EEWeb Pulse - Issue 9, 2011

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


2/22

Contact Us For Advertising Opportunities

[email protected]

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/kDYHm7


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



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



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



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



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



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



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



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/jebqhL


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/jpn9V5



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



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



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.


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



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



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



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



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



RETURN TO ZERO


22/22

RETURN TO ZERO

EEWebElectrical Engineering Community

Contact Us For Advertising Opportunities

[email protected]
http://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cxhttp://bit.ly/l6u6Cx

EEWeb Pulse - Issue 9, 2011

Documents