Presenter:
Douglas C. Smith
BE SURE TO CHECK OUT THE PICTURES
AT THE END....
What we saw:
On Feb 26th your local EMC Society was part
of the IEEE Denver
Expo 2000 - a Professional Activities Conference
in the Denver Tech Center. The speaker was Doug
Smith and the topic he covered was, as the title
described, an indepth study of signal and noise
measuring techniques for characterizing and
troubleshooting electronic circuits and designs
in the frequency range of 20MHz to 1GHZ and
higher. As a bonus he also covered techniques for
finding and fixing signal integrity,noise and EMC
problems.
The RMCEMCS would like to thank
Auspex Systems for allowing Doug to speak at our
show.
Doug started the talk with an
interesting demonstration. Using a home-made
50Mhz oscillator driving a "Black Box"
Doug attached a clibrated scope probe to each of
the the two outputs of the box. He then
demonstrated that the output of the box were not
idential. Not terribly strange - until he open
the box to reveal that the two outputs were
actually attached together. The question he
started the day was - Why?.
The talk scope started with an
overview of basic theory on inductance, magnetic
coupling, shielded cables and skin effect with an
emphasis on measurement tools such as scope
probes and differential measurement techniques.
Doug then proceeded to show some examples of
clock waveforms that were less than perfect and
proved conclusively that there is a direct
relationship between the cost of the measurment
equipment and confidence in the result. We then
covered a section on scope probe characteristics
of various probe designs including Dougs
favourite - the balanced coaxial probe. Using
simple math and graphical displays, Doug showed
how scope probes - even the expensive low
capacitance ones - can dramatically change in
characteristic as the frequency increases.
In the next section Doug covered
induced voltages and its effects on measurements.
In this section we reviewed basic theory on
inductance, mutual inductance and shielded cable
operation. Doug walked us through some simple
concepts on mutual inductance and introduced the
ides that a coaxial cable is really a 1:1
transformer between the shield and the center
conductor. He also demonstrated that by using
inductance and mutual inductance as constants one
could design a probe that will give a qualitative
analysis of the noise in a circuit in question.
In this case the probe was square - a bent paper
clip and a piece of 50ohm coaxial cable. Doug
also presented another method of visualizing a
shielded cable, such as a coax cable in that a
shielded cable often works by inducing noise on
the center conductor rather than keeping noise
out. Doug then introduced us to the
"NULL" experiment which amounts to
validating the measurement.
The third section covered
differential measurements. Doug first covered all
the techniques that are used and reviewed some of
the more common mistakes that can occur in such
measurements. After looking at all other methods,
Doug then went into some detail on the Balanced
Coaxial Probe. He showed, using some ESD induced
waveforms, how the balanced coaxial probe was
more immune to induced noise as it has a high
balanced common-mode rejection ratio.
The final section dealt with
non-contact measurements. In this section Doug
covered exactly how the square loop worked and
reviewed some of the more interesting aspects of
the more ubiquitous round loops that we are all
familiar with.
During the talk Doug punctuated
his points with demonstrations and experiments.
Overall the presentation was well received and
appreciated by all. Thank you Doug.
For more information on Doug
please goto his
website.
Some pictures from the
presentation (Note: Click on the picture for a
larger image)
1. Doug demonstrates a nicely
distorted waveform. Is the measurement real
or...??
2.Doug demonstrating the
measurment of current paths on a printed wiring
board.
3. Doug and Lyle share some
amusing insights into printed wiring board EMC...
4. Doug with the officers of the
RMCEMCS at the IEEE event sponsored luncheon.
|
