Transcript Sean Halloran JSHS Posterx - Manchester Essex Regional High

Electrostatic Discharge Protection for Analog Digital I/O Boards
Sean Halloran
Authentic Science Research Program
Manchester-Essex Regional High School, Manchester-by-the-Sea, MA 01944
Abstract
Electrostatic Discharge Protection for Analog Digital I/O Boards
Sean Halloran
Manchester Essex Regional High School, Manchester-by-the-Sea, MA
Teacher, Dr. Maria Lonnett Burgess Manchester Essex Regional High School
Mentor, Eric Wilson, Engineer, Varian Semiconductor Equipment Associates
The precise control of machines responsible for manufacturing requires several
interconnected, embedded electrical systems, or circuit boards, that take input in the form
of analog and digital signals and then output digital and analog signals to control
subsequent components. We added electrostatic discharge, or ESD, protection to output
and input lines of an analog digital I/O board with the goal of preventing damage to the
board and allowing it uninterrupted functioning. We hypothesized that this configuration
would discharge the ESD to ground, allowing board function to continue. To test this, a
variety of digital and analog outputs were subjected to the discharge of an ESD gun for
durations from 3 sec to 1 min between 5-20kV. Results show that none of the components
of the board sustained damage that could not be resolved by power cycling. Applying ESD
to an analog output channel resulted in that channel being held low, and at higher voltages
the other functioning output channels of the DAC chip would be held low more frequently.
We conclude that this combination of components proved effective at discharging ESD
away from sensitive components of the board at voltages as high as 30kV for as long as
a minute.
Methods
Results
On each I/O line of the board, there was a 0.127mm spark gap from the input or output line to
chassis ground, a ferrite bead, a 47Ω carbon resistor, a 33µH inductor, and either a 12V or
5V zener diode leading to chassis ground, all as seen in Fig. 1.
An eclectically isolated microcontroller cycled the outputs of the board. An oscilloscope was
used to measure the outputs, and if the waveform before the ESD was administered
matched the waveform afterwards, the board survived.
This chart shows the effect of electrostatic discharge, ESD, of various high voltages for
different amounts of time on a sampling of input and output lines of an analog digital I/O
board. Observations of the function of outputs after the application of ESD, or the
“zapping,” are recorded.
The setup for testing the ESD protection was as illustrated below.
Fig. 3. The setup for testing ESD protection. As seen above, ESD gun was attached to be in
direct contact with a wire leading to a given input channel on the board.
Fig 1. Schematic for the ESD ion
protection on one input line of the
analog digital I/O board
Introduction
Precise electronics are becoming increasingly integral to modern manufacturing processes,
like the manufacturing of semiconductors. In many such processes, the electronics in place
are vulnerable to high voltage in the form of electrostatic discharge, or ESD. Whether this
from static shock or a naturally-occurring lightning bolt, ESD can destroy the sensitive
electronics responsible for manufacturing expensive items, and therefore must have
effective ESD protection. We sought to evaluate the effectiveness of one such design.
In our experiment we added ESD protection to the output and input lines of an analog digital
I/O board with the goal of preventing damage to the board and allowing it to continue
functioning uninterrupted. Our ESD protection starts with a spark gap where high voltage
should be forced to ground by the resistance of the other components. Then a ferrite bead,
47ohm carbon resistor, and 33µH inductor should provide more resistance to high voltage
forcing the current across the spark gap. After these components, a Transient Voltage
Suppression (TVS) Zener diode attached from the channel to chassis ground allows for the
high voltages of the ESD to flow straight to ground. We hypothesize that this combination
of components will discharge the ESD to ground resulting in uninterrupted board function.
Fig. 2. Example of naturallyoccurring Electrostatic Discharge
via lightening, ESD
RESEARCH POSTER PRESENTATION DESIGN © 2011
www.PosterPresentations.com
Zap Target
Zap Voltage (kv)
Duration of Zap
(sec)
AOUT0
5
3
AOUT0
5
3
AOUT0
5
5
AOUT0
5
8
AOUT2
5
8
DOUT0
5
6
DOUT0
5
10
DOUT6
5
10
DOUT6
15
10
DOUT0
15
10
DOUT6
30
10
DOUT6
30
60
DOUT6
30
60
DOUT6
30
60
AOUT0
15
10
AOUT2
15
10
AOUT2
30
10
AOUT0
30
10
AOUT0
30
60
AOUT2
30
60
Affect on Digital Outputs
Affect on Analog Outputs
Recovery
DOUT LED’s flickered upon zap, but
continued to cycle
DOUT LED’s flickered upon zap, but
continued to cycle
DOUT LED’s flickered upon zap, but
continued to cycle
DOUT LED’s flickered upon zap, but
continued to cycle
DOUT LED’s flickered upon zap, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
AOUT0 died
AOUT2 survived
AOUT0 died
AOUT2 survived
AOUT0 died
AOUT2 died
AOUT0 died
AOUT2 survived
AOUT0 died
AOUT2 died
None
Upon board reset
None
No recovery needed
None
No recovery needed
None
No recovery needed
None
No recovery needed
None
No recovery needed
AOUT0 survived
AOUT2 died
AOUT0 held at 4v
AOUT2 died
Recovery upon board reset
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
Bright flickering on DOUT LED’s, but
continued to cycle
AOUT0 died
AOUT2 died
AOUT0 died
AOUT2 survived
AOUT0 survived
AOUT2 died
AOUT0 died
AOUT2 died
AOUT0 died
AOUT2 died
AOUT0 died
AOUT2 died
AOUT0 died
AOUT2 died
Recovered upon board reset
Upon board reset
Upon board reset
Upon board reset
Upon board reset
No recovery needed
AOUT0 and AOUT2 held high upon
board reset, recovered after power
cycling
Recovered upon board reset
Recovered upon board reset
Recovered upon board reset
Recovered upon board reset
Recovered upon board reset
Recovered upon board reset
Conclusion
Fig. 4. The setup for adjustment of the arcing distance of the ESD gun. As seen above, this
wire was cut and attached to the jaw of a pair of venire calipers. By adjusting the calipers,
the arcing distance of the ESD gun could be tuned. For each test, this distance was tuned
to be just small enough to allow the current to arc, but large enough so that only high
voltage would arc.
The combination of components used for ESD protection proved effective at discharging
ESD away from sensitive components of the board at voltages as high as 30kV for up to
1 min. The repeated success of these tests suggests that the ESD protection can protect
against ESD as high as 30kV indefinably, or at least for longer than any natural world ESD
would last. Problems with the failure of analog digital to analog chips can be solved by
programming the software to reset the chips during such failure. Further investigation to
determine changes in the waveform output by the digital outputs may be needed.
Fluctuation of the waveform ~5V could be solved with components designed to clamp the
output at 5V, like a zener diode. The success of this experiment may lead to further
experiments to test ESD protection with fewer components.
References and Acknowledgements
The procedure for testing the ESD protection was as follows:
•
•
•
•
•
•
•
Manipulate the ESD voltage and i/o channel targeted
Measure the waveform from the new i/o channel
Fire the ESD gun and close the gap on the calipers until arcing occurs
Continue to fire the ESD gun for a variable amount of time
Measure the waveform after ESD application
If necessary, attempt to recover the board and remeasure the waveform
Repeat
Fig. 5. Entrance to
Varian Semiconductor
Equipment facility, where
I did my research.
1. Littelfuse, Inc. Electronic Product Selection Guide, Chicago, IL, 2011.
2. Wilson E. Personal communication
I thank my fantastic mentor Eric Wilson for giving me the opportunity to intern Varian and
do such interesting research. I thank everyone at Varian for being so helpful and willing
to further my research experience.
Contact [email protected] for further information on this research.