synthesized pulse generator

7/30/2019 Synthesized Pulse Generator

1/5

This synthesized pulse g enerator has a range of from 1/100 Hz to 1.000 MHz.Use it far wo rking w ith logic circuits-or with analog devices wel l into the HE range.GARY McCLELMN

N E OF T H E H A N D I E S T P I E C E S OF ELEC-uipm ent y ou can have is a good

The Programma 1 will change that.

e gene rators, or oth er sources. The1 marries the frequency

This design has many exciting fea-via four BC D (Binary Coded

9990 possible frequencyavailable, with e ach oneI is a

the frequ ency range even farther!t, you c an readily generate signals0.01 H z t o 1.OWI MHz. The* 0.005%, if the ge nera tor is ac-calibrated. As far as the outputages ar e concerned , you have your

ndard TT LICM OS output,lsta ble to 5-volt output. Thisrposes like running

R FI is also use-

ful for AM radio alignment. Still otherfeat ures include drive capability for oneTTL load, and an ERROR lamp that tellsyou that the frequency selected is cor-rect . T his lam p is helpful as a diagnosticdevice, should troubleshooting becomenecessary.There's more

N ot t o be overlooked is the design ofthis instrument. Thanks to the latestCMOS circuitry, it uses just ten IC's.Contrast that number with the seventeenIC's that are normally required in acomparable T TL system. Besides a re-duced IC count, you get CMOS ad-vantages like low power consumption,abs enc e of drift-causing heat, and a lessnoisy signal. Also, the construction hasbeen simplified to one small, single-sided PC board, that you can easilymake or buy. Not to be neglected, theothe r parts have been kept to a minimurnby careful engineering, to m ake buyingthem easier. In fac t, great care has beentaken to insure that all parts for thisproject are readily available. You canexpect to be able to assemble the Pro-gramma I in just a few evenings, thanksto its simplified circuitry and goodparts-availability .For the futureWith "smart" test equipment on thehorizon, or instruments that interfacewith com puters, this project will becomemore useful. By replacing the program-ming switches with appropriate IC buf-

fers , the Programma 1 may be controlledby a microprocessor, automatically gen-erating the frequencies required. Thistechnique is being used in industry fortesting, a nd even alignment, of finishedequipment. It's a big money saver, andyou'll be hearing a lot more about auto-matic testing. The Programma 1 ha sthis automatic test-capability built inright now, ready for the future-some-day you'll appreciate that!Theory of operationFigure 1 show s a block diagram of thepulse gen erator, so refer to it for detailsas you read the circuit description.Although the diagram has been strippeddow n t o just the basics, the actual cir-cuitry isn't much more complex. In fact,you are going to read about one of thesimplest frequency synthesizers everdesigned.Why a synthesizer?You may be wondering why a synthe-sizer has been used in this project, andeven, for that matter, what it is. Basi-cally. a frequency synthesizer is a cir-cuit that takes a single frequency from aqua rtz crystal, and uses it to generatemany others, each with the accuracyand stability of the crystal. In the Pro-

0gramma I , a single color-TV crystal isused to generate 9990 different fie- +quencies. In other words, you replace9990 crystals with one single-crystal afrequency synthesizer. (Now you know ;why they are found in CB radios, and


2/5

by only three IC'alother- places where lots of crystals once reference oscillator input is pin 14. 'Thewere used.) Through frequency synthe- output is pin 13. It drives resistors R19sis you gain accuracy and save money. and R20, and capacitor C6, forming aAs you can see from Fig. i , the syn- network known as a loop filter. Basi-thesizer consists of four basic pam. cally, this filter does nothing more thanThe first is the VCO (Voltage Controlled clean up the VCO control voltage.Oscillator. 11 simply generates a signal Other phase-detector circuitry includeswhose freq~lency can be varied by transistor Q I , which connects a LEDchanging a DC control voltage. In this to the error-detecting circuitry in IC2.project, the VCO freqtlency can be If something goes wrong with the cir-swept from i kHz to 1 MHz. cuitry, and the frequency is off, theThe next section is the re fermce &EDwill light.o,scillaiov. It provides a stable, fixed The V@O poi-tion of IC2 is simplereference-frequency. I n the Programma and straightforward. The DC control1, a single IC , combined with a 3.579545- voltage is applied to pin 9. Resistor R17MHz coiior-'TV crystal. does the job. and capacitor CI set the maximumNext, a programmable divider takes operating frequency of the VCO. Theover , dividing the VCO frequency. The squarewave output signal appears onexact divisor is selected by the front- pin 4, ready for use elsewhere. Thepane! switches. In this design, another reference osciilator circuit consists ofI43 peifiims the entire task , and can [C3, and it contains all the devices re-divide the VCO signal by up to iO,iKN. quired to excite TV crystal XTALI andThe final section is the phase detec- to produce a 100-Hz reference signal.t or . It compares the signal from the The balance of the circuitry on thisreference osciliator with the signal from board consists of five decade-dividers,the programmable divider. En operation. IC4--1438, that simply divide down thethe phase detector outputs a CC voltage output signal. giving a symmetricalcomparable to the difference of the two waveform. Since the outputs of thesesignals. The greater the difference, the IC's are all at CMOS levels, with a 10-greater the DC output is. On the other voit swing, buffer IC9 has been includedhand, if the two input signals are the to convert the voltages to TTL-com-same frequency, the DC output doesn't patable values.change. Since the DC voltage drives the The power requirements of this cir-VCO, it can now adjust the frequency cuit are provided by ICI0 and Zeneruntil the signals on the phase detector diode Dl. These components provide ainput are the same. The result is a VCO well-regulated 10 volts for the synthe-output-frequency equal to the product sizer, and 5 volts for IC9, which is usedof the divisor of the programmable to drive 5-volt 'ITL devices. Power todivider and the reference osciIlator fre- the PC board is supplied by a 14-voltquency. In this project, the VCO and surplus battery charger. Not much cur-phase detector sections are all included rent is required (about 10mA K), oinside one easy-to-obtain IC. That the entire unit can be battery-poweredtakes care of the basics. if desired.Refer to the schematic diagram inFig. 2 for the circuit details. The pro- C~ti~l~trltrudiongrammable divider is IC1, whose divisor Wow that you know how the Pro-is selected by the front panel switches. gramma I works, let's put one together.It is a single LSI IC, and it costs less, One important reminder is in order ifworks better, and is easier to use than y o ~ i re considering breadboarding theany other divider scheme. A unique project-the output signal will be noisyfeature of this circuit is that setting the unless you are careful. Like most otherswitches to 0000 I-esuiis n a divisor of frequency synthesizers, this one has a1 0 , ~ .hat saves the cost of an extra high loop-sensitivity, and is suceptibleswitch. The phase detector and the to noise pickup. So if you wish to get aVCO are in IC2. The divider input of high-quality signal from this project, bethe phase detector is pin 3, and the sure to use a PC board. If desired, you

can buy one, together with assembly in-structions and troubleshooting hints.from the supplier indicated in the partslist. Or you can "roll your own" usingFig. 3.Another important reminder concernsthe quality of the parts you use. Itshouldn't be necessary to remind youto use top-quality components, but ifthe urge to use cheap substitutes isove~powerdng,you may wind up withproblems. Generally, the quality of theoutput signal will suffer, and frequentservicing may be required. Play it safe,and save time and money in the long

PARTS LISTAQ I esistors 1/4 watt, 5% unless &hew isenoted.R1-R16, R1%-100,000 ohmsR17, R18-10,000 ohmsR20, R22, R23-2,200 ohmsR21-22 megohmsR24-47 ohmsR25-10,000 ohms l~nea r aper pot (car-bon) with SPST switchCapacitorsC 1 4 7 pF micaC2, CB-O.1 pF discC3, (29-33 pF mlcaC4-10 pF micaC5-6 to 20 pF trimmer (E.F. Johnson275-0320-005 or equivalent)C W . 7 pF, 16 volts, tantalumC7, C10--10 IF, 16 volts, tantalumC11-220 nF, 25 volts, electrolyt~cSemiconductorsDl-5 1-volt, 1-watt Zener diode (IN4733or equivalent)D2, D3-IN4148 or IN914Q1-2N3906IC1-CD4059AE CMOS divider (RCA)IC2-CD4046 CMOS PLL (RCA)IC3-MM5369EST CMOS oscillator (Na-tional)IC4-IC8-MM74C90N CMOS counter (Na-tional)ICS-CD4050 CMOS hex buffer (RCA)IC10-MC78L05 5-volt regulator (Motorola)LED1-,200-inch d~screteLEDSl-S4--BCD thumbwheel or lever-typeswitches (C&K 3321 10000, EECO 1800Series, or equivalent)S5--6-position, single-pole rotary switchSGS PST switch (mounted on R25)XTAL1-color-TV crystal, 3.579545 MHz,32 pF parallel-resonant, HC-33 caseJILjack to match connector from powersource usedJ2-RCA-type jackJ3-BNC connectorMiscellaneous: PC board, 14-volt DCpower supply or battery elirn~nator, ne8-pin IC socket, five 14-pin lC sockets,two 16-pin IC sockets, one 24-pin ICsocket, enclosure, knobs, solder, r~b-bon cable, etc.

PC boards are available. Order partSCG-I. Price, postpaid in USA, $IQ.QO;CaliRrnia msldents add 6% tax. Foreignorders please add $3.00 or shipping andhandling. Order from: Technic 0 Setvices.Box 2 0 m ~ rangehurst, Fullerton, GA92633.


3/5

2-PROGRAMMA 1 has a frequency range of from 0.01 H z to 1.000 MHz . BCD panel-mountare used for exact selection of pulse frequency.

tz4318 INCHES4S F O I L P ATE RM for the Programma1.See parts list for supplier if you prefer not to make yourPC board.

using top-quality parts. This is or tantalu m) and value specified. This on the PC board. A good place to startially imp ortant with respect to the will insure the best possible signal is with the IC socke ts. Begin by install-stability and purity at a small additional ing a 24-pin sock et at 1C1, then an 8-pinfo r quality IC's is obvio us, the cos t. unit at IC3. Check to be sure all pins arecitors should be the type (e.g. mica Refer to Fig 4 as you install the parts solde red in place on the sockets-


4/5

ove r very carefully for errors. T hen setthe board asid e for a while.

FIG.&OBSERVE POLARITIES shown in parts-placement diagram. Beaware that 78L05pinouts maybe shown differently from manufacturer to manufacturer. Orientation given here is correct tor ailversions.especia lly on the 24-pin one. Continue by you r hand on the screwdriver! Con-by installing 16-pin sockets at IC2 and tinue with C6, a 4.7 p F tantalum, andIC9 location s. Finish up the socket in- just ab ove it install a 47 pF mica at C1.stallation by adding 14-pin sockets to Move up the board and install a 0.1 pFIC4--1C8 locations. This board has one disc at C2, and anothe r at C8, at thejum per wire, which you can insert next. left. The n install a 33 pF mica at C9.Lo cate it in Fig. 3 first (between IC4 Mount anot her 10 p F tantalum at C10,and C1 I ) , then install it on the board. A below IC9. Finish up the c apacito rspiece of bare wire cut from a resistor with a 220 p F electrolytic at C11. Sto pwill work fine. for a m oment, and check your capacitorNo w you are ready for the resistors. installation. Correct any mistakes youStart with the IOOK units, placing 16 of may find and then continue with thethem around ICI (Rl-R16). After that, construct ion.install R21, 22 megohms, next to the By now your clrcuit board will beIC3 socket. Then mount a 10K resistor nearly complete and will look like theon either side of IC2. Note that, while one in Fig. 5. There are just a few partsthe leads of R 17 are simply bent and in- to go. so let's finish up the board.sert ed in the board , R18's leads must be Mo unt crystal XT AL l first, pressingleft longer (about %-inch) to cover the the case down firmly against the boarddistance between the holes. Ne xt, in- before soldering the leads. Then installstall 2.2K resistors at R20 and R22, and ICIO, a 78L 05 regulator next t o C11.a lOOK resistor at R19. Move over to (Note : The 78L05 pinout given by somethe o ther edge of the board and mount a ma nufac turers ma y dlffer from that47-ohm resistor at R24. And finish up shown here. T o the best of our knowl-with R23, 2.2K. Be careful not to con- ed ge , our pinout holds true for all ver-fuse the location with that for D3, just sions of the 78LO5-Editor).below it! Finish up with the IC's, starting withT he diodes are next, and the installa- IC I. No te that the foil side of the boardtion will go quickly . Be careful to install an d Fig. 4 indicate th e orientation ofthem correctly, and double-check eac h IC. Use them to guide you. Afteragainst Fig. 3 afterwards. Start with D l, the IC 's ar e installed, check the boarda IN4733 5.1-volt Zener diode, andthen install IN4148 diodes at D2 andD3. That's it .Th e next step is to install the capaci-tors. You can start with C7, 10 pF.Orient it as shown in Fig. 3. Then in-stall a 33 pF m ica capacito r at C3, and a10 p F m ica capacitor at C4. The trimmeris next , s o examine C5 and note that theground terminal is probably marked insom e way. If there's no arrow or paintdo t, then trace out the pin that attachesto the adjustment screw. Install it sothe ground terminal faces the edge ofthe board. If the trimmer is reversed, FIG. are con-the project will work, but be nected to PC board by ribbon cable. Note ww erto calibrate due tc capacitance added jack, J1, at back of e-nclosure (upper leftj.

Preparing the caseAlthough the original version of thisproject was built in an old meter case,you ar e welcome to use any suitable en-closure. It should be metal, though, toprevent radiation of stray signals thatcan interfere with your tests. As far as

the layout is concerned, you can ex-ercise your judgment in the matter, orduplicate the box layout shown in thephotos. Here are a few helpful tips ifyou decide to "roll yo ur own:" First,be sure to locate the ERRO R LED a ndF R E Q U E N C Y switches close together.This is important because they are usedtogether. A lso, the outpu t jacks andL E V E L pot should be located close to-gether. In fact, they should be positionedcloser to one an other than they are onthe prototype (see Fig. 7). since longleads degrade the sh ape of the signal athigh frequencies. AII signal-carryingleads, for that m atter should be kept assho rt as possible. T he rest of this part isstraightforward.Final assembly

After you have the enclosure pre-pared, you're almost done. Probablythe best place to start is to wire theboard to the F R E Q U E N C Y switches.Refer t o Fig. 2 (schematic) and Fig. 6for details.Start by wiring all the common pinsof the switches together with a piece ofbus wire. Then attach a short piece ofstrande d wire to it . This is the "COM"lead to the circuit board. N ext, you canwire the switches themselves, startingwith S I . Note that S1 is the MSD (MostSignificant Digit), and that it is theswitch section on the far left of thepanel as you view it from the front Useshort pieces of four-conductor ribboncable for the connections. You canattach the end s of the switches first. Infact, it might be a good idea to solder alength of cable to each switch first, andthen to the circuit board later. This iseasierif you have mounted the switchesin the box already.After the wires are attached to theswitch es, connect the cable from S1 tothe holes on the board. N ote that someBC D swi tches a re coded " 2 4 8" andthat corresponds with the "A B C D"marked on the board. In the same man-ner, wire the remaining switches.Switch S4 will be the section on theright when viewed from the front.Finally, conne ct the "COM" wire, andyou are through with SI-S4.

Now for switch S5. Prepare a shortlength of six-conductor ribbon cableand connect one end to the fixed con-tacts of S5. Then attach a single pieceof wire to the wiper terminal. Connectthe o the r ends of the ribbon-cable wires


5/5

I'components. ResistorR25 is a part of the onloff switch,be kept as short as possibleto avoid difficulties at high frequencies.put jacks and "level" pot should be

the "E" through "J" outputs, andsingle wire from the wiper to "IN,"6. Nex t, th e L ED may be installed.

ctors, and the power leads. Don' tbox. A

jack , J 1 (on the back of the box)are finished.

Although this project should work

reasonably well without any calibra-tion, you might wan t to m ake a simpleadjustment for the best frequency ac-curacy. T o do this you'll need an ac -curate frequency counter and an x 10oscilloscope probe. Supply 14 volts DCto J1, then rotate the LEVEL pot to turnon the power. At this point there's noneed to set any of the switches on theproject. Connect the probe to the count-er, and clip its ground lead to the pulsegenerator. Then, carefully touch theprobe t o pin 7 of the MM5369EST (IC3).You should get a reading of 3,579,Sx x(-=variable) Hz. Adjust the trimmer sothat you get exactly 3,579,545 Hz andyou are all set. D isconnect the counterand you can close up the box.Operation

Operating the Programma 1 is a snap!Simply set the frequency you want onthe thumbwheel switches, and watchth e ERROR L E D . It will blink about fouror five times, then go out. W hen it does,you a re locked on frequency. Sw itch S5selects the frequency you get out. Forexample, on the I M H Z range, you'll getan output f rom about 908 H z to exact ly1.000 MH z. Switch to th e 1 0 0 - k ~ z angeand you' ll get a tenth of that or 90 Hz to100 kHz. The rest of the ranges work

in the same manner. If you would likean adjustable output instead of the7T L-l evel signal from 53, simply use 52 ,and adjust the L E V E L control for thevoltage you want. There's nothing tousing this project!Here are a few tips to help you getthe most out of your project. First, dueto the design of VCO and divider cir-cuits, switch positions from 0001through 0009 will be inoperative. TheERROR light will come o n as a reminderthat these numbers are invalid. Notethat the setting of is O K; in fact itwill give you 1.000 MHz, but watchthose oth er settings. As far as the out-put signal is concerne d, it is a constant-amplitude squarewave with a 50% dutycycle. However, if you start to load itdown, the amplitude will change. Also,the waveform quality will tend todeteriorate as the frequency goes up.So , for best results when are in-terested in waveform quality, use avery light load, and watch out for theeffects of coaxial cables at the higherfrequencies. Finally, some degradationof the squarewave will be noted at theadjustable ou tput ( Jl ) at high fre-quencie s. T his is normal where a simplepot-attenuator is used:Some uses for the Programma 1

There are a great many uses for thispulse generator. Although it was de-signed for operating digital circuits, itdoe s well in other areas, too. Here are afew things that can be done with it:checking 7TL divider circuits, decimal-counting uses (w hy not make a timer?),general logic-troubleshooting, andmuch more .In the analog area, it can be used foramplifier squarewave-testing, electronicmusic (it generates a wild glide tone!),A M radio alignment, and more. Howabout using it as a short-wave radiomarker-generator? (The harmonics gowell in to the H F spectrum.) Or as aprogrammable sinewave generator?(A ctive filtering can change th e square-wave to a sinewave.) There arenumerous uses for the Programma I .How many can you think up? R-E

synthesized pulse generator

Documents