8/4/2019 EEWeb Pulse - Issue 9, 2011
1/22
EEWeb.c
Issu
August 30, 20
EEWeb
PULSE
John LaddRoman Systems
Engineering
Electrical Engineering Commun
8/4/2019 EEWeb Pulse - Issue 9, 2011
2/22
Contact Us For Advertising Opportunities
www.eeweb.com/advertising
Electrical Engineering CommunityEEWeb
Digi-Key is an authorized distributor for all supplier partners. New products added daily. 2011 Digi-Key Corporation, 701 Brooks Ave. South, Thief River Falls, MN 56701, USA
www.digikey.com/techxchange
Its all about
connections.
The user-to-user forum is for everyone, fromdesign engineers to hobbyists, to discusstechnology, products, designs and more.
Join the discussions that match your interestor offer your expertise to others.
Join the discussion now at:
discussions
hobbyists
engineers
industry experts
resourceslinks
technical documentswhite papers
reference designs
application notes
community
power
microcontroller
lighting
wireless
sensor
students
http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm7http://bit.ly/kDYHm78/4/2019 EEWeb Pulse - Issue 9, 2011
3/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 3
TABLE OF C ONTENTS
John Ladd 4CEO, Futurist, and Inventorof RSE Technology
Controlling Latches Before TheyRuin Your DayBY RAY SALEMI
Selecting Precision Op Amps for 16Sensor-Input Processing DesignsBY TAMARA SCHMITZ WITH INTERSIL
RTZ - Return to Zero Comic 20
Salemi examines and discusses a possible side effect of combinatorial
proceduresunintended latches.
Interview with John Ladd - Co-Developer of Roman Systems Engineering
How to select the best precision operational amplifier for implementing topquality sensor-input processing designs.
12
8/4/2019 EEWeb Pulse - Issue 9, 2011
4/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 4
INTERVIEW
Roman Systems EngineeringSum up Roman SystemsEngineering in one sentence.
RSE solves some of the problems
facing todays 3d laser scanner
systems by utilizing ancient Roman
technologies in conjunction withmodern dielectric fluids and
novel hybrid binary tree branch
computational solvers.
What is your valueproposition?
Modern 3d laser scanners which
utilize time-of-flight approaches have
difficulty imaging transparent or
highly reflective samples. Because
RSE liquid scan technology does
not rely on radiation, it can overcome
these limitations. In addition, RSE
scanning allows penetration into
the cavities of a porous medium.
Imagine being able to scan a
crouton that has a hollow cavity, or
a rigid but porous food that containsa jelly filling. There are of course
many interesting applications.
Can you tell us about theearly start-up days at RomanSystems Engineering?
We still consider ourselves to be
in the early start-up mode until
mass production is achieved. Our
defining moment occurred on
March 9th, 2011, when Peng Tian,
Guanbo Chen, and I decided to
run a MATLAB simulation to test
my theory about the purpose of the
Roman Dodecahedron. We were
busy in our graduate microwaves
course with 50-page lab write-ups,
but the idea that the Romans usedthe dodecahedron as a liquid-
displacement based 3d recording
device was inescapable. I dropped
out of my plasmas course to
pursue this theory and related
experimentation as a full time
endeavor. I began writing patents,
copyrighting our material, and
pursuing trademark protection. We
were lucky to have the mentoring
and advices from a top U.S. patent
holder, Salman Akram, and some valuable advices from technology
leaders such as David Orton (CEO
of Aptina), Gennady Agranov (V.P. of
imaging technology), and especially
theoretical physicist Dr. Sergey
Prokushkin. The inspiration to do
something beneficial for the U.S.
economy came from an inspiration
we (co-founder Megan Albrightson
and I)had felt after taking Rick G.
Branners electromagnetics and RF
courses at U.C. Davis.
By April 25th, we were already
ready to present our technology in
a public forum and had achieved
hundreds of provisionally patented
ideas surrounding our core
technology. It was a one-month
patenting binge and the most
productive period of my life. To date,
John Ladd
RSE Co-Developers (left to right): Larissa Prokushkin, Sergey Prokushkin,John Ladd, Megan Albrightson, Nail Khaliullin
8/4/2019 EEWeb Pulse - Issue 9, 2011
5/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 5
INTERVIEW
our theory about the Purpose of the
Roman Dodecahedron has been
unchallenged. It was also thoroughly
reviewed and then defended for
two hours in front of technology
leaders in San Jose on June 29th. We have also been interviewed by
the Fox News reporter who ran the
story that the Mystery of the Roman
Dodecahedron may never be
solved, and we think we answered
all of her questions to her liking.
Please explain what theRoman Dodecahedron is andwhy you believe you have
solved this mystery.The Roman Dodecahedron is an
ancient bronze (or sometimes
stone) artifact that has turned up
by the hundreds and looks like a
highly optimized characterization
device. Not all dodecahedrons or
icosahedrons appear the same at
first glance, but there are common
traits that give evidence to its
purpose. Dozens of theories have
been developed, all of which have
been discounted. In fact, doctoraltheses have been written attempting
to discover the intended use. We
are confident that the purpose
has finally been revealed and that
I discovered it in a microwave
engineering course while in
graduate school in Ann Arbor,
Michigan. We think that the Roman
Dodecahedron was used to record
the three dimensional shape of an
object under study by measuring
the fluid displacement of the object
under various angles and depth of
immersion.
We dont buy it. Give us moredetails.
Not all dodecahedrons or icosa-
hedrons appear the same at first
glance, but there are common traits
that give a clear indication to its
probable use. There is evidence
(e.g., scuff marks) that objects
were placed in it, but we found that
in order to wedge a device holderbetween the supporting vertices, it
required a flexible object holder in
order to achieve high repeatability
in the device placement. We sus-
pect that the Romans used wood,
and similar Egyptian engineering
regarding the construction of pyra-
mids, the tools rotted and left his-
torians scratching their heads due
to a lack of empirical evidence. In
order to fully appreciate our theory,
you have to think about how the
Romans designed the device from
scratch. You must start asking fun-
damental questions that lead you to
the design of the Roman Dodeca-
hedron by considering all aspects
of a 3d recording device. In fact,
we were unaware of the existence
of the Roman Dodecahedron and
asked how to solve a practical en-
gineering problem. It was after our
design was finished and we wereperforming some volume calcula-
tions did we notice the artifact on
Wikipedia and was quite shocked
to see its purpose was unknown.
We immediately began compiling a
list of traits that Roman Dodecahe-
drons must exhibit in order for our
theory to be valid. We saw that all
the dodecahedrons found so far did
exhibit those traits and reminded
us so much of the debates we had
when designing our own device. Itlooks like the Romans had the same
debates about the optimal design
of the dodecahedron that we had,
which was pretty exciting for us.
Please explain whatfundamental questions ledyou to the design of theRoman Dodecahedron.
At the beginning of our design
project, we asked the question
about how we could measure the
spatial dipole impulse response
at an arbitrary location around a
two terminal conductor of arbitrary
shape (analogous to a Green
function in electrostatics). We
began with ideas for a rotating
sphere (i.e., trackball design)
that could have dielectric fluids
injected into the hamster ball at
various levels and rotate at variousmeasurement angles. It was an idea
for a variational technique that we
believed would lead to information
to determine the lower and upper
bound for the effect of a dielectric
raindrop on the total capacitance
seen at the two terminals after the
dielectric raindrop was brought into
proximity of the electric fields. We
were to develop a measurement
system and algorithms to study this
idea, determine how useful it wouldbe for mapping tensor behavior,
and we were ultimately to place a
dielectric drop on a Styrofoam rod
to test our experimental or analytical
algorithms we would uncover. It
was a very aggressive project for
a three-credit course where we
were obviously extremely busy
with the laboratory write-ups and
teaching and research assistant
commitments.
To implement a practical prototype
in a realistic timeframe, Guanbo
was adamant that we use an open
immersion system so that fluid
leaks would not be an issue, though
we heavily debated the choice of
whether to use a closed or an open
8/4/2019 EEWeb Pulse - Issue 9, 2011
6/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 6
INTERVIEW
system. I looked to platonic devices
that would have enough angles of
measurement for reasonable spatial
accuracy based on the theory of
binary integer solving theory. I
found that to obtain a reasonableresolution we would need between
12 and 24 faces on our immersion
cage to be able to perform liquid
scans and continuously measure
the capacitance of the enclosed
two-terminal device with a practical
accuracy (enough to study 3d
flowering algorithms that would
allow a method for interpolation to
map out the spatial dipole response
everywhere in the vicinity of the
terminals). When a dodecahedron
is inscribed in a sphere, it has a
greater volume than an icosahedron,
which for fluid based imaging, is
great. You dont want the device
being tested to look like a gnat
relative to the size of the immersion
bowl. The dodecahedron also has
more vertices than the icosahedron
and a larger entrance hole on the
face for placing the device to be
tested. It definitely wins on mostlevels compared to its icosahedron
cousin. Even where an icosahedron
holds an advantage (the number
of fundamental angles that can be
scanned) the dodecahedron can
hold its own by having the large
feet buttressed on supporting
legs to gain additional angles of
measurement (when a pentagon is
supported in this fashion it rests like a
wheel-barrow). The dodecahedron
is very able to be manufacturedcompared to the icosahedron and
it only required two hours with a
hack saw to put together quite an
accurate device (both halves of the
cage came together perfectly on
the first shot). What is interesting is
that there are signs that the Romans
also had this debate on whether
the bowl, the fluid level would rise
in accordance with Archimedes
principle. The degree to which
the fluid level rose would depend
upon the displacement slice of
the object being measured, and thisfluid level would be recorded into
the 12 columns of displacement
data representing the 12 angles
under study. By performing this
measurement on all faces, a
reasonable rendering of a smooth
shape could be performed, if they
had only had a computer at their
disposal. Of course we are not
suggesting they had a computer
or even rendered. What matters
is that the data set is unique to the
device being tested and can be
compared with other devices for
a type of one dimensional quality
control. By maintaining a high
quality control of perhaps projectile
manufacturing, a greater kill ratio of
the Roman army could be achieved.
Laser scanners today are used for
quality control, and the Roman
Dodecahedron could have easily
been used at that time with only 18minutes of recording effort for the
device under study. It was practical
and it was useful. It would even be
useful if placed side by side with the
micrometer in every hardware store
on the planet.
Are you suggesting that the 12inches-in-a-foot also stemmedfrom the repeated use of thisdevice?
Yes, but it is more difficult to prove.
At the very least, it should become
a leading hypothesis because it
is based upon a fundamental and
consistent principle. If you are
recording 12 columns of data every
day in a table that is approximately
one foot wide (they probably didnt
use an 8.5 inch wide paper with fine
to use the dodecahedron or the
icosahedron since they did find a
single icosahedron prototype. We
added solder balls on the outside
of the cage as feet and ground
them down on my basement floor toachieve a very level device.
RSE solves some ofthe problems facing
todays 3d laserscanner systems by
utilizing ancientRoman technologiesin conjunction withmodern dielectricfluids and novel
hybrid binary treebranch computational
solvers.
I am missing something.What exactly did theRomans do with the RomanDodecahedron, in the simplestof terms?
They attached the object to be
tested (e.g., a projectile) into
the dodecahedron by means of
a flexible support (most likely wood or wax) that was wedged
between two interior vertices (or
corner reflectors). They put the
dodecahedron into a bowl and
repeatedly added fixed amounts
of water to the bowl, probably
using something equivalent to an
Erlenmeyer flask for high accuracy.
Each time water was poured into
8/4/2019 EEWeb Pulse - Issue 9, 2011
7/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 7
INTERVIEW
point writing utensils) recording
scroll, and that the data represented
your three dimensional ruler, but
you were viewing the data and
comparing it in one dimension,
wouldnt you also adopt the samestandard for a 1d ruler? It definitely
is a possibility if they preferred a
consistent unit of measurement
between shapes and distances.
I dont see any other reason they
would adopt a 12-inches-in-a-foot
standard. I think it was based on
the fifth element of the Zodiac.
When the Normans arrived, they
brought back the tradition of the
Roman 12-inches in a foot. Although
no single document on the subject
can be found, it appears that during
the Reign of Henry I (1100 1135)
the 12-inch foot became official.
What is your strongest shortsummary that supports thispossible historic nding? Ifyou are right, the world needsto know about it. What wouldyou say to a skeptic?
I would ask skeptics whether they would think it would be useful or
not for the Romans to have the
ability to record the 3d shape of
projectiles or other devices? If so,
I would like to ask them to propose
a better way than fluid immersion
and a more optimized structure
than the Roman Dodecahedron
to perform this important task.
If they know a better way, then
I would agree that skepticism
makes sense. If not, I would ask
them to look at the commonalities
behind all the dodecahedrons,
which include large feet for angle
adjustment, a large hole on at
least one side for DUT placement,
and the hole patterns such that all
dodecahedrons (or icosahedron)
allow for horizontal fluid level
settling around the device that is at
the center. If the Romans didnt use
this device for 3d recording, they
were missing out on an ideal and
highly optimal and practical use for
this structure. In general, it is very
rare for a highly optimized piece of
characterization equipment to notbe used for an ideal practical use for
its design. We think that our theory
is the first which gives an optimized
solution to a practical problem that
the Romans would be facing. The
other theories, simply put, dont
make any sense. This device was
not an optimized paperweight or
a candlestick holder. They are
not going to be measuring the
diameter of pipes horizontally (it
is not ergonomic and hard on the
shoulders to do this repeatedly) and
would simply place a square on the
ground with a crescendo of holes
to plant the pipe in. In addition, the
icosahedron that was found was
clearly not used for this purpose
(with the many equally sized small
holes) so that theory should be
discarded.
The use of this device in quality
control applications did not require
the dodecahedron to even be
placed perfectly flat relative to
gravity, or be manufactured with
perfect precision. It only had tocompare two devices for likeness,
so it was clearly well engineered
for this purpose. Even the feet
themselves give indication to the
need for repeatability. A close
inspection shows that they often took
care to not only make the feet large
but make the area of the feet that
touches the bowl somewhat small,
guaranteeing a higher repeatability
because of a lowered chance of
dirt to be trapped in between thecontact point of the structure and
the measurement tank.
We think you are prettyadamant about yourdodecahedron theory.Lets talk about the future of
3D Scanner Dodecahedron Prototype by RSE
8/4/2019 EEWeb Pulse - Issue 9, 2011
8/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 8
INTERVIEW
engineering. What can you dowith this device?
We extend the accuracy and speed
of this device by adding capacitive
fluid level probes within the do-
decahedron structure. By snapping
two of these dodecahedrons to-
gether in hourglass form, and seal-
ing the dodecahedron, accuracy is
unparalleled because the top do-
decahedron knows very accurately
how much fluid was dispensed into
the bottom and the bottom structure
obviously knows the displacement
through the capacitance level read-
ing. By using modern fluids, such
as fluorinert, we can even penetrateinto some test objects to map out
the interior of a cavity. Fluorinert
has extremely weak intermolecular
forces and is almost twice as heavy
as water. Its ability to penetrate a
medium and then quickly drain out
(without hysteresis) is demonstrated
in our videos. That is the big advan-
tage here over photons, the persis-
tence of molecules to penetrate into
a medium, and their determination
to achieve a state of minimum po-tential energy. The software pack-
age delivers a useful rendering of
the subject (in voxel space) on the
home computer. Additional angles
of measurement are performed in
real-time to isolate areas of interest
where higher resolution is required
(by utilizing the statistical engine),
and to avoid air-bubble traps that
are flagged by the time displace-
ment information that is captured.
Additionally, the fluid level opera-tion serves as a slant-edge line-
of-sight calibration procedure that
maps out the imaging zones for
traditional color image sensors
that are embedded in the dodeca-
hedron vertices. This approach
allows a full color surround image
of the DUTs surface to be overlaid
onto the geometrically accurate
liquid scan volume image, morph-
ing the RGB surface information to
coincide with the geometrical liquid
scan data. We anticipate that users
will be stunned to be able to scanunique objects (such as Twinkies,
for example) that contain filling and
render density scan images along
with the color surface on websites.
We recognize that the fundamental
technologies that have proven
themselves throughout time are
ideal building blocks for the device
of the future. People have proven
that they have always been willing
to carry around water bottles orflasks, and sometimes on the hip.
Tape measures are worn on the
hip. And this slick device will be
worn on the hip, and since it is a
fluid containment system, it will
naturally be able to hold hard
alcohol, which is actually a decent
lower-end substitute for fluorinert
to do some basic 3d rendering
hard alcohol, depending upon
type, has low surface tension and
reasonable levels of measurement
hysteresis. Fluorinert is a very
inert and non-toxic substance and
the risk of cross-contamination
being an issue to human health
is nonexistent. In the future we
expect that the dodecahedrons that
are joined in hourglass function to
perform scanning functions will
have a social proximity based tie-in
that will allow for meaningful face-
to-face interaction between people.This might be an acceptable
replacement for the problem
presented by modern social media
which leave people feeling isolated
and lonely hitting refresh on their
computers on a Saturday night.
By having the two dodecahedrons
talk to each other, with a natural
hand-shake operation, a certain
level of privacy is achieved while
allowing both devices to screen
for commonalities that would
serve as excellent ice-breakers. As
engineers, we recognize that face-
to-face interaction and problemsolving is a key ingredient in
accelerating our economies forward.
Not much good development
happens without a whiteboard, no
matter how advanced our virtual
conference rooms become. I fear
that I am getting too far ahead into
the future market for our device and
have strayed. My apologies.
How long before we see these3d Roman Ruler products onthe store shelves?
Oh the weather outside is frightful
but the fire is so delightful. And if
there is no place to go, let it snow,
let it snow, let it snow! I think Santa
will lend EEWeb one of the first to
stuff their stockings with. Thank you
for your time.
Wait. If the Romans used the
Dodecahedron and measuredthe water level, wouldnt itbe difcult to measure a verysmall displacement causedby say a small defect in anarrow head?
Exactly, and surprisingly, this adds a
great deal of credibility to the theory.
We are working on an app that allows
volume intercept injection of a table
of Roman style hand measurements
into both a 3d rendering programand a simple quality assurance
program that would be more in
keeping with the Roman technology.
We plan to issue a significant cash
prize for both the best rendering
of a projectile and the best quality
control measurements (lowest
standard deviation error between
two projectile standards that can
8/4/2019 EEWeb Pulse - Issue 9, 2011
9/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 9
INTERVIEW
be resolved). Anybody is welcome
to beat us to the app and it should
be an interesting competition. We
can only answer your question at
the time with two references for an
upper and lower bound on what canbe achieved.
The lower bound would be
dropping a lead ball into a bowl
and measuring the fluid level
displacementthat would be a
single angle with a single slice,
and also be called an Archimedes
experiment. That is a lower bound
that anybody can achieve and is still
the most accurate way to measure
the volume of a solid object. Toperform fine resolution slicing and
more angles takes increasing skill,
both for the fluid level measurement
and the repeatability of the DUT
placement methodology. We have
played with the measurement and
found that we are still on the steep
learning curve where repetition will
only make us better.
We can get guidance about the
practical upper bound fromthe mathematics of 3d volume
displacement imaging. For
example, from six angles of
measurement, and 10 data slices
per angle, we can mathematically
resolve a simple object with only 0.7
percent voxel error on a 10x10x10
volume grid. With this as the upper
bound, it is unlikely that a small
point defect could be recorded
and rendered; it would have to be
a significant chip or deviation in
the shape of the arrowhead. If the
Romans intended to only perform a
comparison between two projectiles
for consistency, they need not
record in such a manner that would
even be able to be injected into a
modern computational engine and
render. It need only provide error
data that over many samples could
be indicative of the quality of a
manufacturing line. In other words,
they may only need to know which
slave to whip.
We think our historicalefforts will help
people understandand appreciatethe technical
sophistication of
people outside ourown century. Everytime I begin to havea slight doubt, I takeanother look at the
Roman Dodecahedronand recognize
the proficiency ofmanufacturing theseengineers possessed.
For the practical upper bound, one
can imagine some highly skilled
characterization engineer in a tower
performing weeks of measurements
without interruption. The engineer
may provide many angles of
measurement using supports under
the feet of the dodecahedron. His
experience with the device maycover a lifetime of trial and error
and meticulous experimentation.
He may look across the fluid level
like a sharpshooter aligns the open
sights of a gun, closing one eye for
added precision. Or he may place
a wooden stick into the bowl and
have a pigment or oil floating on the
surface of the water which soaks
into the wood and dries, recording
each increment. After removing the
stick, he may be able to directly
compare a column of data with
a stick measured from anotherprojectile. This is analogous to the
method of checking fluid level from
an automobile engine.
In summary, we think that this method
of 3d recording is challenging,
and there are significant sources
of error. It is clear that the Romans
understood the error and took many
efforts to reduce it. The choice of
the dodecahedron itself shows an
understanding of the challenge of3d volume displacement recording
and choosing a structure that has the
lowest practical error. They showed
an understanding that the mass of
the object under study must match
the interior volume of the cage and
the bowl as well as possible because
the holes for DUT placement were
large on at least one side of the
dodecahedron. The range of sizes
of dodecahedrons found (4cm to
11cm) exhibit an understanding
that the dodecahedron must closely
match the size of the object under
study for maximum practical
accuracy.
Did they nd any wateringbowls in the eld near thedodecahedrons?
This would be nice data to have
and is a profoundly important
question if we are to get muchtraction from the non-engineering
crowd. We would like a smoking
gun. However, it appears that
gathering this kind of data could
be a challenge for several reasons.
First, although dodecahedrons were
found near military sites, projectile
manufacture can also take place
8/4/2019 EEWeb Pulse - Issue 9, 2011
10/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 10
INTERVIEW
in civilian territory. The number
of samples found is relatively
low. In addition, dodecahedrons
were found in graves which led tospeculation that it was a religious
artifact. That is a hurdle. We are not
surprised that some would choose
to be buried with a device if it was
their lifelong craft. When you hold
the device, it does grow on you.
Imagine holding a baseball for most
of your life, but something that feels
even more ergonomic in the hand
because it has corners and will not
slip out. Engineers who worked on
the NASA missions were very proudof their slide-rules and wouldnt go
anywhere without them.
Another problem is the disparity
in cost between the bowl and
the dodecahedron. While the
dodecahedron was expensive, a
bowl simply needed to closely match
the dodecahedron in size. Because
of the difference in cost between the
two devices, you would naturally
expect that dodecahedrons wouldbe stored with some care, while
the bowl may not be close by. The
bowl obviously had secondary uses
as well. Our strategy for the search
for a smoking gun is actually to
start with electrical engineers who
are familiar with our particular
terminology. If the interest is there,
historians will take a closer look at
what we are doing and hopefully
provide some assistance. It may not
be easy. We dont know what materialthe bowl was composed of (it may
have shattered or rotted) so we
could be facing a lot of the problems
that surround theories of how the
pyramids were built, especially if
wooden tools surrounded the use of
the dodecahedron.
Ultimately, our efforts are aimed
at bringing a useful product into
production and creating some jobs
in the United States. However, tosolve a historical mystery would
be neat. We are confident we have
found the answer, but our reasoning
is fortified through engineering
principles, and not a smoking
gun. We think our historical
efforts will help people understand
and appreciate the technical
sophistication of people outside our
own century. Every time I begin to
have a slight doubt, I take another
look at the Roman Dodecahedronand recognize the proficiency of
manufacturing these engineers
possessed.
For more information please see the
Roman Systems Engineering Web-
site: www.romansystemsengineer-
ing.com.
http://bit.ly/jebqhLhttp://bit.ly/jebqhL8/4/2019 EEWeb Pulse - Issue 9, 2011
11/22
Avago Technologies new AEAT-6600
Hall Eect Magnetic Encoder delivers optimalsolutions for Robotic, Industrial and Medical
systems designers.
Worlds highest resolution
16-bit absolute positiion through SSI
Programmable Magnetic Rotary Encoder IC
16-pin TSSOP package
Power down mode
New Encoder for theWorst Case Environments
Avago Technologies Motion Control Products
For more information and to request
a free sample go to:
www.avagotech.com/motioncontrol
http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V5http://bit.ly/jpn9V58/4/2019 EEWeb Pulse - Issue 9, 2011
12/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 12
Before TheyRuin Your Day
Controlling
Latches
Ray SalemiVerification Consultant
In my last article (Creating Combinatorial Logic (Part
1)) we learned how to use procedural code to create
complex combinational logic. We saw that we could saveconsiderable space with an adder by creating our own
logic within a procedural process. In this article we are
going to examine a possible side effect of combinatorial
proceduresunintended latches.
Unintended latches are bad. In addition to making dogs
howl, causing children to cry, and curving your spine,
they will create simulation mismatches between your
RTL and gate level simulations, take up extra space in
your FPGA, and screw up your timing analysis. You really
dont want unintended latches in your design.
Synthesis tools create unintended latches when we forget
to handle all the conditions possible in our combinatorial
code. Lets look at an example of an unintended latch.
This design is supposed to take eight bits of input and
either increment or decrement it. We have an increment
signal to allow us to increment the data and a decrement
signal to decrement it. But weve forgotten the case
where neither increment, nor decrement is raised:
17
18
19
20
21
22
23
24
25
26
27
28
29
ARCHITECTURE rtl OF inc_dec_vhd IS
BEGIN
process (all)
beginif inc = 1 then
data_out
8/4/2019 EEWeb Pulse - Issue 9, 2011
13/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 13
TECHNICAL ARTICLE
21
22
23
24
25
26
27
28
29
3031
32
33
34
35
36
ARCHITECTURE rtl OF inc_dec_fixed_vhd IS
BEGIN
process (all)
begin
if inc = 1 then
data_out
8/4/2019 EEWeb Pulse - Issue 9, 2011
14/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 14
TECHNICAL ARTICLE
The solution is to explicitly add a reset to our design as
shown here:
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
ARCHITECTURE rtl OF alu IS
BEGIN
process (all)begin
if reset = 0then
result 0);
else
case op is
when 00 =>
result result result -- no change on 11
end case ;
end if;
end process;
END ARCHITECTURE rtl;
1
2
3
4
5
67
8
9
10
11
12
13
14
15
16
module alu (input [1:0] op,
input [7:0] A, B,
input reset,
output logic [7:0] result);
always_combif (!reset)
result = 0
else
case (op)
2b00: result = A + B;
2b01: result = A & B;
2b10: result = A B;
2b11: begin end// no change
endcase // case (op)
endmodule // alu
reset
op[10]
B(7:0)
A(7:0)
in out
ix29
result_lat Bus1(7:0)
LATRS_8_7_-1_set_1_reset
D
GQ
R
S
in
in(0)
in(1)
ou
out
t
ix
result_0n1s2_andBus3(7:0)
in[0]
in[1]out
result_0n1s3_xorBus4(7:0)
result_max_02Bus2(7:0)
a(7:0)
a(2)
LOR
LOR
b(1)
c(0)
cin
b(7:0)d(7:0)
dresult_add8_01
in(0)
in(1)out
ix41 19
Figure 4
Now weve added a reset signal to lines 24 and 25 and
this reset is reflected in the code. We have an explicit
asynchronous reset attached to our latch. Now we can
reset it when we start our simulation, and well get the
same results whether we simulate RTL or gates.
Summary
When we use procedural code to create combinatorial
logic, we need to be careful to define all the paths
through the logic. If we dont, we can unintentionally
create latches in our design. These latches can screw up
our simulation results, timing results, and area results.
Most synthesis tools will warn you if they are creating
latches. Be sure to take those warnings seriously and
either remove latches from your designs or give them
resets.
About the Author
Ray Salemi is a veteran of the EDA industry and has been
working with Hardware Description Languages since he
joined Gateway Design Automationthe company that
invented Verilog. Over the course of his career he has
worked at Cadence, Sun Microsystems, and Mentor
Graphics. Ray is currently an Applications Engineer
Consultant with Mentor Graphics.
8/4/2019 EEWeb Pulse - Issue 9, 2011
15/22
Wideband, Low-Power, Ultra-High Dynamic Range
Differential Amplifier
ISL55210The ISL55210 is a very wide band, Fully Differential Amplifier
(FDA) intended for high dynamic range ADC input interface
applications. This voltage feedback FDA design includes an
independent output common mode voltage control.
Intended for very high dynamic range ADC interface
applications, at the lowest quiescent power (115mW), the
ISL55210 offers a 4.0GHz Gain Bandwidth Product with a very
low input noise of 0.85nV/(Hz). In a balanced differential I/O
configuration, with 2VP-P output into a 200 load configured
for a gain of 15dB, the IM3 terms are
8/4/2019 EEWeb Pulse - Issue 9, 2011
16/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 16
forSensor-InputProcessingDesigns
Op Amps
Selecting Precision
Tamara SchmitzSenior Principal Applications Engineer
and Global Training Coordinator
A
s the basic building blocks used in an
extensive array of consumer, industrial,
scientific, and other applications, Operation Amplifiers (Op Amps) are among the most widely
used electronic devices, and for most low-end
applications, the requirements are straightforward
and the device choice is relatively easy. However,
there are challenges to selecting the optimal precision
op amps for implementing many higher-end sensor-
input processing designs.
The op amp selection can be especially challenging
when the types of sensors and/or the deployment
environments create special demands such as ultra
low-power, low-noise, zero-drift, rail-to-rail input andoutput, solid thermal stability, and the repeatability to
deliver consistent performance across thousands of
readings and/or in harsh operating conditions.
For precision op amps to be used in complex sensor-
based applications, designers need to look at multiple
aspects to get the best combination of specs and
performance, while balancing cost considerations as
well. In particular, chopper-stabilized op amps (Zero
Drift Amplifiers) offer excellent solutions for ultralow
offset voltage and zero drift over time and temperature.Chopper op amps achieve high DC precision through
a continuously running calibration mechanism that is
implemented on-chip.
Although there is no easy one-size-fits-all formula, the
following examples show how the op amp selection
can help achieve critical application objectives.
Weigh Scales & Pressure Sensors
Weigh scales and pressure sensing applications
typically use a highly sensitive analog front-end sensor,such as a strain gage, that can provide very accurate
measurements but output very tiny signals. For high-
precision weigh scale applications, designers may
use a bridge sensor network, in which individual op
amps are paired with gain resistors chosen to provide
common mode extraction and to deliver 10-20 PPM
of accuracy. Such advanced roll your own designs
require stringent performance from the op amps to
8/4/2019 EEWeb Pulse - Issue 9, 2011
17/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 17
TECHNICAL ARTICLE
extract very small signals riding on relatively large
inputs.
In order to successfully amplify these small signals,
the op amp must have ultralow input offset voltage
and minimal offset temperature drift, with wide gainbandwidth and rail-to-rail input/output swing. (Rail-to-
rail input swing is not needed for small input signals,
of course.) It is also critical for the op amp to offer
very stable ultralow frequency noise characteristics at
close to DC conditions such as 0.1Hz to 10Hz.
For high-precision weigh scale bridge network sensor
applications, designers should look for a single zero-
drift op amp that features very low input offset voltage
and low noise with no 1/f to 1mHz.
As illustrated in Figure 1, a good example is thechopper-stabilized zero-drift ISL28134 op amp delivers
excellent noise voltage (nV) across the range from
10Hz down to 0.1Hz, thus providing virtually flat noise
band to DC level. Leveraging the inherently stable
chopper-based design, the ISL28134 specification
Time (s)
Voltage(nV)
0
300
200
100
0
-100
-200
-300
1 2 3 4 5 6 7 8 9 10
Figure 1: ISL28134: 0.1Hz to 10Hz Peak-to-Peak Noise Voltage
actually includes a maximum noise gain of 10 PPM(Seven Sigma) to offer optimal performance for high-
gain applications while minimizing noise gain error.
For portable weigh scale applications where low-
power is also an important consideration, designers
may want to consider the ISL28133, which combines
ultralow micropower (25A max) and low voltage
offset (6V max) characteristics with a chopper-
stabilized design that delivers flat noise band to DC
and near-zero drift. For other strain gage applications
that need to use higher reference voltages, such as
10V instead of 5V, designers should also consider the
ISL28217 or ISL28227.
Current Sensing & Control Applications
There are a number of different ways to sense
current levels depending on the specific application
requirements. These include shunt sensors using
resistors, Hall Effect sensors and current transformers.
In this example, we will look at op amp requirements
for use in shunt sensor applications. Todays shunt
sensor techniques have evolved to provide a high level
of accuracy and also offer the advantages of lower cost
and applicability across a wide range of requirementsand deployment scenarios.
Basically, the shunt sense methodology places a
resistor in the path of the power supply source being
measured. Because the resistor drop impacts power
efficiency, it is generally desirable to use the smallest
resistor value possible. Once again, this means
that the current sensing application must amplify a
relatively small differential power drop in resistance
into a large gain.
Therefore the op amp circuit must offer high commonmode range and high accuracy. Low power is also an
important requirement, especially for current sensing
in battery applications. Embedded current sensing
circuits also need to be relatively inexpensive so as
to not add significantly to the BOM cost of the product
that is being monitored.
In addition, for many industrial, utility and
communications current sensing applications, the
op amp needs to minimize drift over extremes of
temperature and extended time periods. For example,
current sensors deployed on top of utility poles areexposed to relatively harsh environmental swings
and need to provide consistent performance over
long periods of time without incurring the expense
maintenance requirements.
Many shunt based current sensing applications are
built using op amps such as the ISL28133 or ISL28233,
which are chopper-based, zero-drift amplifiers that
8/4/2019 EEWeb Pulse - Issue 9, 2011
18/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 18
TECHNICAL ARTICLE
combine both low power and high accuracy in the
smallest package size on the market. In addition, as
illustrated in Figure 2, these chopper-stabilized CMOS
devices provide excellent low drift characteristics
over both temperature extremes and extended time
periods.
Temperature (C)
VOS(nV)
-40
8
7
6
5
4
3
2
1
0-20 0 20 40 60 80 100 120
ISL28133
Months
VOSDRIFT
(nV)
0
0.50
0.40
0.30
0.20
0.10
0.00
-0.10
-0.20
-0.30
-0.40
-0.50
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
ISL28233
Figure 2: Minimizing Vos Drift over Temperature and Time, the
ISL28133 is a single chopper-stabilized op amp and the ISL28233 is
a dual of the same amp.
Current sensing is already one of the most pervasiveapplications used across a wide range of industry
segments (consumer, industrial, communications,
utility, etc.) and it is only becoming more important
with the proliferation of new electronic devices and the
increasing emphasis on green power management
techniques. The chopper-stabilized precision op amp
devices described above offer very low offset voltage
and offset drift, rail-to-rail input and output, and low
power consumption needed to support the escalating
demand for embedded current sensing applications.
Handheld Toxic Environment
Safety Monitor
The final application example brings together a
number of different sensor inputs within a single device
and illustrates how well-designed op amp circuitry can
help to efficiently handle such a multi-sensor signal
chain within a compact portable device. Handheld
devices used to monitor hazardous environments are
increasingly combining multiple sensors in order to
minimize size while maximizing capabilities. Such
a device might combine a combustible gas sensor,
oxygen sensor and catalytic heat band sensor.
As illustrated by the block diagram in Figure 3, usingmultiple instances of an ultralow power op amp such
as the ISL28194 provides advantages for multi-sensor
signal chains within a small handheld device.
Because these safety devices typically need to operate
in an always-on mode, the ISL28194 ultralow micro-
power profile (450nA max and 2nA when idle) allows
for extended battery life without compromising on
performance. The ISL28194 is designed for single-
supply operation from 1.8V to 5.5V, making it suitable
for handheld devices powered by two 1.5V alkalinebatteries. In addition, because the multiple ISL28194
signal chains can feed into a single ADC (ISL26132),
the overall system-level circuit complexity and parts
count can be minimized.
Because the combustible gas sensors, oxygen sensors
and heat sensors can typically take as much as 10
seconds to settle, the bandwidth of the op amps is
less critical but they need to have a constant bias on
the sensors. Also, as with the previous examples, the
outputs from the sensors tend to be very small signals
so the op amp must provide peak-to-peak noiseflatness and drift characteristics over a large gain step.
Widening Range of Op Amp
Alternatives Is Ready
Already among the most prolifically deployed
electronic components in the world, the usage of op
amps continues to increase. The op amp deployment
8/4/2019 EEWeb Pulse - Issue 9, 2011
19/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 19
TECHNICAL ARTICLE
curve is exponentially accelerating as more devices
incorporate analog sensor functionality, ranging from
the examples described in this article to the exploding
use of millions of motion, proximity, light and other
sensors in industrial and consumer devices.
As with any good design practices, the first criteria
always must be to achieve the systems operational
objectives for accuracy and performance, so low-
noise, low-drift and precision in high-gain scenarios
will always be critical factors for success. Fortunately,
system designers are now able to choose from a
widening range of precision op amp alternatives that
allow them to effectively meet even the most stringent
Low Power, Precision Signal Chain
ChargingSafety
BatteryCharger
System PowerManagement
Battery
OxygenSensor
Heat BeadSensor
FuelGauge
VREF
EA
Buffer/Driver Amp
Buffer Amp
Transimpedance Amps
Transimpedance Amps
Transimpedance Amps
CombustibleGas Sensor
Gain Amps
Gain AmpsSP1 Bus
Gain Amps
ER
EW
USBHot Plug
RS-232
ADC
AlarmSpeaker
EEPROM
ISL28194/5
ISL28194/5
ISL28194/5
ISL28194/5
EL8170/72
ISL28230
ICL3238E
ISL6118/19
ISL21070/80ISL60002
LDO:ISL80101/A and ISL9021Buck Converters: ISL8009A and ISL9104
Fuel Gauge: ISL6295LiIon charge: ISL9205Charging Safety: ISL9200, ISL9212
ISL26132
ISL12030
ISL88001/2/3
UserInterface
Handheld/
PortableDisplay
Integrated Solution
Supervisor
RTC
performance and accuracy requirements while also
balancing power usage, size, parts count and overall
cost.
About the Author
Tamara Schmitz is a Senior Principal Applications
Engineer and Global Technical Training Coordinator
at Intersil Corporation, where she has been employed
since 2007. Tamara holds a BSEE and MSEE in
electrical engineering and a PhD in RF CMOS Circuit
Design from Stanford University. From 1997 until 2002
she was a lecturer in electrical engineering at Stanford;
from 2002 until 2007, she served as assistant professor
of electrical engineering at San Jose State University.
Figure 3: Multi-sensor Handheld Toxic Environment Safety Monitor
8/4/2019 EEWeb Pulse - Issue 9, 2011
20/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 20
RETURN TO ZERO
8/4/2019 EEWeb Pulse - Issue 9, 2011
21/22EEWeb |Electrical Engineering Community Visit www.eeweb.com 21
RETURN TO ZERO
8/4/2019 EEWeb Pulse - Issue 9, 2011
22/22
RETURN TO ZERO
EEWebElectrical Engineering Community
Contact Us For Advertising Opportunities
http://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cx