Volume I- Number 10

SPECTRUM February, 1993

ISSN 1066 - 0550

The Wireless, PCS and Advanced Diiital CommunicationsMonthly News Magazette

Single Copy Price $5.00

-t 1993 by RFlss

2.4 GHz Band is Latest SS Target! Wireless Expo Combwon --page 4.

Wireless in 1923 . . .

The RF/SS 1923 Atwater Kent Model1 OB “Breadboard” TRF Radio

Wireless in 1993!See what 70 years hath wrought

(more on page 4)

Spread Spectrum SceneAn RF/SS Publication

P.O. Box 2199El Granada, CA 94018-2199

WR\VARI)ING & AI)I)HESS COHRECI’ION RlXJUlCS’l’El~1’EIIMI’I‘ NO. 7

HighlightsArticle Page

Full Table of Contents 2March Preview 2Letters 3Editorial 4Consultant’s Corner 6Cartoon 7Int’l & Wash. Scenes 9Technical Tricks 10Tale of Two Chips, Pt. 2 12

SPREAD SPECTRUM SCENE isdedicated to the Spread SpeCtrtJm

professional and is committed tobeing the primary source for thelatest news and information aboutthe growth, regulation, andopportunities in this emergingscience.

SSS provides a forum for publi-cation of technical information,advertising, editorials, opinions,and news relating to the emergingf ie lds of our coverage andemphasis. SSS is a targetedcirculation publication. We havechosen a distinguished cadre of themost important decision makers inthis new industry to be the readersof SSS. SSS is sent by first classmail to this important cadre ofindustry “movers and shakers”each month. SSS can present youradvertising message to the keydesigners, equipment developers,programmers, system integratorsand end users in this new industry.Call our 800 number Hotline torequest a Media Kit.

Editor & Publisher:Randy Roberts

Editorial Consultants:Norm Holsing, Marty Roberts

Contributors:Dan Doberstein, MatthewJohnson, Peter Onnigian

R&D Staff:Chris Kilgus, Gary Mitchell

ISSN 1066 - 0550

Published bv: RF/SSP. 0. Box 2199

El Granada, CA 94018-2199

Telemhone Numbers:Voice: 415-726-6849FAX: 415-726-0118

InterneffUUCP Email:hithr@well.sf.ca.us

Adrertisina & SubscriptionHotline:

Voice: 800-524-9285

SSS is published monthly and isavailable by subscription. You canreceive a sample copy by sending$1 .oo.

Subscriotion Rates:12 Months - US First Class Mail - $49.9512 Months - Foreign, AIR MAIL - $70.00

US Funds

Rumora 45 Ramblinpl Ardis (or Motorola/Ardis) isplanning something big! Theyare holding a special press ann-ouncement breakfast on Tuesdaymorning, February 16, at the Mo-bile 93 conference in San Jose.It seems that Ardis’ president andCEO Frank Wapole and WilliamA. Hipp, president of RadioMail,will be pitchin their new jointventure.

l Go Corporation along withpartners AT&T and EO, Inc. arealso planning something for theMobile 93 show. We got a handwritten, pen-based “wireless orig-inated” FAX from them exhortingus to stop by their booth, #2 16.Go shipped its latest application,GO Fax, on January 8 and sentits PenPoint 2.0 Japanese out inmid-December.

l The latest marketing study doneby BIS Strategic Decisions, Nor-well, MA, shows the exciting andprofitable potential of PDAs andother consumer oriented personalcomputing devices. The reportcites work being done by Casio,Tandy, Toshiba, TI, Sharp, Appleand AT&T aimed at this burg-eoning market.

j Decipherings /

._ .__.

Don’t miss an issue ofSpread Spectrum Scene.

Subscribe now!

I

Table of Contents

Article Page

Letters 3

Editorial 4

The Aerial 5

Consultant’s Corner 6

Equipment Corner 7

Ham Radio Happenings 8

International Scene 9

Washington Scene 9

Technical Tricks -- AllAbout Correlators 10

A tale of Two Chips --Part 2, The PMI TRlOOl 12

Nibbles & Bits 16

Think 8pread 8pectrum!h

I

March Preview

l Conclusion of DanDoberstein’s GPSSTransceiver idea article.

l Mobile 93Conference report.

0 Hint’s for more usesof the MC13176 chip.

l Highlights andreviews of RF MicroDevices new chips.

l New Products,columns, SS news andother features galore.t

.%xcad Suectrum Sccnc.._

February, 1993

PACIFIC ~TELESIS..Group

January 21, 1992

Mr. Bandy RobertsSpread Spectrum SceneP.O. Box 2199El Granada, CA 94018-2199

Dear Bandy:

please allow us to set the record straight regarding Pacific Telesis’upcoming spinoff of its wireless businesses. Your article in theNovember-December issue contained several errors of fact.

First, Pacific Telesis is not also known as Pacific Telephone and PatBell. Pacific Telephone ceased to exist in 1984 following the breakup ofAmerican Telephone & Telegraph. Pacif ic Bel l succeeded it , providinglocal telephone service to most of California. Pacific Telesis is theholding company that owns Pacific Bell, as well as several othersubs id iar i e s .

Second, we have no plans to offer long-distance service in a deal withMCI.

Third, we’d love to have 33 million cellular customers, as you stated.The actual number is approximately 722,000. Your figure is based on thepopulation of the regions where we provide service.

Fourth, the separation in no way reduces the authority of either thestate Public Utilities Commission or Judge Harold Greene over PacificBe l l . In fa c t , the PUC retains regulatory authority over Pacific Bell aswell as over the California operations of PacTel Cellular and PacTelPaging. The spinoff will remove the remaining divestiture-imposedrestriction8 from PacTel Cellular and our international operations.Those restrictions, however, will remain in place against Pacific Bell.

Final ly , this move does not “flaunt any old regulations.” The planenhances shareholder value by creating separate, publicly tradedcompanies that can react quickly and responsively in a rapidly changingtelecommunications market.

Sincerely,

SSS regrets aJJy coJ+sicJn we nluy huve caused, coJlcerJ~iJlg this suiIjcct. Our originul sources jijr this

piece were iJmJrrect. We will publish the jidl PacTeI press release on this subject in u future issue.Than.ks, Michael,ji)r straightening us out!

Felwuary, 1993___~ ._______ ~~__

Spread Spectrum Scunc 3

Pseudo Random Tidbits

Magazine deadlines arehumbling events ! In 0t.n Januaryissue, I incorrectly credited thecover picture of the lovely ladyto Qualcomm and some non-existent magazine. The magazinedue the credit is Microwaves undRF magazine -- sosry Jack &John. Unfortunately this samemistake was carried into the arti-cle on the Wireless Symposium& Exhibition, also. All I can sayis: sorry guys.

We also need to learnhow to use our spelling checker alittle better -- in our editorial, onpage 2, we were unable to spellpredictions! Oh well, we’re stillmaturing and will probably al-ways need a little more proof-reading than time allows.

While we are talkingabout our goofs in January, letme correct the ad on page 13 foiJohn Kratz -- his phone numbersare reversed, in case you didn’tnotice when you tried tocall/FAX him. Very sorry John.

Another Wireless Newsletter

WCCN Publishing hasstat-ted an interesting new news-letter, covering mostly the appl-ication side of IDP (Interactive I

_

Data Processing) and wirelessPDA (Personal Digital Assistant)type devices. Not really a com-petitor (it’s not highly technical innature) WCCN is interestingreading, if you can afford itssteep price. Call or FAX Mr.Thomas A. Polizzi, President ofWCCN Publishing for infoi-ma-tion or a sample.

WCC&WCC~TubCis&iing

T-0. 230~ 426

Sprhgdi, IA 52336-0426

3 1 9 - 8 5 4 - 7 3 3 3

fq: 319-854-7336

Wireless Exposition W r a pThe Mo\v to 2.4 GH,

As our cover headlinestates, our main conclusion fromthe January Wireless show wasthat everyone wants to move tothe 2.4 GHz ISM band. HKmicrowave showed an FM TVlink at this frequency that workedvery well. HP and Triquint hadnew chips for this band. Britishgiant GEC-Marconi demonstrateda really neat GaAs transceiverchip for this band. RF Associ-ates, the local GEC-Marconi repassures us that samples will beavailable for $500 each in March-_ with production prices comingdown under $20 per chip thisyear. The session that I chairedat the show on Friday featured atechnical paper on this chip. Itpacks a lot of functionality into asmall slice of GaAs! Look into

this chip if you want to jump intothe 2.4 GHz waters. However,beware of the capricious Britishindustrial policies and economy --I’ve been burned more than onceusing Plessy parts!

What 70 Years HathWrought

Cove,. Feciture

Did we get your attentionwith my old 1923 Atwater Kent,model 10 breadboard radio?What a contrast, huh’! 70 yearsdoes make a difference in tech-nology -- doesn’t it’.’ Wireless in1923 was the latest craze -- muchlike “Wireless” is today. Butyesterday’s wireless became radioby the thirties, the nation entereda severe depression and the morethan 2000 radio manufacturersthat had sprung up to answer themagic of “wireless” mostly wentaway. I think history does repeatitself! Today wireless while littlelike yesterday’s, is ripe for feastor famine.

The contrasting photo onour front cover shows some ofYamada’s prototypes which use aSAW (Surface Acoustic Wave)convolver to rapidly synchronizeSS signals at 900 MHz graphical-ly shows how technology has ad-vanced in 70 years. But in thebroadest sense have we learnedany lessons -- have we improvedmankind’s lot’! I think not!

Spread Spectrum devicescan help mankind if we put OUIminds to it! Let’s come up withsome educational, health care orbaby monitoring applications --not just more burglar alarms!

4 Spread Spcctrutn Scene February, 1993

by Peter Onnigian, P.E., WGQEU

ELECTROMAGNETICRADIATION

Wireless data transmissionutilizes electromagnetics. Just whatis this‘! The phenomena that arerepresented by the broad term elec-tromagnetic radiation are amazinglydiverse and far reaching. They areall characterized by interaction be-tween objects, the radiation sourceand the radiation receiver. The em-itter (transmitter) may be widelyseparated in a vacuum, in air. ininterstellar space, buildings, walls.trees, or what have you, from thereceiver.

Examples ofelectromagneticradiation are the warming of yourbody by sun light, perception of astar by the eye, or reception of a RFsignal from Voyager, millions ofmiles away! Human tissue and boneis illuminated from an X-ray sourcetransmitted thru tissue to the X-Rayfilm (receiver). These are only afew examples of electro-magneticradiation, which we take for granted.

The change in the receiveroccurs after a time interval that isconsistent with the understandingthat something has traveled from theemitter to the receiver with thespeed of light. The speed dependson the velocity factor of the medi-lm.

In a configuration of con-ductors and dielectrics (antenna)current flow will produce electro-magnetic fields and electrostaticcharges.

We have all heard of Max-well’s equations and their relation-ship to electromagnetic currents andvoltage charges. Maybe you’ve for-

Fcbruarv, 1993

Jotten much of it though, since tak-

11g Electromagnetics 10 1, say thirty{ears ago. Maxwell was a genius,vho explained this type of radiation;vith math equations. One of his:quations worthwhile recalling is:hat radiation decreases about 6 dBevery time you double the distanceoetween transmitter and receiver.

Perhaps it’s best to leave thetheory of electromagnetics at thispoint, without getting into the Diracdelta function free space Green’sfunction, or scaler and vector poten-tials. It’s a very complicated subject,taking about two years to cover atmost specialty universities.

It’s very important to re-member electromagnetic radiationcontains both electric and magneticfields. Their relationship is verycomplicated, but nevertheless theyexist together.

We have all heard of Max-well’s equations and theirrelationship to electromag-netic currents and voltagecharges.

Now in antennas used forRF data transmission, electrons (cur-rent) are forced to oscillate in sync.Such oscillation is limited to thefinite time during which the trans-mitter is active, causing current tooscillate at the carrier frequency if itis unmodulated. The current tlowoscillates by changing flow directiontwice each cycle. It’s this currentflow that causes magnetic fields,which causes radio frequency radia-tion, the good stuff we need forinformation transmission from pointA to Point B!

Basic Radiation

There are two inseparablefields associated with the transmittedsignal, an electric field due to volt-

Sprwd Spectrum Scene

age and a magnetic field due tocurrent flow. The oscillations ofcurrent and voltage occur at twicethe RF frequency. These fields areat right angles to each other and tothe direction of propagation.

The electric field is mea-sured by the potential per unit dis-tance. This value is referred to asthe field strength.

Polarization

By convention, the directionof the electric lines of force definethe direction of polarization of radiowaves. Horizontal dipoles (parallelwith the earth) radiate horizontallypolarized radiation. And verticaldipoles propagate vertically polar-ized waves.

In free space, very remotefrom ground effects and the earth’satmosphere, these senses remainconstant. A suitably aligned receiv-ing antenna would respond best tothe whole of the incident field prop-erly aligned.

Send your antennaquestions to

Peter Onnigian.

ANTENNAS FOR WIRELESSALARMS AND DATA

Extend your wireless range upto 30 miles using Ham-Proantennas specifically designedfor 150,450 and 915 MHz.Spread spectrum and alarmsystem reliability is greatlyimproved using these Yagiantennas indoors or outdoors.

Hc~M-?ROA N T E N N A S

6199 El Warehouse WaySacramento, CA 95826

(916) 381-4469. Fax: (916) 381-4332

5

By Gary Mitchell

This month marks the begin-ning of a new monthly feature in SSS.tentatively titled Consultant’s Corner.This column is for those readers who.by choice or not, are in the process ofbecoming consultants rather than full-time employees.

An Introduction: As an introduction,I thought I would say a few wordsabout myself. I have been au engineerfor about 14 years now, most of thattime as an employee, usually of largecompGes.

In 1990, after two successivelayoffs, I decided that it was time totake charge of my own career, ratherthl leave it in the hands of some com-pany. That’s when I started consulting.It’s been a slow start, and it definitely isllot a way to get rich quick, but to meit’s a lot better than being an employee.

I doubt that there is anyonewho hasn’t noticed that we have under-gone a very painful economic restructur-ing the past couple of years. Partly as aresult of this painful restructuring, theUS (and Canada) are now poised to be& low cost producers of electronic equ-ipment.

When you consider the down-sizing of companies, the low intereslrates and the pent-up consumer demand,it’s obvious that we are ready for a bigchange. I think that we are going to seea resurgence of manufacturing, but youwon’t see it in the big-name companiesthat you think of when you think aboutthe electronics industry.

The new growth will be insmall companies (and individuals) thatrecognize the singular opportunity thatnow presents itself. We RF engineerscan (and must) be a big part of that.Manufacturing is where the high-payingjobs are, not just for us, but for the restof the country as well, and we hold the“keys to the kingdom,” the ability tomake things happen with hardware.

___~_ ..-..6

When you consider that, and all of thenew te1ecommunications that is liappeii-ing now (most involving RF), you real-ize that it’s up to us, there simply _ noone else to do the job. So let’s get to it!

Getting Started: I want this column tobe an interactive, two-way forum. Pleasewrite to me or call me with questions,suggestions, complaints or just aboutanything -- I like talking to RF peopleabout most anything. My address andphone number are printed at the end ofthis column and I would like to hearfrom you. You can write to me if youwish, but I would rather talk on thephone. Call me during the day, leave amessage. and I’ll get back to you in theevening.

One of the first things I wouldlike to hear is a better name for thiscolumn. We are calling it Consultant’sCorner right now, but we would like to

have a better-sounding name, and it’s up

to you.

Product Review: As I will discuss infuture columns, one of the most impor-taut things we have going for us as coii-sultauts is an elicyclopedic knowledge ofthe parts available to do the job at hand.To facilitate this, I will frequently re-view new RF products.

This month’s product is a quad-rature modulator, the RF2402, from RFMicro-Devices. The chip, shown below,is a complete quadrature modulator inone inexpensive monolithic package.

This single component replacesa quadrature phase splitter. two mixersaud a summer with one inexpensivepackage. The device operates over a fre-quency range of below 100 to above1000 MHz.

So how might you use thisdevice’.’ Well, suppose you had a chanceat a job that required a QPSK modulatorat 222 MHz for a mobile data service.You could use a conventional approach.with mixers, splitters and combiners,

moduulating at a low IF and then up-converting.

Or you could use RF Micro-Devices quadrature modulator chip thatmakes it possible to do the same jobusing just a synthesizer at the finaloutput frequency and the quadraturemodulator chip (with some appropriatefiltering added). We’ll be talking aboutinexpensive synthesizer chips in the nearfuture.

All of this simplicity comes ata cost of only $12.32 per device, inquantities of 100. You really can’t affordflit to use this device, and if you callChris Fisher at RF Micro-Devices, I bethe’ll be happy to send you a sample ortwo. (Help Randy, tell Chris that youheard about the part in Spread SpectrumScene). They also make a companion&modulator chip and we’ll talk about ituext month.

One more thing. There is notli-ing as aggravating to me as reading adata sheet or product review and findingeverything except a price. It’s like theydon’t think price is an important designparameter! So I promise you, any prod-uct reviewed in my column will includea price, or it won’t be in the colunu~ atall!

Next Month: The fhure topics of thiscolu~nn depend to a great extent on whatyou want to hear about. I have someideas, but if you write or call me, I’ll tryto get to what is on your mind.

RF Micro Devices can be reached at:7341-D W. Friendly Ave.Greensboro, NC 27410(Y 1Y) 855-X085

NOVARF Systems, Inc.The Complete RF/Microwave Solution

-RF/Microwave Systems-Custom Design/Consulting-Simulation Software-Synthesizers (PLL/DDS)-Complete Lab/Machine Shop-TDMNCDMA/Spread Spectrum

11740 Pine VaII;;5,_;~na, VA 22182

Spread 3pcctruw Sccnc February, lY93

E q u i p m e n tC o r n e r

by Chr is Ki lgus

EQUIPMENT CORNER

I promised to do a review ofthe receiver design workshop byRichard Webb at the WirelessSymposium in San Jose last month.However, Mr. Webb is a little latemailing out his software -- maybenext month.

The San Jose ConventionCenter, where the WirelessSymposium and Exhibition washeld, is a beautiful facility. RFExpo is coming March 17-19. Ithink it will be a real challenge forthem to out do the WirelessSymposium.

Why do people, like CrystalFilter vendors, list all their standardproducts in a catalog, and then areunable to offer a sample or smallquantities for prototyping in lessthan lo-12 weeks? A 10.7 MHzcrystal filter for example, can onlyhave a few standard varieties, can’tit? Two or four pole, with fourdifferent bandwidths. That is onlyeight crystals, yet it took me monthsto obtain a few for evaluation. Theyshould be off the shelf.

Nicad’s are improving.Smart chargers and better chemistryhave helped, but we have all hadbad experiences with them. Did youknow that it is accepted engineeringpractice to float charge alkalinebatteries? Yes, there are warninglabels on the batteries, “Do NotRecharge”, etc. That is to preventgross charging currents. It isperfectly okay to replenish analkaline battery with a l\lOO to1\200 capacity charge. Tiy it onyour next low power design project.

The latest media blitz about

EM Fields causing brain cancer hasmade everyone think about what weare doing. Apparently the brain andeyeball resonate at about 1 GHz.Some Japanese scientists observedthe greatest temperature rise at thatfrequency. The eyes have thegreatest vulnerability, since they areessentially unprotected. Cataractscan be the result. The ham operatorrunning 1500 watts into a beam only20 feet away certainly is at risk. Ifthere is in fact a danger, we onlyhave to look at how long cigaretteswere considered a non health hazardto possibly estimate the time beforethe truth is known.

To design an unlicensedradio system, one almost has tojump immediately to 900 MHz tofind a modicum of spectrum to workwith. It is disgusting the pittance ofspectrum that is allocated forpersonal unlicensed operation.There are only a few possibilitiesbelow 900, the 80 KHz from 49.82 -49.90 MHz is one. The non-critical

February, lYY3 Spread Spectrum Sccnc

RF circuits and readily availableIC’s make life easy here. Theproblem is that you will be sharingthe band with radio controlled toysand cordless telephones. Unless youwant to try working with the 160 -190 KHz band, that’s about it.There are other bands, but thepower and duty cycle limitationsmake them rather useless except forremote control. I hope that the FCCwill take a better look at how ourspectrum is being utilized. Thevarious commercial spectrum usersshould all tighten their specs andmove closer together -- or join the20th century and use efficient digitalmodulations! How about that sacredcow of the Hams, 6 meters? Rightthere in the middle of it all (SO - 54MHz) is prime spectrum that isalmost unused. Use it or loose it!

I appreciate the commentsthat I have received from variousreaders in the last few months --keep those cards and letters coming.Until next month.

___-I

Ham Rai3ioHappenings

by HCXZR;~ Ham, NXGX

Welcome to this newcolumn! We plan to cover avariety of topics of interest toamateur radio operators, studentsand advanced experimenters inthis column. We welcome mate-rial for use in this column --justsend us anything that we mightuse, if it really belongs here, we’lluse it and acknowledge thesource.

Experimental Group within theCalgary Amateur Radio Club.The group is currently focused onthe completion of the conversionof a number of low priced, sur-plus cellular rigs to the 900 MHzband. The rigs present an oppor-tunity to not only run voice, butalso high speed digital acrosstown. Plans are in the works fora local repeater also.

First off, I’d like to share

OtherExperimental Groupinterests involve DSP and SpreadSpectrum. These two interestsare currently on the back burneruntil the cellular conversionproject is finished. Rustyexpressed continuing interest inour SS kit and wanted furtherdetails. If some more of you areinterested, we may still resurrectour SS kit idea -- let us know.

some information we received Other items: the picturesfrom Rusty Rushton, VE6TL, of below are from the January meet-Calgary, Alberta, Canada. Rusty ing of the High Tech Experim-says that they have formed an enters Group (formerly PPRS, for

Peninsula Packet Radio Society).This meeting was attended bysome 30 “high tech” bay areahams and featured demonstrationsof working RDF (radio directionfinding) equipment by some ind-ustrious Bay Area hams. If youhave interesting photos or storiesfrom your experimentally inclinedham club -- send them along!

Final ly , Gerd Schrick,WB81FM of Dayton is sending ushis club’s newsletter titled: Ano-malous Propagation. Very int-eresting, the group is The Mid-west VHF-UHF Society and itsannual membership is only $6.00.The publication shows a realexperimental flavor in the groupsactivities and interests -- my kindof club. The newsletter remindsus that the Dayton HAMVENT-ION is April 23, 24 and 25 thisyear.

- - -

Tune in to Qpread 8pectrum &ene!

_ __~_. ______.__.___- - __.__-. _ _.. ~__Spread Spectrum Scene February, lYY3

l One of the best articles I haveseen on the subject of the Euro-pean Digital Cellular Market waspublished in the January issue ofMicrowave Journal. The article,by Guy Daniels, a ContributingInternational Editor, is notewor-thy for its clarity, coverage andcompleteness. SSS can makecopies available for this article ifyou have trouble finding it. It’srecommended reading.

l L. M. Ericson, Stockholm,Sweden and Hewlett-Packard,Palo Alto, CA have formed ajoint venture to provide telecom-munications operators worldwidewith systems that combine net-work management and adminis-trative/customer support. Thenew venture will be based inStockholm and will have facilitiesthroughout Europe.

l Inmarsat Project 21 {a poorlydefined, worldwide low-earth-orbit (LEO) mobile satellitephone service is still giving Mo-torola fits. Motorola’s Iridiumsystem, which could cost thenand their partners up to $3.4 bil-lion, is certainly a competitor olwhatever happens with Projec,21. Motorola and on-again, offagain partner Lockheed Corpor,ation, Missile and Space Division, are once again looking foinvestors. Can Inmarsat one-u1Motorola/Lockheed?

February, 1993

Ft

r3

l Three regional telephonecomppanies have asked the Fed-eral Communications Commis-sion to investigate the proposedmerger of American Telephone &Telegraph Co. and McCaw Cellu-lar Communication. Bell Atlanticcharged that the proposed merger“raises competitive and publicinterest concerns of the most fun-damental kind,” in a petition filedwith the FCC in conjunction withNYNEX and Southwestern Bell.The proposal would combine thecompany that dominates the long-distance market with the nation’slargest provider of wireless ser-vices.

l Under pressure because of therecent scare over possible linksbetween cellular phones and braincancer, FCC officials last weekproposed to change a longstand-ing guideline on allowable EMFexposures -- an action that couldslow development of next- gener-ation wireless technologies. Thenew standard would reduce thelimit for the power radiating fromcurrent cellular phones to one-tenth the value of that allowedunder the old standard and wouldrequire next-generatiion systemsto operate at half the power ofexisting ones. The commissionplans to replace an Amercan Na-tional Standards Institute standard

Spread Spectrum Scene

rdopted in 1985 with one pub-ished by the Institute of Elec-rical and Electronics Engineers‘IEEE).

l QUALCOMM Inc. and partnerTelesis Technologies Laboratory(TTL), subsidiary of Pacific Tele-sis, are jointly testing a PCSsystem in San Diego. The systemis expected to operate in the 1850to 1990 MHz band using QUAL-COMM’s code division multipleaccess (CDMA) digital transmis-sion technology. The FCC grant-ed QUALCOMM an experimen-tal license to conduct PCS fieldtrials last year. According to theFCC’s notice of proposed rule-making, published in July 1992,the advent of PCS could have agreat impact on the future devel-opment and configuration of alltelecommunications networks.The 1.25 MHz, CDMA channeloperating in the 1800 MHz freq-uency is considered narrowenough to allow spectrum shar-ing, while road enough to offerall the advantages that can begained from a CDMA-based PCSsystem.

l NASA’s Jet Propulsion Labo-ratory (Pasadena, CA) has awar-ded a contract to the Pacific Ad-vanced Communications Consor-tium (PACC) to examine the useof satellites in personal commu-nications networks (PCNs) forrural users. The PACC is com-prised of several Oregon andWashington communicationscompanies, universities, and gov-ernment agencies. For the JPLstudy, the PACC will evaluatethe role of satellites in remoteand rural PCNs, focusing onusers in the Pacific Northwest.

(1

Technical Tricks

by R. H. Roberts, DirectorRF/Spread Spectrunl

The real “art” or t radesecret technology of Spread Sp-ectrum is in the acquisition andtracking of code phase, carrierfrequency and data clock. Theonly “magic” involved is a prac-tical knowledge of how to do itwith cot-relators. Cot-relatorscome in various types:

AnalogSAW correlatotSAW convolvesCCD delay lineDoubly balanced mixer

DigitalFull parallelSlidingHybridDSP algorithm basedAI “smart” correlatoi

There are plenty of references oncot-relators and lots of theoretical

analysis of how they are sup-posed to work. However, in thereal world it is best to have afavorite circuit or two that hasworked well for you in the pastand then adapt or modify it for anew application. I will present afew such circuit ideas here.

First, the simplest andquickest cot-relator to get up andrunning, by fdr, is the simpleserial, sliding cot-relator witheither two (Tau Dither) or three(Delay Lock/Early-Late) chan-nels, each containing EXOR’s orDBM (doubly balanced mixer)cot-relators. In this approach, onechannel is devoted to “on-time”or data channel correlation. Inthe Tau-Dither, two channel sys-tem, the second channel is timeshared between a slightly “early”and a slightly “late” timing offsetchannel used to form a “discrim-inator” function for code trackingput-poses. In the three channelsystem one timing channel isalways “early,” while the other isalways “late.” Again a discrim-inator-like error function is gen-erated to enable code tracking.

To better understand the

HSP45256Binary Correlator

strategy behind the sliding corr-elator, assume that the receiverhas no knowledge of the codephase or frequency to be re-ceived, at all. The simplest stra-tegy is just to sequentially tryeach possible code position, untilcorrelation is found. The “data”channel, mentioned above, isused to detect “code lock,” sincethe signal instantly de-spreadsand a narrow band carrier (possi-bly with data modulation) magi-cally appears when code lock isachieved.

Sliding correlators aresimple, reliable and slow! Ahybrid, serial/parallel or “pipe-lined” approach can speed up thistype of correlator by a factor ofN’ , where N is the number ofseparate, parallel pipelined cha-nnels. Thus a 3 way pipelinedhybrid sliding cot-relator, whereeach parallel pipelined sectionexamines a different section ofthe code, can acquire sync about9 times faster than the simplesequential, sliding correlator. Agreat return for a nominal addi-tion of circuitry. Today’s PLD(Programmable Logic Device)

TMC2023Advance Information

AMOS Oigital Output Carrelatori4-Bit. 30MHz

hBlChdBbekDii

Spread Spectrum Scene

-1

February, 1993

technology makes it easy to im-plement hybrid sliding correlatorsup to near the complexity of afull parallel digital correlator.

However, the fastest cor-relators are fully parallel devices__ they search the entire codeepoch length all at once. Thesedevices can use CCDs, SAW(Surface Acoustic Wave) or di-gital LSUASIC technology.SAW convolvers can designed tobe programmable for any code --but, the most useful and generalpurpose parallel correlator is theall digital device. The chip blockdiagrams at left and below showsome of the available ASIC offe-rings from Harris, TRW andZilog. Actually Zilog has li-censed the SS technology devel-oped by Stanford telecommuni-cations, Inc. for consumer scalecommercial development.

The chips shown here arejust a sampling of what’s avail-able from these and other vendorsout there. All three chips shownperform superbly in a correctlyinterfaced SS system. There isan art to using any of these chips,however. It seems that to even

read the data sheets of thesechips you need a PhD in micro-processors and silicon BiCMOStechnology. Each vendor doesmake available a certain level ofapplication support -- StanfordTelecom sells evaluation boardsand c o m p l e t e d e v e l o p -ment/simulation circuit boardsubsystems. My recommendationis to select a chip based on theperformance you need, build up asimple all digital test circuit first,then proceed slowly, in smallsteps, to integrate your new cor-relator into your SS system. Thisway you will learn some of theidiosyncracies of the chip at eachstep of your design/integrationproject.

Many companies havespent hundreds of thousands ormillions of dollars developingtheir own full parallel digitalcorrelators. Save your companyand your project (as well a yourreputation) the time, trouble andexpense -- use an existing LSI /ASIC parallel digital correlatorchip.

Parallel correlators cansync up in as little as one code

23340DUAL 64-TAP, 11 MCPSDIGITAL MATCHED FILTER

GENERAL DESCRlPllON (Contmued)

epoch (the code repeat time inter-val). However, noise and statis-tics usually enter the picture byforcing certain P,, and P, req-uirements on you. It is thustypical that all digital parallelcorrelators synch in perhaps 3 to5 PN code epochs (data bittimes). Even this speed is blaz-ingly fast compared to the slidingcorrelator which syncs up, atbest, in the code length’s numberof data bits.

The use of digital circuitryfor correlation provides interest-ing challenges to the SS innova-tor -- first it forces him to includeanalog and digital circuitry bothin his design. Next, he mustlearn something of the rudimentsof Digital Signal Processing, if heis to succeed in his efforts. Fi-nally he must learn, by trial offire and smoke, that SS design isfield for those brave, perseveringfew, who can master multipletechnologies and disciplines.

Next month we will givecomplete block diagrams andcircuit descriptiions for both thesimple sliding and the hybridcorrelator designs.

U. S. Headquarters -- Harris Semiconductor, 1301Woody Burke Road, Melbuurne, FL 32902, TEL:407-724-3000

ASIC and Custam Products Grnup -- StanfordTelecom, 2421 Mission College Blvd., Santa Clara,CA 95054-1298, TEL: 408-980-5684, FAX: 408-980-1066

TRW LSI Products Inc. -- 4243 Campus PointCourt, San Diego, CA 92121, TEL: 619-457-1000,FAX: 619-455-6314

Zilog, Inc. - 210 Hacienda Ave. , Campbell, CA95008-6600, TEL: 408-370-X000, FAX: 408-370-8056

February, 1993 Spread Spectrum Scene 11

A Tale of Two ChipsPut-t 2

by R. H. Roberts, DirectorRFiSS & The PM1 Staff

The PM TR-1001 -- A GaAsMMlC Transceiver Chip fol

915 MHz

Cordless phones, PDAs,spread spectrum radios and otherLAN/WAN or “Wireless” devicesneed RF transmit/receive frontends . The Pacif ic Monolithics,Inc. PM TR-100 1 provides a veryinteresting complete, one-chipsolution for this “need.”

The MMIC t ransceiverchip, in the figure at the right,incorporates an upconverter and

L

downconverter both fed by thesame on-chip local oscillator. Theupconverter is composed of apre-amplifier, a single ended mix-er and a post- amplifier. Outputpower is +15 dBm into a 50 ohmload at 915 MHz with 0 dBmi n p u t a t 90 M H z . T h edownconverter consists of a lownoise RF amplifier, a single-endedmixer and an IF amplifier. Con-version gain is 25 dB with a SSBnoise figure of 6 dB. The localoscillator is a capacitive sourcefeedback design and tunes about 4MHz using a tuning voltage of 0to -3 volts on an off-chip varactor.The tuning bandwidth is enoughto cover 160 channels with a 25KHz channel spacing. By lower-ing the resonator Q, the oscillatorcan be tuned over a wider band-

12 Spread Spectrum Sccnc February, 1993

width at the expense of higherphase noise. For high volume andlow cost applications, the chip ispackaged in a 24 pin surfacemount plastic package as shownat left. In the transmit mode, thechip draws 60 mA off a 3.6 VoltNi-Cad battery used in manycordless phones. In the standbymode, the chip draws 15 mA.

Functionality of the GaAsMMIC chip was demonstrated bybuilding two FM transceiversystems (see functional blockdiagram below). This provided atest bed for correlation of keyelectrical parameters (such asphase noise and output power) toaudio quality and area coverage.Phase noise is one importantaspect in this application, sincephase noise in the 300 Hz to 3kHz offset is directly convertedto audio noise in FM systems.The GaAs oscillator with anoff-chip coaxial resonator exhibitsa root mean-square (rms) phasenoise of 110 Hz over the range of300 Hz to 3 kHz offset from the

carrier. A comparable siliconbipolar oscillator was also de-signed and exhibited 60 Hz ofphase noise. Although theperformance of the silicon designis slightly better, the GaAs designexhibits respectable performance.A good audio signal could beheard for rms phase noise lessthan about 150 Hz, above whichbackground noise becomesirritating. The line-of-sight rangeachieved was well over 50meters.

This chip provides the RFheart of any communicationssystem using the 900 MHzfrequency range. A low poweredSS transceiver can be designedwith this chip very simply byadding the requisite spectrumspreading mixers to the IF inputand output signals.

An evaluation kit isava i l ab l e . Con tac t PM1 at:Pacific Monolithics, Inc., 245Santa Ana Court, Sunnyvale, CA,94086. Phone: (408)732-8000,FAX: (408)732-3413.

New! =MATCH=Suite of matching network solutions~Narmw or broadband. comDlex loads

=OSCILLATOR=Unified oscillator design process:>L-C, distributed, SAW, & crystal designsP !Sstimates SSB, residual FM & PM noiseP$795, i @I& $495 with =&qerStar=:& .* ‘ : II ~yf&vy.i~gw.ti~~~~ “&&~1.7

=FILTER=Complete L-C filter synthesis package:P ConventionaI, narrow, s mmetric&zig-zagaNumerous all-pole & e&ptic functionsh =EQJALIZ@=..for delay equalizersp $795” 1‘ ,@I& -$495, @itIt %supj?+r=, Yx” b” *‘id

=TLINE=Relate physical & electrical parameters:

=SuperStar=RF/Microwave circuit simulation, tuning,optimization and statistical analysis:&True real-time tuning and optimization&Huge library of corn onents & modeis&Runs at;&l&rb% lx!%

For information or to orderCall (404) 939-0156FAX (404) 939-0157INFO/CARD 51Or write:

’ ”Eaaleware Corvoration LiiHiBl175uO Mount& GlenStone Mountain, GA 30087USA

Progrnms run on standard IBM andcompatible PCs, include a 30 day satisfaction

fuarrtntee, free support and no annual fees.mmedute shipment aanilable.

Same prices internationally. Direct factorysales and user support by phone or FAX.

~~~___________-February, 1993 Spread Spectrum Sccnc 13

@ The subject of “Wireless” hasfinally arrived! February’s BYTEmagazine covers the field in 3feature articles and a resourceguide. It makes very good read-ing for those new to the subjector those hungry for lay levels ofexplanation. Pick it up, look itover -- it really isn’t bad. Makesure you browse pages 146 to177.

0 Pulson Communications Inc.,McLean, VA has announced its“successor” to Spread Spectrumtechnology. The controversialtechnique is known as “ImpulseRadio.” Using pulse-positionmodulation and pulses as short as0.5 nanoseconds, the system “sp-reads” its information across theentire spectrum instead of justacross one frequency band. TheFCC has awarded Pulson a Pio-neer’s Preference to use the sys-tem in field trials. However, the

I FCC has no allocations where/ this goofy system may be used.

Q Even staid Electrokc Design,in. its February issue, covers som-ething about our vast subject

area. On pages 61 to 67, Elect-rotzic Desigrz presents the article“Communications Terminals GetPersonal.” It’s pretty good read-ing material on the subject ofPCSs, in general, and regulatoryproblems facing PCSs, in partic-ular.

DlAGNOSlSHW SW Behavioral TestingSpectral Intelligence AnalysisMULTIPLE IwxEuLxGmTcm

TREATMENTGroup TherapyLeft/Right Brain TransplantsCosmetic MakeoverReConstnxtive Surgery

House Calk-Out-Patient ClinicPhone/FAX 510 527-8736

Subscribe toSpread SpectrumScene -- today!

Spread Spectrum Sccnc February, lYY3