the amount of
parasitic capacitance each structure will add to the I/O
under protection.)

Before getting into the selection criteria, let us review a


few basic ESD topologies and their typical turn-on
characteristics. (Note: Throughout this paper the ESD
threat/pulse referenced is defined by the IEC61000-4-2
standard shown below in Figure 1.)
Figure 1

Each structure is intended to steer both positive and


negative ESD pulses away from the IC being protected.
For positive transients the TVS/Zener of Figure 2 will
turn-on after a voltage VZ (typically 6-8V) is reached
and provide a resistive shunt to GND. Likewise the
diode array in Figure 3 will steer a positive current
through the upper diode and into the internal TVS
device when a voltage VF+VZ is reached (typically 6-8V).
For negative ESD pulses, both structures behave the
same in that they each will conduct when -VF (typically
0.6-0.8V) is exceeded or the bus being protected falls
one diode drop below GND. With this basic
understanding in mind, let us examine the electrical
characteristics typically given by ESD vendors on their
device datasheets.
Figure 2

To IC

Figure 3

VCC

ESD
Entry
ESD
Entry

To IC

2009 Littelfuse, Inc

Selecting an Appropriate ESD Device

Electrical Characteristics
The majority of TVS/Zener ESD devices (Figure 2) are
specified with four main characteristics: ESD Level,
Reverse Standoff Voltage/Leakage, Capacitance, and
Breakdown Voltage. (There is a fifth characteristic that
is not as prevalent, which well discuss later.)
The ESD level only tells the designer what level of ESD
the device itself can withstand before damage occurs.
It gives no guarantee that the IC being protected will
survive the same level of ESD given on the devices
datasheet. Additionally, the breakdown voltage is
usually between 6V-8V at 1mA or 10mA. This only gives
the designer a data point to ensure the ESD device
remains inactive during normal circuit operation. It gives
no indication about the shunt resistance or clamping
voltage he/she can expect under an ESD strike. The
other two characteristics relate to the devices
parasitics and again offer no insight about the ESD
devices performance during an ESD transient.
Similarly, the rail clamp or diode array structure in Figure
3 is usually specified with the same four characteristics:
ESD Level, Reverse Standoff Voltage/Leakage,
Capacitance, and Breakdown Voltage. The breakdown
voltage of the internal TVS diode is usually given (i.e. 67V) but there is no mention how the steering diodes
plus TVS will behave or clamp during an ESD pulse.
In summary, the electrical characteristics mentioned
above provide characterization data on the ESD
protector only, and give the designer no information
about the effectiveness of the ESD device under stress.
We will revisit the electrical characteristics later but for
now let us examine some of the common plots/graphs
found in ESD device datasheets.

Characteristic Plots
A board designer can find many waveforms in an ESD
device datasheet. The following list only names a few
of the more common plots:
1

Capacitance vs. Reverse Bias


Power Derating Curve
3
Insertion Loss (S21), and
4
ESD Response or Clamp
2

The first three do not give any information about the


ESD devices effectiveness to clamp an ESD pulse. The
last plot ESD Response appears to be the most
helpful, but there are a couple of unknowns.

2009 Littelfuse, Inc

First of all, some datasheets state that the input pulse is


8kV, the maximum level given per the IEC61000-4-2
standard, but others give no such indication. Most all
board designers want to meet at a minimum the 8kV
level so its important to understand what ESD level the
protection device was measured against. An example
of this datasheet waveform is shown in Figure 4 which
has the same shape of the input pulse (Figure 1), but its
a dampened or clamped version of it.
Figure 4

Secondly, there is seldom any mention about the setup


these waveforms were captured under. To
give the best results, the ESD strike would need to be
applied right at the pin of the ESD device to limit the
voltage spike generated by any parasitic trace
inductance or

VLP = LP

di
dt

These setup conditions are very ideal which rarely


occurs in the real world. Correctly designed boards
would try to place the ESD device as close as possible
to the ESD point of entry, but in any event, there is
almost always some distance of PCB trace between the
two. (Note: As little as 1nH of inductance can induce a
30V spike on the PCB trace which is in addition to the
actual clamping voltage obtained by the ESD device).
Lastly, many of the ESD Response waveforms fail to
mention the external attenuation or attenuator size that
was used to protect the oscilloscope while the
measurement was being made. Some datasheets point
out the attenuator size used (such as 10x or

Selecting an Appropriate ESD Device

100x) which one can easily use to calculate the true


clamping voltage; however, some datasheets give no
indication. Therefore, the designer must be cautious of
this and never take the screenshot voltage as the final
clamping voltage he/she can expect from the ESD
device.

Most ESD vendors include the results of this test on


their datasheets either in the electrical table or by
means of a plot. Most designers tend to overlook these
data points since they are either unfamiliar with the -45 test or believe that an 8/20s current pulse is
outside the scope of their design.

Outside of the aforementioned electrical characteristics


and waveforms, the board designer has very little
information to decide which ESD device will give
him/her the best performance. The following section
will identify a key parameter that can be universally
applied to put all ESD devices on equal ground.

Figure 6

Dynamic Resistance
As mentioned earlier, the ultimate goal of a protection
device is to provide the lowest resistance shunt path to
GND under an ESD event. Ideally, all current would be
steered into the ESD device so that the protected IC
would not have to dissipate any of the energy input into
the circuit. Figure 5 depicts the ESD protection device
as a variable resistor which will be high impedance (low
leakage) during normal circuit operation and low
impedance during any EOS (Electrical Over Stress) or
ESD event.
Figure 5
To IC

ESD Entry

ESD Device

No matter how the data is presented (electrical table or


plot) calculating the dynamic resistance is a simple
exercise. First, find two points which include a clamping
voltage and current level. It is usually advisable to select
the two lowest current levels where internal device
heating has the smallest effect. Next, just do the math.
The example below uses data taken from the electrical
table of Littelfuses SP3001 series datasheet which is a
low capacitance (0.65pF) diode/rail clamp array. In the
table, a parameter called Clamping Voltage can be
found that lists input currents of 1A and 2A which
(typically) gives a 9.5V and 10.6V clamp voltage,
respectively. Therefore:

R DYNAMIC =

To compare ESD devices and evaluate their effective


resistance during an ESD pulse, it is pertinent to remove
all outside factors or variables some of which were
previously discussed. To that end, there is another
standard, IEC61000-4-5, that defines a current pulse
(Figure 6) with a much slower leading edge than an ESD
pulse. This waveform has an 8s risetime in
comparison to the 1ns risetime of an ESD pulse. This
allows any lead or trace inductance to be mitigated so
test setups cannot affect the test data, and quite often
no external attenuator is needed.

2009 Littelfuse, Inc

10.6V 9.5V
= 1.1
2 A 1A

This value is essentially the shunt resistance the ESD


device will provide under an ESD pulse. With this
simple calculation, the board designer can now
effectively evaluate any ESD device on equal ground
irrespective of the possible datasheet differences.

Selecting an Appropriate ESD Device

Important Considerations
It is usually common practice to think of an ESD pulse in
terms of a current pulse. As can be seen in Figure 7, a
simplified model of an ESD generator, the voltage on
the capacitor (i.e. 8kV) is discharged through a 330
resistor into an ESD protection device. The ESD device
is typically <10 so the output of the ESD generator
more closely resembles a current pulse as opposed to a
voltage spike.

input pin for a few nanoseconds. Often the board


designer only knows what DC voltage he/she cannot
exceed without causing irreparable damage to the IC.
So while its unclear if the very brief existence of 66V on
the input will cause damage, its apparent that Device A
will give him/her a far better chance at achieving a
successful first pass design.

Conclusions
This paper has given an overview of a couple of ESD
topologies and explained the difficulties a designer may
face in choosing the correct ESD component for his/her
application. Time and again limiting criteria such a
parasitic capacitance, standoff voltage, ESD withstand,
etc. are not enough to sort through and vet all the
possible choices for a suitable protection device.

Figure 7

And now going back to Figure 1, an ESD pulse reaches


its highest voltage in the first nanosecond and is quickly
dissipated thereafter in under 100ns. Table 1 shows the
(IEC61000-4-2 specified) relation between the two.

Using the sometimes overlooked Clamping Voltage or


tP=8/20s specification, the board designer can
quickly calculate the dynamic resistance of each device
to see which one offers him/her the lowest shunt
resistance to GND.

Citations/References:
1)

Figure 1 was recreated from the IEC61000-4-2 specification.


Figure 6 was recreated from the IEC61000-4-5 specification.
3)
Figure 7 was recreated from the IEC61000-4-2 specification.
4)
Table 1 was recreated from the IEC61000-4-2 specification.
2)

Table 1

IEC
Level

Voltage
(kV)

1
2
3
4

2
4
6
8

Peak
Current
10% (A)
7.5
15
22.5
30

Current
@ 30ns
(A)
4
8
12
16

Current
@ 60ns
(A)
2
4
6
8

This background was given to help illustrate the point


below, that the dynamic resistance can only be used in
relative terms or for comparison purposes. For example,
take two TVS/Zener ESD devices with the same
breakdown voltage (VZ) but each with a different
dynamic resistance. For arguments sake, lets say
Device A=1.0 and Device B=2.0. Device B will
clearly give a higher peak clamping voltage to an 8kV
ESD strike.

Written by Chad Marak and Jim Colby

Littelfuse, Inc.
8755 West Higgins Road, Suite 500
Chicago, IL 60631 USA
Phone: (773) 628-1000
[email protected]

www.littelfuse.com

Clamp A = VZ + I PEAK R DYN , A = 6V + 30 A 1.0 = 36V


Clamp B = VZ + I PEAK R DYN , B = 6V + 30 A 2.0 = 66V
These values may not mean much to the board designer
as these peak voltages only exist on the data line or

2009 Littelfuse, Inc

You might also like