Contrary to popular belief, the shunting of leads or terminals of static-sensitive devices or boards doesn’t always protect them from ESD.1 What, then, is the risk of damage? And what should the shunt material be? To find these answers, some high-voltage discharge tests were conducted. Here are the results.
Test Procedures
The test apparatus was a loop of 0.041″-dia copper wire with three pairs of alligator clips on a 4″ x 6″ x 0.25″ TeflonTM slab. The clips held a 3M Static Event DetectorTM (SED), a resistor (R) and a shunt as diagrammed in Figure 1.
For shielding from electrical fields and radiation from the discharge, the slab was mounted on a 1″-high Teflon pedestal inside a 10″ x 15″ x 3″ cardboard box externally covered with 0.001″ aluminum foil. In some tests, a clip (not shown in the figure) connected the shunt material to the foil on the box. This setup represented a circuit board with shunted edge contacts or connector pins.
One SED was tripped by surges of approximately 95 V and represented 100-V-sensitive devices such as MOSFETs.2 Another SED was tripped by surges of approximately 15 V and represented very highly ESD-sensitive devices.
The SED was protected from overvoltage by connecting an NE-2 neon bulb from the lid to the body. The bulb would then flash when the detector was tripped. An SED had an advantage over an actual device because the SED registered the moment of failure, whereas the device could fail at any time, even during removal from the test apparatus.
The discharge was from a hemispherically tipped 0.25″-dia steel probe on a 200-pF ceramic capacitor charged to 0.5, 1, 6 or 12 kV. This probe represented a blunt hand-held tool. In fact, a screwdriver shaft held by an insulated, charged person gave similar test results.
For a discharge to the circuit, rather than to the shunt material, the probe was touched to one of the clips (on R side of circuit) holding the shunt.
The Results
The shunt materials were copper wire or 0.5″ x 2″ plastic strips held by alligator clips 1″ apart (Table 1). The resistance was measured, and all testing was done at approximately 50% relative humidity.
In Tables 2 and 3, an unqualified protection rating required that the SED not be tripped when R was either a copper wire or an open gap. When the rating was qualified by a resistance range for R, the SED was tripped at the upper level but not at the lower. As a result, shunting became somewhat less effective as R became larger and the circuit became more unbalanced.
Successful shunting should control a 6-kV discharge, which easily occurs from ungrounded personnel at 20% to 30% relative humidity. For discharges to the circuit rather than to the shunt itself, the 6-kV criteria was not met for 15-V surges with any shunt (Table 2). This result suggested that highly ESD-sensitive devices required design protection. The criteria was met for 95-V surges when the copper shunt was connected to the foil-covered box (Table 3).3 Results were the same with the box lid open or closed and the foil grounded or ungrounded. As a result, an incomplete and ungrounded Faraday cage connected to the shunt was sufficient.
Shunting the 95-V SED from the lid to the body with a large metal clip gave 12-kV protection. Similarly, grounding both of the alligator clips holding the SED prevented 95-V surges at 12 kV with the copper shunt and no box.
Generally, high-voltage, short rise-time discharges to the circuitry created momentary currents even with the circuit closed (copper shunt). The only way to prevent these currents with associated voltage surges was to short the circuitry near the device.
High-voltage discharges to the shunt itself, on the other hand, could be tamed by using a conductive shunt connected to a (partial) Faraday cage shielding the circuitry from spark-discharge radiation, or using a shunt of antistatic plastic (surface resistivity at least 109 W /sq but less than 1012 W /sq) that prevents a spark.
A 12-kV discharge sparked to the carbon-loaded polyethylene (PE) but not to the static-dissipative vinyl or antistatic PE. Carbon-loaded PE is often used for shunting; for example, in the form of strips slipped over edge contacts of boards. This material, however, gave disappointing results shown in Table 3, and was not effective unless covered by antistatic plastic.1,4,5 The limited conductivity of carbon-loaded PE promoted spark discharges that could not be dissipated rapidly.
Recommendations for shunting are given in Table 4. A conductive shunt covered with antistatic plastic can serve all purposes, although an antistatic shunt will suffice for Faraday-caged boards with only the shunt exposed to discharges. A bare metallic shunt is less desirable because it requires a low-resistance connection to the case that might be degraded by soiling or corrosion (Table 3).
When unshielded boards cannot be grounded at various points, protection from discharges to the circuitry is limited and testing is in order (Table 4). Hardening of the mounted devices, which is always recommended, is especially important in this instance.3
If connector pins on a Faraday cage are to be merely covered, not shunted, any solid material in Table 1 might suffice for a cap with an air space between it and the pins. However, SED tests would be prudent for critical items.
Conclusion
Shunting of circuitry, as opposed to shorting device leads, provides limited ESD protection. Many factors are involved, such as unbalanced resistance and capacitance in the circuit. Actual circuits should be evaluated using an SED tripped at the human body model sensitivity level of the most sensitive device.
References
1. MIL-HDBK-263B, paragraph 40.1.13.
2. Kolyer, J.M., “Testing Surfaces for ESD Safety,” EE-Evaluation Engineering, October 1994, pp. S-36-S-40.
3. MIL-STD-1686B, paragraph 5.3.
4. NAVSEA SE 003-AA-TRN-010 Electrostatic Discharge Training Manual, p. 130.
5. Kolyer, J. M., and Watson, D.E., ESD from A to Z: Electrostatic Discharge Control for Electronics, Van Nostrand Reinhold, 1990, p. 22.
About the Author
John M. Kolyer joined Rockwell International in 1973 and today is a Senior Engineering Specialist in ESD control and nonmetallic materials sciences. He received a Ph.D. in chemistry from the University of Pennsylvania in 1960 and has authored many papers and articles and a book on ESD control. Rockwell International, Mail Code 031-GE22, 3370 Miraloma Ave., Anaheim, CA 92803, (714) 762-6144.
Table 1
Material
Thickness
(in inches)
Resistance
(in W )
Copper Wire
0.041
<0.1 @ 1 V
Carbon-Loaded PE
0.050
1.3k @ 1 V
Static-Dissipative Vinyl Sheet
0.085
8 x 108 @
100 V
Antistatic PE Sheet
0.036
4 x 1011 @ 100 V
Antistatic Foam
0.25
5 x 1011 @ 100 V
Table 2
Shunt
Protection Level
(in kV)
Discharge to Circuit
Discharge to Shunt
Copper Wire or Carbon-Loaded PE
< 0.5
< 0.5
Static-Dissipative Vinyl
< 0.5
1
Antistatic PE or Foam
< 0.5
6
Table 3
PROTECTION LEVEL, kV
SHUNT
SHUNT CONNECTED TO BOX
SHUNT ISOLATED FROM BOX
DISCHARGE TO CIRCUIT
DISCHARGE TO SHUNT
DISCHARGE TO CIRCUIT
DISCHARGE TO SHUNT
Copper wire
12*
12*
0.5 (R = 1K-10K)
1 (R = 0-1K)
0.5 (R = 1K-10K)
1 (R = 0-1K)
Carbon-loaded PE
0.5
1
1
0.5 (R = 100K-1M)
1 (R = 1K-10K)
Stat.-diss. vinyl or antistatic PE or foam
< 0.5
12
< 0.5
12
Copper wire covered by two layers of antistatic PE
12*
12
0.5 (R = 1K-10K)
1 (R = 0-1K)
12
Table 4
|
Condition |
Shunting Method |
|
Handling of boards (modules) during mounting of components (circuitry exposed). |
Conductive shunt covered with antistatic plastic. Circuitry grounded at various points. Confirmatory SED test recommended. |
|
Boards assembled in a metallic case (circuitry Faraday-caged, shunt exposed). |
Antistatic plastic shunt or conductive shunt covered with antistatic plastic. |
|
Boards assembled with no case or a nonconductive plastic case (circuitry not Faraday-caged or groundable at various points). |
Conductive shunt covered with antistatic plastic. SED tests must establish discharge protection level. |
Copyright 1996 Nelson Publishing Inc.
May 1996
|
|