AudioXpress magazine

AUGUST 2003US $7.00 Canada $10.00

THE AUDIO TECHNOLOGY AUTHORITY

TEST AMP BEHAVIOR WITH A REACTIVE LOAD BOX

www.audioXpress.com

Continuing Audio Electronics, Glass Audio & Speaker Builder

2 audioXpress 8/03 www.audioXpress.com

audioXpress (US ISSN 0004-7546) is published monthly, at $34.95 per year, $58.95 for two years. Canada add $12 per year;overseas rates $59.95 per year, $110 for two years; by Audio Amateur Inc., Edward T. Dell, Jr., President, at 305 Union St., POBox 876, Peterborough, NH 03458-0876. Periodicals postage paid at Peterborough, NH, and additional mailing offices.

POSTMASTER: Send address changes to: audioXpress, PO Box 876, Peterborough, NH 03458-0876.

F E AT U R E S

A SIMPLE, HIGH QUALITY AM TUBE RECEIVERTune in to clean, quality sound with this easy-to-build tube-basedAM receiver.By Scott K. Reynolds.................................................................... 6EXTENDING THE 30 TRACTRIX HORNExperience the high-efficiency sound of horn loudspeakers withthis detailed construction project.By Robert Roggeveen ................................................................20IMPROVING DYNACOS FM-5 AND AF-6TUNERSHeres a modification to the power supply to enhance the perfor-mance of your Dynaco tuner.By Daniel Dufresne ............................................................28ALPHA TRANSMISSION LINESAn interesting look at transmission-line construction, with detailson how to build your own unit.By Rick Schultz ................................................................34SYSTEM POWER CONTROLHeres a simple circuit to control the on/off sequence of youraudio system.By Frank S. Thomas III ....................................................46BUILD A REACTIVE LOADAudio constructors will enjoy building this simple load box designto test power amps.By Dick Crawford..............................................................48CARE AND MAINTENANCE OF DUKANEIONOVAC TWEETERSHeres a guide to restoring classic high-frequency speakers,which include no moving parts . . . only the air itself.By Daniel Schoo ..............................................................50

C O N T E N T SVOLUME 34 NUMBER 8 AUGUST 2003

page 20

page 34

audioXpress August 2003 3

LEGAL NOTICEEach design published in audioXpress is theintellectual property of its author and is of-fered to readers for their personal use only.Any commercial use of such ideas or de-signs without prior written permission is aninfringement of the copyright protection ofthe work of each contributing author.

SUBSCRIPTION/CUSTOMERSERVICE INQUIRIESA one-year subscription costs $34.95; a twoyear subscription costs $58.95. Canadaplease add $12 per year. Overseas rate is$59.95 for one year; $110 for two years. Tosubscribe, renew or change address write tothe Customer Service Department (PO Box876, Peterborough, NH 03458-0876) or tele-phone toll-free (US/Canada only) 888-924-9465 or (603) 924-9464 or FAX (603) 924-9467. E-mail: [email protected]. Oronline at www.audioXpress.com

For gift subscriptions please include gift recipi-ents name and your own, with remittance. Agift card will be sent.

EDITORIAL INQUIRIESSend editorial correspondence and manu-scripts to audioXpress, Editorial Dept., PO Box 876, Peterborough, NH 03458-0876.E-mail: [email protected]. No re-sponsibility is assumed for unsolicited manu-scripts. Include a self-addressed envelopewith return postage. The staff will not answertechnical queries by telephone.

CLASSIFIEDS & WEB LISTINGSContact Nancy Vernazzaro, Advertising Department, audioXpress, PO Box 876, Peterborough, NH 03458, 603-924-7292,FAX 603-924-6230, E-mail [email protected].

Printed in the USA. Copyright 2003 by AudioAmateur Corporation. All rights reserved.

IN EVERY ISSUE

CLASSIFIEDSAudio-related items forsale or wanted .......................... 70

AD INDEX .......................... 71

YARD SALEFree classifiedsfor subscribers ........................ 71

CO

DE

Editor and PublisherEdward T. Dell, Jr.

Regular ContributorsEric Barbour Erno BorbelyRobert Bullock Richard CampbellJoseph DAppolito Vance DickasonBill Fitzmaurice Gary GaloCharles Hansen G.R. KoonceNelson Pass Richard PierceReg Williamson

Vice PresidentKaren Hebert

Dennis Brisson Assistant Publisher

Marianne Conway Editorial Assistant

Amity Palmer Editorial Assistant

Tina Hoppock Graphics Director

Amy Wyman Production Assistant

Laurel Humphrey Marketing Director

Kelly Bennett Customer Service

Advertising Department

Strategic Media Marketing1187 Washington St.Gloucester, MA 01930Peter Wostrel

Phone: 978-281-7708Fax: 978-283-4372E-mail: [email protected]

Nancy Vernazzaro

Advertising/Account Coordinator

T H E S TA F F

The peculiar evil ofsilencing the expression

of an opinion is, that it is

robbing the human race;posterity as well as the

existing generation; thosewho dissent from the

opinion, still more than those who hold it.

JOHN STUART MILL

D E PA RT M E N T S

SHOWCASEAn Italian THORBy Atto Rinaldo..............................................................................58XPRESS MAILReaders speak out ........................................................................62NEW CHIPS ON THE BLOCKAKM AK4395 and AK4554By Charles Hansen ........................................................................72

Articles and letters identified with the following symbols:

page 6

solid state tubes speakers

page 58

page 50

Pleasure just knows no compromises

8 0 1 - 6 2 1 - 1 5 0 0 w w w . w b t u s a . c o m

Every WBT product has been designed and ref ined for t rouble f ree rel iabi l i ty and superior sonic performance. The formula is s imple:develop innovat ive designs and bui ld them with the f inest mater ials .

The construction of the 330 is built-to-last andfeatures an elegant polycarbonate rear housingavailable in either black or clear with a durableNylon-Tough front housing. Additionally, thereis a dust seal to protect the wiring chamber.Marked with Audio Grade and Blue Star.

AUDIO GRADE POWER CONNECTOR

801-621-1501 www.wattgate.com

Tr iple plated contacts

Adaptable cord c lamp insert

Terminal funnels andholes

Wire(st rand conductor)

This construction project beganwhen I read a letter from VinceRoberts in the February 2001issue.1 Mr. Roberts requested an

article on an AM tuner using vacuumtubes, and reading his letter inspiredme to build one. I tried to come up witha design that would be easy for thehome constructor to build and also givegood audio quality. Both of these goalsmeant that I had to keep the radio fre-quency (RF) circuits as simple as possi-ble, but a large, efficient antenna morethan compensated for the low RF gainand resulted in a receiver that has sur-prisingly good sensitivity.

My wife was at first skeptical of thisproject, but in the end she liked the

radio so much that I had to move itfrom my workshop into the study. Mywife is a National Public Radio (NPR)enthusiast, and we live in a fringe re-ception area where the local NPR AMstation (WNYC) usually comes in with a

fair amount of static and noise. To evenmy surprise, this new AM receiverbrings in WNYC with consistentlylower noise than any other radio in thehouse, and it has a rich and detailedsound the other radios lack. It made my

You, too, can re-discover how good

AM radio can sound when it is

done right. By Scott K. Reynolds

A Simple, High Quality AM Tube Receiver


PHOTO 1: Front view of the radio receiver. The RF circuits are on the left, with tubes V1and V2 at the left rear.

FIGURE 1: Schematic of the tuner section ofthe AM receiver.

G-2109-1

wife appreciate (and me remember)how good AM radio can sound when itis done right.

BACKGROUNDWhen I read Mr. Roberts letter, Ithought about what sort of receiverwould be best for home construction. Ieventually settled on a tuned radio fre-quency (TRF) receiver with a singlestage of RF amplification (Fig. 1). Thereare only two tuned circuits in this re-ceiver, so construction and alignmentare easy, as I will describe later. So thatthe trade-offs I made in arriving at thisdesign will be clear, Ill first discusssome other receiver circuits that I con-sidered but did not pursue.

The simplest receiver capable of giv-ing satisfactory results in everyday op-eration is undoubtedly a regenerativedetector followed by audio amplifica-tion. In its simplest formas it was firstbuilt in the 1920sthe regenerative de-tector circuit consists of a triode grid-leak detector with positive feedbackaround it. The positive feedback greatlyincreases the gain of the stage and theselectivity of the tuned circuit. Unfortu-nately, it also increases the distortionand noise and makes the detectorprone to oscillation.

These simple regenerative circuitscan work amazingly well at pulling indistant signals, but their poor audio

quality makes them better suited to ex-perimentation than pleasant listening.The early editions of the Radio Ama-teurs Handbook described these regen-erative receivers in detail, but the morerecent editions discuss them only inpassing.2 Antique Electronic Supplyhas books on regenerative receivers ifyou would like to experiment withthem.

Most radio receivers are superhetero-dyne3 circuits in which the incomingRF signal is first converted to an inter-mediate frequency (IF) signal beforebeing amplified and detected. Super-

heterodyne circuits are popular be-cause they can have high gain and se-lectivity while also being stable andeasy for you to tune.4

Superhets need to be carefully de-signed to give good audio quality for AMsignals, however. If the IF transformersare too highly selective, the high-fre-quency components of the received sig-nal can be attenuated. Also, the frequen-cy conversion process can generatehigh intermodulation distortion.

When I considered building a super-heterodyne receiver and writing aboutit for audioXpress, I was concerned

PHOTO 2: Rear view of the radio receiver.

FIGURE 2: Schematic of theaudio power amplifier section ofthe AM receiver.

G-2109-2


about the complexity of the RF circuitryand the difficulty that the home con-structor might have in aligning a newly

built radio (aligning several stages in anewly built superheterodyne radio ismore confusing than re-aligning an al-

ready working radio). I was alsoconcerned about the difficulty offinding parts, especially the tube-compatible IF transformers,since I dont think they are beingmanufactured anymore. AntiqueElectronic Supply has a few sur-plus IF transformers, but its dif-ficult to know how long theyll beavailable. I considered windingmy own transformers, but I de-cided to follow Mr. Roberts ad-vice to start simple.

A TRF receiver offered a sim-pler alternative . . . if I could getthe required RF gain in a singlestage. Multiple-stage TRF re-ceivers arent so simple any-more, because of the problem ofaligning multiple-tuned stagesand having them track togetherover the band as the receiver istuned. A high gain, multi-stageTRF receiver can be unstabledue to unintended feedback.

Finally, TRF receivers suffer

from variations in gain and selectivityas you tune from one end of the band toanother, a problem which becomesworse as the number of stages increas-es. Back in the 1920s and 30s, design-ers came up with elaborate solutions tothese problems, but I didnt want any-thing elaborate.

ANTENNA REQUIREMENTSHaving decided on a single-stage TRFdesign, I wanted to get the most perfor-mance I could out of it, and the antenna(component L1 in Fig. 1) seemed likethe best place to start. I decided to buildthe largest loop antenna that would stillbe practical to receive as much RF sig-nal as I could (without setting up anoutdoor antenna). An amateur canreadily build a better antenna than amanufacturer can, not having the samesort of space, shipping, or manufactur-ing limitations.

The RF signal received by a loop an-tenna is proportional to both the num-ber of turns and the area, so it wouldseem that you want a large loop areaand lots of turns. But in order for theantenna to resonate with the 365pF tun-

PHOTO 3: Distant rear view of the radio, showingthe construction of the antenna L1.


USA: 153 Mountain Home Road Port Angeles WA98362 Washington tel 360-452-9373Great Britain: 29 Drake Avenue Penkridge Stafford ST19 5UA tel 0 1785 711232Australia: POBox 307 Surry Hills NSW 2010 tel 02 9660 1228Scandinavia: Christiansgade 1 DK7500 Holstebro tel +45 97410863

DIYCABLEAudio Atmosphere

Soundlabs Group pty.L C Audio Technology

www.diycable.comwww.audioatmosphere.com

www.soundlabsgroup.comwww.lcaudio.com

+Out

+Out

-Out

-Out

1 23

+60V

-60V

PowerGND

10.000uF63V

10.000uF63V

HiFi Digital Amplification, that simply works, and sounds... GREAT!

Predator SE 250 Watts RMS in 8 OhmsStereo Power Amplifier

DIY ModulesSE Version w/Black Gate VK

as shown:

Efficiency ~ 96% =environment friendly

Special Edition Black Gate VK19 Watts Total Loss 500 Watts RMS Sound!

Predator SE 2 x 250 Watts RMS

SIMPLE!

Audio Power Amplifier using New advancedTechnology to produce 400 Watt RMS (4 Ohms)on a Credit Card size Footprint !

Available as a ready to play audio appliance,or modules for DIY projects. Almost no heatsinking is called for, a thin steel sheet metalfloor of the enclosure will do! Build as monoblocks, stereo, 5 ch, 7 ch in one enclosure.200 W/8 400W/4 650W/2 RMS cont. pwr.5 - 150.000 Hz +/- 3dB THD 0.02%@10WDamping Factor 130 Max. Current 62 Amp.

ZAPpulse 2.0

EMC Optimized 4 layerPCB

TEFLON/SilverWire

ing capacitor C1A over the entire AMbroadcast band (5401600kHz), you areconstrained to make the loops induc-tance 238H. The loops inductance isapproximately proportional to thesquare of the number of turns and thesquare root of the loop area, so select-ing the antenna size gave me an inter-esting mathematical and experimentalproblem to solve.

I settled on a 24 24 (61cm 61cm)square wooden frame, with nine turnsof AWG 22 enameled magnet wirewound around the outside, with theturns spaced (3.18mm) apart. Thespacing is important in order to mini-mize the self-capacitance of the anten-na. If the self-capacitance is too high,the antenna cannot be tuned over theentire AM band.

My radio tunes from 560 to 1570kHz,which should be adequate for mostreaders. But some of you may wish toreceive a station in the extended AMband from 1600 to 1700kHz (these aremostly low-power local stations, themost interesting kind). To do that,youll need to take a turn (or half aturn) off L1, which may prevent youfrom getting stations below 600kHz.You can also tap L1 to make changing

the number of turns easy. For the antenna to tune the entire

5401700kHz extended AM band with-out changing taps appears to require asmaller loop area, but, of course, thatresults in a less efficient antenna and aless sensitive receiver. The size of theantenna I chose is just one of those in-evitable engineering compromises. Illgive more details on antenna construc-tion and setup later.

CIRCUIT DESCRIPTIONThe main tuning capacitor C1 is a dual-ganged 365pF air-variable capacitor,which used to be standard in AM radios.But these air-variable capacitors are be-coming harder to find. RF Parts Co. andAntique Electronic Supply both havesuitable parts. The fine-tuning capacitorC2, which is a 33pF air variable, is morereadily available. But it can be tough tofind one with a shaft for a knob (most

PHOTO 4: Underside view of the chassis. The RF wiring is at the right rear.

FIGURE 3: Power-supply schematic for the AM receiver. G-2109-3


have slots for screwdriver adjustment).Mine came from RF Parts Co.

All of the RF gain in the receiver (2030dB) comes from remote-cutoff pen-tode V1, a 6BA6 or 5749/6BA6W. Thequiescent cathode current in V1 is setto a relatively high 14.4mA by the valueof R1. The high bias current is used tomaximize the transconductance of V1,giving high gain and low noise. Raisingthe value of R1 lowers the transconduc-tance and the gain, so you couldachieve RF gain control by inserting avariable resistor in series with R1. Illgive more details on this later.

The output of V1 is transformer cou-pled to the detector circuit by T1, a slug-tuned RF transformer. T1 is AntiqueElectronic Supply part number P-C70-RF (Im not aware that these RF trans-formers are available from any othersource, but they arent surplus, so theyshould continue to be available).

T1 is designed so that the primarycircuit should resonate at a frequencybelow the AM band, but if the associat-ed circuit does not have enough capaci-tance, some additional capacitanceneeds to be added to reduce the prima-

ry resonant frequency. That is the pur-pose of C5, a 15pF dipped mica capaci-tor. The secondary of T1 resonates withC1B and C2, forming a second tunedcircuit at the same frequency as the L1-C1A combination. Because of its largephysical size, the antenna L1 has moreself-capacitance than the secondary ofT1, so you need to add the additionalfine-tuning capacitor C2 in order tomake the two tuned circuits track prop-erly over the AM band.

The detector circuit is a classic diodeenvelope detector used in countlessAM radios. To get high sensitivity, use a1N34A (or 1N60A) germanium point-contact diode (D1), which is often usedin crystal radios and is 10100 timesmore sensitive than a vacuum tubediode. These germanium diodes alsohave the advantage of being more lin-ear than a vacuum tube diode at thelow RF signal levels found in this re-ceiver. Note that a standard silicondiode will not work for D1.

A cathode follower (V2) isolates thediode detector from the second tunedcircuit, thereby preventing the diodefrom loading the tuned circuit and pre-

serving high selectivity. Tube V2 (a 6C4or 6100/6C4WA) is biased at a fairlyhigh 10.8mA to give high transconduc-tance. A cathode follower can be unsta-ble at high frequencies when it has acapacitive load and a tuned circuit atthe grid, which is the situation in Fig. 1.Grid stopper R4 prevents any such in-stability. You should install it immedi-ately adjacent to the tube socket.

The output of the diode detector goesthrough a low-pass pi-section filter con-sisting of C10, L3, and C11, which re-moves the residual RF carrier. This fil-ter has an audio bandwidth of about6.6kHz, which is adequate for AM sig-nals. You can increase the audio band-width by reducing the values of C10and C11, if you want to experiment.With the RF carrier removed, the re-maining signal at point A on theschematic (Fig. 1) consists of the audiosignal and a negative DC component(there is also some second harmonicdistortion). Capacitor C12 blocks theDC component and passes the audio toV3A, the first audio amplifier stage.

The amplitude of the DC componentis proportional to the strength of the RF


Clean AC power allows audio and videocomponents to perform better. The cleanerand more perfect the power, the better thesound and the better the picture. Theimprovements you experience areimmediate and astounding.

The ExactPower EP15A is a single box thatperforms seven essential corrections,creating the most perfect AC poweravailable for your audio and videoproducts. Other products on the market do one or twoor, at most, three corrections. Even in combination,these products are totally outperformed by the EP15A.Because it does multiple, essential corrections for the

cleanest, purest power, only the EP15A canprovide maximum high performance andincreased longevity from your audio andvideo components. All units are available insilver or black, are rack-mountable, andfeature surge protection that can handlecurrent peaks up to 25,000 amps.

Our products are smaller, lighter, cooler,more efficient, more technologicallyadvanced, and far more affordable than

youll find elsewhere. Add to that our proven trackrecord of reliability and you will see that ExactPoweroffers the finest choice in power processing/management products.

The ExactPower EP15A allows audio and video equipment to operate at peak performance by providing clean, non-current-limited AC power.

The ExactPower EP15A uses a unique,patented subcycle AC voltage regulatorthat provides a constant 120 volts (0.5%) inaddition to instantaneous waveformcorrection. It also provides up to 1800 wattswithout any current limitation. In fact, theEP15A actually utilizes current storagewhich avoids any energy starvation-relatedartifacts from corrupting the powerconditioning process. The EP15A is the mostaccurate, most efficient, most affordable, mostflexible, and most adaptable powerprocessing/management solution available.

For the absolute ultimate in powerconditioning, couple the EP15A with our highperformance SP15A Symmetrical Power unitproviding balanced current and digitalfiltration, plus a host of other features.

Visit our website to find out additionaloptions and technical details on our full lineof power processing/management productsthat can dramatically improve your audio andvideo systems performance.

Atlantis Power Quality Systems, Inc.9411 Winnetka Avenue, Chatsworth, California 91311

SALES & INFORMATION (866) 392-2822 (866) EXACT ACTECH SUPPORT (800) 773-7977 U.S. only (818) 886-5364 outside U.S.

FAX (818) 886-0504 WEB www.exactpower.com

signal, which makes it useful for operat-ing a tuning indicator. In my area, thestrongest RF signals produce about 1.5V at point A. But the 6E5 tuningindicator tube that Mr. Roberts request-ed requires about 8V for complete eyeclosure.

Therefore, the DC component is am-plified by IC1, an LF351N JFET input opamp configured for a non-inverting gainof 5.6. An op amp with a JFET input isessential in order to provide highenough input impedance so as not toload the diode detector circuit. A stan-dard 741 op amp will not work here.

I think the 6E5 tube is quite attrac-tive, but it is becoming rare and expen-sive. You could also use a 10V full-scaleanalog meter movement at the outputof IC1, or a 1mA meter with a 10k se-ries resistor.

Keep in mind that the output of IC1swings from zero to about 8V, so wirethe meter with the correct polarity. Youcannot wire an analog meter directly topoint A on the schematic because itwill load the diode detector. Digital me-ters will not work for tuning becausetheir response is not continuous. Somesort of tuning indication is essential forboth alignment and normal operation,

so dont be tempted to omit it.

AUDIO FILTERComponents R44R45, L8, and C42C43in Fig. 1 make up a passive low-pass fil-ter that I added to the output of thetuner to eliminate (or reduce) the an-noying 10kHz squeal that AM radiossometimes produce in an environmentcrowded with many stations. This10kHz squeal arises because AM sta-tions are spaced 10kHz apart, andalong with the desired station you mayalso be receiving a weak, distant station10kHz higher or lower in frequency(other local stations are usually 30kHzapart). The two signals beat togetherin the receiver to produce a 10kHz tone,and the only solution is to filter it out.This problem is less noticeable on amodern superheterodyne receiver be-cause the IF filters are so highly selec-tive, giving the receiver less bandwidth.

The filter has a 3dB point at approx-imately 5.5kHz, and it is down morethan 15dB at 10kHz. A tiny filter choke(L8) acts as an audio choke to producean LC filter. Capacitor C43 resonateswith L8 at 10kHz to produce a sharpcutoff characteristic. The filter is de-signed to work best with load imped-

ances between 50k and 100k (here itsees 470k in parallel with the 100kpotentiometer R18, or about 82k).

I dont recommend replacing L8 (aHammond part #154M), because otherfilter chokes may have different audiocharacteristics. I added a switch (SW4)to bypass the filter when it isnt needed,since I didnt want to unnecessarily re-duce the wide bandwidth that my re-ceiver has. You can build the tunerwithout the filter by eliminating R44R45, L8, C42C43, and SW4, taking theaudio output directly from coupling ca-pacitor C15.

AUDIO POWER AMPLIFIERThe output of V3A or the audio filter iscoupled into the audio power amplifier(Fig. 2) through selector switch SW1. Idecided to add an auxiliary input to thepower amplifier so that I could plug inmy FM tuner or CD player (or any linelevel audio signal). The power amplifieris a simple, classic design that manyreaders will have seen before. Tube V3Bis an input amplifier, and V5 is a long-tailed phase splitter. A pair of push-pull6V6s produces about 9W of audio power.

One slightly unusual feature is thatthe 6V6s have separate cathode bias re-


FIGURE 4: Optional test oscillator to supply signals for receiver alignment. G-2109-4

sistors. I did this to accommodate mis-matched tubes. With a single cathoderesistor, one tube can hog most of thebias current if the 6V6s are badly mis-matched. With separate resistors, youcan plug in any two working 6V6s andachieve satisfactory performance.

A 10W Hammond output trans-former (type 1608) couples the push-pull 6V6s to the speaker. About 15dB ofnegative feedback is applied from thespeaker output to the cathode circuit ofV3B. Note that the output transformerT2 is shown wired for an 8 speaker. Ifyou opt to wire it for 4, then R31should be about 750 instead of 1000.

If you wire it for a 16 speaker, thenR31 should be about 1.5k.

POWER SUPPLYThe power supply is shown in Fig. 3. Aconventional full-wave rectifier and pi-section filter supplies 325V for thepower amplifier stages (B+1). Gaseousregulator tubes V8 and V9 produce ashunt-regulated 255V supply for the RFstages (B+2).

I regulated this supply to provide lowAC ripple as well as a stable operatingpoint for tubes V1 and V2, so that linevoltage and temperature fluctuationswouldnt detune the radio. A regulated

supply is probably not essential, howev-er, as long as B+2 is in the 240270Vrange. An RC decoupling filter pro-duces 240V for the low-level audio am-plifier stages (B+3).

The 6.3V secondary on T3 suppliesheater power for tubes V4V7. Note thatone side of the 6.3V winding is ground-ed. However, you should run twistedpairs for the AC heater wiring insteadof running a single wire and groundingone side of each tubes heater at thesocket. Running twisted pairs preventsAC heater current from flowingthrough the ground connection, there-by eliminating a source of 60Hz hum.


TABLE 1COMPOSITE PARTS LIST FOR FIGS. 14

Q1 2N2907A or equivalent, PNP silicon bipolar transistorR1 180, W, 5%R2, R39 33k, 2W, 5%R3, R9, R12, R20, R31, 1k, W, 5%R41, R46R4 220, W, 5%R5, R10, R15, R24, R25 1M, W, 5%R6 470, W, 5%R7, R8 10k, W, 5%R11 33k, W, 5%R13, R27, R28 220k, W, 5%R14 10M, W, 5%R16 4.7k, W, 5%R17 22k, W, 5%R18 100k audio taper potentiometerR19, R35 100, W, 5%R21, R22 51k, 1W, 5%R23 62k, 1W, 5%R26 22k, 2W, 5%R29, R30 750, 3W, 5%R32 1.6k, 10W, 5%, wirewoundR33 2.2k, W, 5%R34 270, W, 5%R36 330, W, 5%R37 250 trimmerR38, R44 47k, W, 5%R40 33k, 1W, 5%R42 82, W, 5%R43 1k linear taper potentiometerR45 470k, W, 5%SW1SW3 Single-pole double-throw toggle switchSW4 Double-pole double-throw toggle switchT1 540-1600kHz slug-tuned RF transformer (Antique

Electronic Supply part #PC-70-RF)T2 10W audio output transformer, 8000 plate-to-plate pri-

mary with ultralinear taps, 416 secondary (Hammondpart #1608 or equivalent.)

T3 Power transformer, 250-0-250V @ 100mA, 6.3V @ 3A, 5V @ 2A (Hammond part #270CBX or equivalent.)

V1, V10 6BA6 or 5749/6BA6WV2 6C4 or 6100/6C4WAV3, V5 12AU7A or 5814AV4 6E5 tuning indicator tube (or substitute an analog meter

movement)V6, V7 6V6 or 6V6GT or 6V6GTAV8 0B2 gaseous regulator tubeV9 0A2 gaseous regulator tube

C1 Dual ganged 365pF air variable tuning capacitor (RFParts Co. Part #10295 or Antique Electronic Supply Part #C-V365-X3 or equivalent)

C2 33pF air variable tuning capacitor (RF Parts Co. Part#074-033 or equivalent)

C3, C4, C7, C8, C35 0.01F, 500V, disc ceramicC5 15pF, 500V, dipped mica, 5%C6, C9, C23, C41, C44 0.022F, 500V, disc ceramicC10 33pF, 100V, dipped mica, 5%C11, C36, C38, C39, C43 100pF, 100V, dipped mica, 5%C12 0.047F, 100V, filmC13, C16 220F, 16V, electrolyticC14 0.01F, 50V, filmC15 0.22F, 400V, filmC17C19 0.1F, 400V, filmC20, C21 470F, 35V, electrolyticC22 4700pF, 100V, film or dipped micaC24, C26 47F, 450V, electrolyticC25 100F, 450V, electrolyticC27C30 100F, 25V, electrolyticC31, C32 4700F, 16V, electrolyticC33 0.1F, 100V, filmC34 0.001F, 500V, disc ceramicC37 470pF, 100V, dipped mica, 5%C40 10pF, 100V, dipped mica, 5%C42 0.001F, 100V, filmD1 Germanium point-contact diode, 1N34A, or 1N60A, or

equivalent. Available from Antique Electronic Supply,RF Parts Co., or All Electronics.

D2, D3 1N4007, 1000V, 1A rectifier diodeD4, D5 1N5401, 100V, 3A rectifier diodeD6D8 1N4001, 50V, 1A rectifier diodeD9, D10 1N4735A, 6.2V, 1W zener diodeIC1 LF351N, JFET input op ampIC2 LM317T, adjustable voltage regulator (mount this part

on a heatsink or on the chassis, using a thermally conductive, electrically insulating gasket)

L1 Nine turns of AWG22 enameled magnet wire, spaced (3.18mm) apart, wound on the outside of a 24 24 (61cm 61cm) square wooden frame. See text.

L2, L3, L5 2.5mH RF choke, Hammond part #1535B or equivalent.

L4 Filter choke, 12H @ 100mA, Hammond part #193B orequivalent.

L6 100H molded RF inductor, 10%, Dale Vishay typeIM2 or equivalent. Mouser part #70-IM2-100

L7 10H molded RF inductor, 10%, Dale-Vishay type IM2` or equivalent. Mouser part #70-IM2-10L8 Filter choke, 2H @ 100mA, Hammond part #154M. (Be

careful about using a substitute for this choke, sinceother chokes may have different audio characteristics.See text.)

I included a 6.3V DC heater supplyfor tubes V1V3 to produce a complete-ly hum-free radio. A voltage doublerproduces about 11V DC from the 5Vheater winding on T3, and a standardslow turn-on regulator circuit using anLM317T produces a practically ripple-free 6.3V DC. The LM317T needs aheatsink, which I provided by boltingthe tab to a piece of aluminum anglestock, which, in turn, is bolted to thechassis. A thermally conductive washerprovides electrical isolation.

Op-amp IC1 requires positive and neg-ative supply rails. The positive rail isfrom the 6.3V DC heater supply, and a 12.4V DC supply rail is provided by ahalf-wave voltage doubler using one sideof the 6.3V AC heater winding. The volt-age doubler supplies approximately 16V, which is regulated down to 12.4Vusing a couple of 6.2V zener diodes.

Power transformer T3 is heavilyloaded, although not overloaded. Afterprolonged operation, the surface temper-ature of T3 reaches 5055C. Most powertransformers are rated for a 60C temper-ature rise from ambient, so the tempera-ture of T3 is not a problem, but youshould allow for good ventilation asshown in the open design of Photos 13.

CONSTRUCTIONI began construction of my radio withantenna L1. I built a 24 24 woodenframe using 1 clear pine trimstock (Photos 13). For the upright sup-

ports I used clear pine trimstock cut to 36 lengths. You should as-semble the frame without metal hard-ware. I used small nails to hold theframe together while the glue dried,and then I removed the nails.

I maintained the center-to-centerspacing of the turns by cutting slotswith a hacksaw in the four outside cor-ners of the frame, starting from theedge of the wood. This leaves room forten slots in the 1 wide trim stock.Prior to winding the antenna, I stainedthe pine frame, then applied a coat ofpolyurethane.

I wound the nine turns of L1 with asmuch tension as I could by hand, andwhen I finished the wire was fairly taut.I ran the wires from the two ends of L1down opposite sides of one of the up-right supports. You should do this orsomething similar in order to keep thewires separate and minimize their mu-tual capacitance. Do not twist the wirestogether. You can hold the turns inplace with spots of quick-setting epoxyat several locations along each side.

After the epoxy dried, I gave the en-tire assembly another coat ofpolyurethane to hold the turns perma-


TABLE 2PARTS SUPPLIERS

ANTIQUE ELECTRONIC SUPPLY (480) 820-5411www.tubesandmore.com RF transformer T1, Hammond components, tubes, airvariable capacitorsRF PARTS CO. (800) 737-2787www.rfparts.com Air variable capacitors, 6E5 tube, germanium diodeMOUSER ELECTRONICS(800) 346-6873 www.mouser.com Inductors, capacitors, semiconductors, metal chassis ALL ELECTRONICS(800) 826-5432www.allelectronics.comGermanium diode, various surplus itemsJAMECO ELECTRONICS (800) 831-4242 www.jameco.comClarostat conductive plastic potentiometer, capacitors,resistors, semiconductors

Chelmer Valve Company Ltd The Stables, Baddow Park, Great Baddow, Chelmsford Essex, CM2 7SY, England.

email: [email protected] ** tel. 44 1245 241 300 fax. 44 1245 241 309 ** www.chelmervalve.com

for High Quality Audio TubesEverybody in the audio tube business knows that the justly famous brand names of yesteryear likeBrimar, GEC, Mullard, RCA , Telefunken etc. etc. are scarce and often quite expensive.Although we supply all major brands as available (and we have many in stock) our policy is to offer arange of tubes, all new and mostly of current manufacture, the best we can find from factories around theworld, which we process to suit audio applications. The result CVC PREMIUM Brand. Our specialprocessing includes selection for low noise, hum & microphony on pre-amp tubes and controlled burn-in on power tubes to improve stability avoid tubes with weaknesses etc.

A selection of CVC PREMIUM Audio Tubes ******PRE-AMP TUBES POWER TUBES POWER TUBES cont. RECTIFIERS cont.ECC81 5.90 EL34G 8.30 6L6/ 5881 WXT 9.00 5Y3GT 4.80ECC82 5.90 EL34 (JJ) 8.50 6V6GT 5.50 5Z4GT 5.80ECC83 5.90 EL34(Large Dia) 11.00 6080 11.50 SOCKETS ETC.ECC85 6.60 EL84 5.50 6146B 11.00 B9A (Ch or PCB) 1.60ECC88 5.70 EL509/519 13.00 6336A 48.00 Ditto, Gold Pl. 3.00ECF82 5.50 E84L/7189 7.50 6550WA/WB 15.00 Octal (Ch or PCB) 1.80ECL82 6.00 KT66 11.00 7581A 12.00 Ditto, Gold Pl. 4.20ECL86 6.30 KT66R 22.50 807 10.70 UX4 (4-Pin) 3.60EF86 6.00 KT77 13.20 811A 11.80 Ditto, Gold Pl. 5.50E80F Gold Pin 11.00 KT88 13.50 812A 31.00 4 Pin Jumbo 10.00E81CC Gold 8.00 KT88 (Special) 17.00 845 (New des) 33.50 Ditto, Gold Pl. 13.00E82CC Gold 9.00 KT88 (GL Type) 30.00 RECTIFIERS 5 Pin (For 807) 3.30E83CC Gold 8.50 PL509/519 9.90 EZ80 5.10 7 Pin (For 6C33C) 4.70E88CC Gold 8.80 2A3 (4 pin) 15.50 EZ81 6.00 9 Pin (For EL509) 5.006EU7 7.00 2A3 (8 Pin) 17.50 GZ32 15.50 Screen can B9A 2.206SL7GT 8.90 211 23.00 GZ33 15.50 Ditto, Gold Pl. 4.306SN7GT 5.30 300B 45.00 GZ34 7.20 Top Con. (For 807) 1.706922 6.40 6C33C-B 25.00 GZ37 15.50 Ditto, (For EL509) 2.007025 7.00 6L6GC 7.60 5U4G 6.30 Retainer (For 5881) 2.20

6L6WGC/5881 8.90 5V4GT 5.00****** And a few Other Brands, inc. rare types ******5R4GY Fivre/GE 8.50 6SL7GT STC 13.00 13E1 STC 100.00 6550C Svetlana 18.005R4WGY Chatham 10.50 6SN7GT Brimar 13.00 211/VT4C GE 120.00 6146B GE 18.505Y3WGT Sylv. 6.50 12AT7WA Mullard 6.00 300B JJ 56.00 A2900 GEC 15.006AS7GT Sylv. 12.00 12AU7 Mullard 12.50 300B Svetlana 80.00 E88CC Mullard 14.606AU6WC Sylv. 5.10 12AY7 GE / RCA 8.40 300B WE 195.00 F2a Siemens 145.006B4G Sylv. 27.00 12AZ7 Westh. 8.00 805 USA 52.00 KT66 GEC 69.006BW6 Brimar 5.40 12BH7A RCA 14.00 5842A GEC 15.00 KT88 JJ 17.406BX7GT GE / RCA 9.00 12BY7A GE 9.50 6080 Telef. 13.30 KT88 Svetlana 35.006CG7/6FQ7 8.50 12E1 STC 12.50 6550A GE 31.50 PX25 KR 128.00

ALL PRICES IN U. K. POUNDS Please note extras: carriage charge (3.00 in U.K.) & in EEC VAT (17.50%). When ordering please state if matchingrequired (add 1.00 per tube) . Payment by credit card (VISA, AMEX etc.) or TRANSFER or CHEQUE (UK only).

FAX email or POST your ORDER for immediate attention We will send PROFORMA INVOICE if required.MILLIONS OF OTHER TUBES & SEMICONDUCTORS IN STOCK!

** Valve Amplifiers sound better still with CVC PREMIUM Valves! ****

PRICE VALIDITY TO END APRIL 2002 ASK ABOUT ANY TYPES NOT ON THIS LIST

******

nently in place. You should wait, how-ever, until you have aligned and testedyour radio before using the epoxy orpolyurethane, in case you need tochange the number of turns on L1. Youcan hold the turns temporarily in placeby using tape.

Loop antennas are directional, andmaximum signal pickup is obtainedwhen the signal arrives in the plane ofthe antenna. Conversely, there is a nullin the signal pickup when the signal di-rection is perpendicular to the plane ofthe antenna. I bolted the antenna up-right supports to the radio chassis inmy receiver, so I need to rotate the en-tire chassis for best reception. But youmay choose to build your radio so thatthe antenna itself rotates.

Assemble the radio itself on a 12 8 2 aluminum chassis. You shoulddefinitely use a metal chassis to obtaina low impedance RF ground. FromPhoto 1, you can see the RF circuitry ison the left-hand side of the chassis,with tubes V1 and V2 at the left rear.

The main tuning capacitor C1mounts on top of the chassis inside a 3 4 5 aluminum box. The box pro-vides a ground shield around C1, pro-tects it mechanically, and keeps dustfrom accumulating between the plates.The power-supply transformer T3 andchoke L4 are at the right front of thechassis, and output transformer T2 is atthe right rear.

Proper grounding is essential in anyRF or audio circuit, and having both onthe same chassis as the power supplycomplicates matters. You should keepall RF grounds short (less than ) andmake them directly to a solder/groundlug bolted to the chassis. Make groundsfor audio tube V3 in the same way.

Isolate the ground for the auxiliaryaudio input jack from the chassis witha nylon washer and run a wire to thesame ground lug that R18 and R19 con-nect to. You should make all groundsfor the power supply and audio poweramplifier (tubes V5V7) to an AWG 14(or larger) copper bus wire, which con-nects to the chassis at a single point,preferably at the same ground lug thatR18 and R19 connect to. Connect theground wire in the AC power cord di-rectly to a ground lug.

For best operation of the radio (aswell as for safety reasons), the radio

needs a grounded three-wire line cord.If you live in an older house and yourelectrical system is ungrounded, youllneed to provide an alternate ground forthe chassis. The Radio Amateurs Hand-book describes safe and effective waysof doing this.5

Keep wiring for the RF circuitry asshort as possibleideally or lessbut I have found that it was not alwayspractical to do this. Photo 4 shows anunderside view of the chassis, with myRF wiring in the right rear. The wiresleading to capacitors C1 and C2 and an-tenna L1 are the longest ones. I keptthese wires away from the chassis andaway from other wires and componentsin order to minimize stray capacitance.

SETUP AND ALIGNMENTTo set up and align your receiver, youllneed a high input impedance (10M)vacuum tube voltmeter (VTVM) or digi-tal multimeter (DMM). Youll also needan RF signal generator to provide testsignals at the low and high ends of theAM band (approximately 600 and1400kHz). An oscilloscope is not neces-sary, but one can be very helpful introubleshooting.

If you dont have an RF signal gener-ator, you can build the circuit of Fig. 4to supply the test signals. This circuit isa Hartley oscillator built with a spare6BA6 tube and using the radios own B+and heater supplies (remove one of theradios 6V6 tubes while you power thisoscillator from the radios supply rails,to avoid overloading the power supply).

To prevent direct pickup of the oscil-lator signal, you should provide shield-ing by building the oscillator in its ownseparate metal minibox. If you build theoscillator with close tolerance compo-nents as specified, the test signalsshould be within 7.5% of 625kHz and1420kHz. If you have an accurately cali-brated oscilloscope, you can check theoutput frequencies to be sure theyrecorrect. If not, you must trust that theoscillator is within tolerance.

When you first power-on the receiv-er, check the supply voltages to be surethey are correct (10%) as shown in Fig.3. Adjust trimmer R37 to set the DCheater supply to 6.3V. Also check thevoltages at selected points in Figs. 1and 2, which should be within about10% of the values indicated.

Specifically, the 6V6 cathodes shouldbe at 22V, the V5 cathode at 92V, the V3Bcathode at 3.3V, the V3A cathode shouldbe at 4.0V, the V2 cathode at 59V, andthe V1 cathode at 2.6V. If these voltagesare all approximately correct, all of thetubes in the receiver are probably wiredproperly. If not, look for a wiring errorassociated with the respective tube.

To align the receiver, connect eitheryour RF signal generator or the circuitof Fig. 4 to the ungrounded side of an-tenna L1 through a 100k resistor (putthe resistor on the receiver side of thewire or cable you use). If your tuningcapacitor C1 has trimmer capacitors,adjust the trimmers for minimum ca-pacity (fully out). Attach your VTVM topoint A on Fig. 1, and set C2 to itshalf-meshed position. Apply a 600kHzsignal and tune C1 for the maximumnegative voltage, which corresponds tomaximum eye closure on V4.

You may need to turn down the signalgenerator level to avoid overlapping thelighted portions of the eye. Conversely,you may need to turn up the signal gen-erator to see any indication on V4. Whencorrectly tuned, C1 should be almostcompletely meshed. Be careful to avoidtuning to a second harmonic of the sig-nal generator at 12001250kHz.

Using a hexagonal adjustment tool(available from Radio Shack as part of acolor TV alignment tool kit), adjust theslug on T1 for maximum eye closure.Then readjust C1 for maximum eye clo-sure, then adjust C2, then T1. Continueadjusting these three in sequence untilthe eye closes no further. You may needto turn down the signal generator asyou approach alignment. Temporarilymark the position of C2.

Apply a 1400kHz signal and adjustC1 for maximum eye closure, then ad-just C2. Alternately adjust C1 and C2until the eye closes no further, and tem-porarily mark the position of C2. Makesure that maximum eye closure is ob-tained before C2 is fully meshed. Ifmaximum eye closure is obtained onlywith C2 fully meshed, youll need to addsome additional capacitance in parallelwith C2.

Try adding a 10pF dipped mica inparallel, then repeat the alignment pro-cedure. But I doubt that youll need anyadditional capacitance if you used a33pF trimmer for C2. Assuming C2 is


not fully meshed, alignment is com-plete. In my radio, C2 is at about one-third meshed at the low end of the AMband and at two-thirds meshed at thehigh end.

OPERATIONAt this point you should be able to re-ceive many AM radio stations. Tune theradio using C1, then adjust C2 as a fine-tuning control. A little practice willmake this easy. Using radio stations asmarkers, you can determine the tuningrange of your receiver.

If you find that you cannot tune tothe top end of the AM band, you can re-move either a full turn or half turn fromL1. Conversely, if you cannot tune tothe low end of the AM band, try addinga full or half turn to L1. Youll need torealign the receiver if you change thenumber of turns on L1. As I mentionedearlier, tapping L1 is one method to ex-tend the tuning range of the receiver,but the need to realign when the tapsare changed limits the practicality ofthis approach.

Once youve settled on the number ofturns for L1 and made your final align-ment of the receiver, you can make acalibrated dial for C1, and perhaps forC2 as well. I simply affixed a piece ofpaper to the aluminum chassis behindthe knob for C1 and marked my favoritestations on it. But you could easily comeup with a more professional-lookingdial.

Youll find that the receiver has moregain and greater selectivity at the lowend of the AM band than at the highend, but this does not seem to be a seri-ous problem in practice. The reason forthis variation is found in transformerT1. Since the primary of this trans-former needs to be resonated at a fre-quency below the AM band, the effec-tive load on V1 varies as a function offrequency, and hence the gain varies.

It is also possible to wind a trans-former with a primary that is resonatedabove the AM band. In fact, that was themore common approach back in theearly days of radio. But that only revers-es the problem, moving the high gainend of the band to the top.

This variation in gain and selectivitydue to the coupling transformers is oneof the limitations of a TRF receiver. Ifind the gain and selectivity more than

adequate across the entire band. But ifyour favorite station is weak and nearthe top end of the band, you may findthe gain or selectivity inadequate. Ithink that is unlikely, but youll only findout for certain by building a prototype.

As I mentioned, the strongest RF sig-nals in my location produce about 1.5V at point A in Fig. 1, and I set thegain of IC1 to produce full eye closurefor this signal level. You may find thatyou have stronger or weaker signals,and you can adjust the value of R17 toincrease or decrease the gain of IC1and keep the tuning eye in a useful

range. If you live in an area with a verystrong AM station, you may get morethan 3V at point A, which is aboutthe largest signal that the audio ampli-fier V3A can handle without distortion.

If that is the case, you have two op-tions for handling the large signals.One option is to use a 1M potentiome-ter in place of R10 to allow you to atten-uate the audio signal (connect thewiper to the grid of V3A). The secondoption is to add an RF gain control byplacing a variable resistor in serieswith R1. Connect one end of a 5k po-tentiometer to ground, and connect


both the wiper and the other end of thepotentiometer to the previously ground-ed end of R1. When the potentiometeris set for zero resistance, the circuit willwork as it does now. As the resistanceis increased, the gain of V1 is reduced.

If you do add this RF gain control,you should use a high-quality poten-tiometer, because DC current flowsthrough it at all times, and standardcarbon pots quickly become noisyunder such conditions. Youll also wantto make sure that the potentiometerhas almost exactly zero resistancewhen it is set to one end of its rotation.I recommend a Clarostat or similar con-ductive plastic potentiometer. Avoidwirewound potentiometers becausethey are inductive. You can also use aswitch to select resistors of differentvalues for R1.

AUDIO QUALITYTo get the best audio performancefrom this receiver, I recommend ahigh-quality speaker, although it neednot be expensive. Most AM radios havepoor speakers. I use a MadisoundSledgeling, a small two-way book-shelf speaker. The audio quality youget will also depend on the RF signalstrength and how much RF interfer-ence there is in your location, and bothof these factors can vary from day today.

Finally, the audio quality you get de-pends on the quality of the signal thatthe radio station itself transmits. Al-most all radio stations (AM and FM)use some sort of signal processing toincrease the average modulation per-centage of their signal. This puts morepower into the sidebands and increasesthe effective amount of power that thestation is transmitting, making the sta-tion sound louder by increasing the signal-to-noise ratio.

There is nothing really wrong withthis, but it obviously impacts audioquality, and if it is overdone the qualitycan be poor. Stations also sometimes

indulge in equalization, effectively turn-ing up the bass and treble controlswhether you like it or not. I find thatWNYC and other NPR stations broad-cast high-quality signals.

CONCLUSIONSThere are several options for buildingyour radio. Obviously, you can build theentire receiver as shown in Figs. 13, asI did. You could also use the AM tuneras part of your audio system and omitthe power amplifier circuitry in Fig. 2.In this case the power transformer T2and filter choke L4 can both be smaller(a 50mA current rating would be suffi-cient). If you just want to experiment,you could build the tuner circuitry inFig. 1 and power it from one or morebench power supplies, but you shouldstill use a metal chassis.

An AM receiver may seem like ananachronism in these days of radio pro-grams streamed over the Internet, but ithas brought pleasure to me and to mywife. Perhaps the apparent anachro-nism is part of the pleasure. In spite ofthe radios simplicity, it is the most sen-sitive radio in the house. The large an-


Recorded by Brian H. Preston

A wonderful collection of music recorded using a crossedcardioid microphone technique. This technique enablesimpressive stereo sound reproduction in keeping with

Blumleins original theory. Selections have been chosen toinclude a variety of sounds including vocal, chamber

ensemble, organ, and jazz band. The music is engagingand the soundstage heard from this technique is remark-

able. As well as being entertaining, this CD will also makean excellent test disc for your audio system. 2003.

Sh. wt: 1 lb.

Old Colony Sound Laboratory, P.O. Box 876 Dept. X2, Peterborough, NH 03458-0876 USA Toll-Free: 888-924-9465Phone: 603-924-9464 Fax: 603-924-9467 www.audioXpress.com E-mail: [email protected]

To order call 1-888-924-9465or order on-line at www.audioXpress.com!*Shipping add $4.00 in US; $5.00 in Canada;

$7.00, all other surface; $11.25 all other air.

UDIOGOA Nwww. .comBuy low. Sell hi-fi.

H I G H E N D A U D I O M A R K E T P L A C E

tenna and a sensitive detector are thekeys to its RF performance.

On recent winter evenings, my wifeand I have listened to a French lan-guage station from Montreal, which isover 300 miles away. It is certainly pos-sible to build a more sensitive receiverusing the superheterodyne approach,but I doubt youll be able to beat theaudio quality of this receiver. It is notCD quality, of course, nor even FMquality, but I think I can promise some-thing much better than youve come toexpect from AM radio. I think you willbe most pleased with it.


REFERENCES1. Vince Roberts, Letters, audioXpress, February2001, p. 81.2. My favorite regenerative receiver appeared on p. 210 of the 1944 Radio Amateurs Handbook, al-though this reference is now practically unobtainable.3. The invention of the superheterodyne was the workof Edwin Howard Armstrong, who also invented FM.See Man of High Fidelity by Lawrence Lessing, Ban-tam Books, Chapter 7, 1969.4. The ARRL Handbook for Radio Amateurs, ISBN #0-87259-186-7, p. 17.18, 2000.5. The ARRL Handbook for Radio Amateurs, ISBN #0-87259-186-7, p. 9.3, 2000.

LANGREX SUPPLIES LTDDISTRIBUTORS OF ELECTRONIC VALVES, TUBES & SEMICONDUCTORS AND I.C.S.

1 MAYO ROAD, CROYDON, SURREY, ENGLAND CR0 2QP24 HOUR EXPRESS MAIL ORDER SERVICE ON STOCK ITEMS

E-MAIL: [email protected]

A SELECTION OF OUR STOCKS OF NEW ORIGINAL VALVES/TUBES MANY OTHER BRANDS AVAILABLE

MANY OTHER BRANDS AVAILABLEThese are a selection from our stock of over 6,000 types. Please call or FAX for

an immediate quotation on any types not listed. We are one of the largest distributors of valves in the UK. Same day dispatch. Visa/Mastercard acceptable.

Air Post/ Packing (Please Enquire). Obsolete types are our specialty.

5R4GY RCA 7.505U4GB SYLVANIA 15.00 5Y3WGT SYLVANIA 5.006BX7GT GE 7.506FQ7 EI 5.006L6GC SYLVANIA 20.006L6WGB SYLVANIA 15.006SL7GT USA 7.50 6SN7GT USA 7.506V6GT BRIMAR 7.5012AX7WA SYLVANIA 7.5012BH7 BRIMAR 12.0012BY7A G.E. 7.00211/VT4C G.E. 85.00807 HYTRON 7.505687WB ECG 6.006072A G.E. 10.006080 RCA 10.006146B G.E. 15.006922 E.C.G. 6.006973 RCA 15.007308 SYLVANIA 5.00SV6550C SVETLANA 20.00

PHONE44-208-684-1166

FAX44-208-684-3056

ECC81 RFT 3.00ECC82 RFT 6.00ECC83 RFT 8.00ECC83 EI 4.00ECC85 RFT 5.00ECC88 BRIMAR 6.00ECC88 MULLARD 10.00ECL82 MULLARD 5.00ECL86 TUNGSRAM 10.00EF86 USSR 5.00EF86 MULLARD 15.00EL34 EI 6.00EL37 MULLARD 30.00EL84 USSR 3.00EL509 MULLARD 10.00EL519 EI 7.50EZ80 MULLARD 5.00EZ81 MULLARD 10.00GZ32 MULLARD 25.00GZ33/37 MULLARD 20.00PL509 MULLARD 10.00UCH81 MULLARD 3.00UCL82 MULLARD 2.00

A2900/CV6091 G.E.C. 17.50E82CC SIEMENS 7.50E83CC TESLA 7.50E88CC G. PIN TESLA 8.50E188CC MULLARD 20.00ECC81/6201 G.E. 5.00ECC81/CV4024 MULLARD 6.00ECC81/M8162 MULLARD 7.50ECC81/6201 G. PIN MULLARD 10.00ECC82/CV4003 MULLARD 15.00ECC82/M8136 MULLARD 17.50ECC83/CV4004 MULLARD 40.00

SOCKETSB7G CHASSIS 0.60B9A CHASSIS 1.00OCTAL CHASSIS 1.00OCTAL MAZDA 2.00LOCTAL B8G CHASSIS 2.50

SCREENING CANSALL SIZES 1.00

STANDARD TYPES AMERICAN TYPES SPECIAL QUALITY TYPES

Alarge mouth concrete axisym-metrical tractrix horn with asmall throat is likely made upof two sections: the neck and

the bell (Photo 1). The reason istwofold: weight and manufacture tech-nique as a result of gravity. The neck isthe length from the throat to roughlyhalfway along the axis toward themouth. The bell is the rest of the expan-sion to the terminus mouth flair.

This article describes how to makean extension to the tractrix horn de-scribed in Build Your Own Axisym-metric Tractrix Horn (aX Sept. 02, p.28); in particular, to the 30 diametermouth with 9 throat (Photo 2) thatwas mated 1:1 to a 12 EV. (I misstatedthe diameter as 36 in the article. It real-ly measures 30. Sorry for the mistake.)

STRETCHING THE NECKApplying theory, you determine thatthe lower cutoff of a 15 radius mouth is150Hz. By lengthening the neck alongthe tractrix curve path and thus effec-tively narrowing the throats diameter,you can mate this large horn to variousother drivers needing a smaller throat. Ichose a 2 diameter throat.

This neck section has two woodenplatesone for the driver and one toconnect with the bell section. Use ma-rine plywood for the plates. In betweenthe two plates is concrete, molded to fol-low the tractrix curve. The mold is builtup of sections of disks, each with a ra-dius of 15 made into several cones andstacked on top of one another. Becauseof the nearly vertical shape, you alsoneed a retention mold to keep the mor-tar from falling off, which is what setsthis project apart from the first.

You must transpose thebells eight mounting holes ac-curately to the connecting plate, sowhen you fasten it the inner circlesmatch up smoothly and the tapers arefacing the same direction. Ideally, thesetwo wooden rings are made by joiningtwo plywood disks of 15 diameter,cutting the inner throat circle andmounting holes for the driver (12 EV),at once. But I decided on a differentprocess.

MAKING CONNECTIONI used a test bolt, cut short to with adull point ground on one end. Once youhave inserted and screwed it into one ofthe eight T-nuts in the bells throat plate,with the ground tip reaching slightlyabove the plate surface plane, lift theconnecting plate, align the center holesby lifting and shifting, and then lower it.Once it is aligned, press lightly wherethe bolt is inserted. This leaves an in-dent that marks the spot to be drilledout, slightly oversized relative to the hex bolt used to connect the two plates.

Remove and insert the test bolt intothe T-nut at 125 to the first, againslightly raised above the plane. Insert ahex bolt through the hole drilled in theconnecting plate and into the first T-nut. Since that hex bolt allows the con-necting plate to move slightly, align theinner circles and press down gentlywhere the test bolt is inserted, leavingan indent which you will again drillout. Repeat this sequence for all T-nuts,each time placing more hex boltsthrough the added drilled-out holes inthe connecting plate.

The connecting plate (Photo 3) lookslike a thick flat round washer with a

15 outer diameter and 9 taperedinner diameter. The slant of the taper isabout 70 to the surface plane. Thereare eight diameter holes equidis-tantly placed on a virtual circle corre-sponding to the T-nuts embedded in thebells throat plate, roughly in the centerof the flat washers ring area.

This virtual circle produces twoareas on the flat connecting plate: aninner surface from the 9 diameterhole to the virtual circle and an outersurface from the outer circumference tothe virtual circle. The inner surfacegives the image of what is roughly the

This author shows you how to lengthen his horn design

to increase its compatibi l i ty with other drivers .

By Robert Roggeveen

Extending the 30 Tractrix Horn


PHOTO 1: The completedconcrete tractrix horn.

PHOTO 2: Horn extension inner moldand connecting plate.

cross section of the concrete horn wallat the connecting plate, revealing a 1thickness.

Thirty-two drywall fasteners arescrewed equidistant onto this inner sur-face area of the connecting plate: 1,2, and 2 in length at a 17 angle withthe axis of the horn, slightly more steepcompared to the 20 inward slant of theinner ring. These are drawn as far intothe predrilled holes and placed in asine rhythm: 2, 2, 1, 2, 2, 2, 1, 2

sequence. All slant inward with the con-tour of the horns neck. If they reachoutward into the open, then youve ap-plied them to the wrong side of thewooden ring. Think of all these screwsas anchor rods embedded in the con-crete, holding the wooden ring to it.

The eight mounting holes in the con-necting plate are fitted with the follow-ing ensemble: a hex bolt, a washer, twospacers, and a nut (Photo 4). Thisarrangement is necessary because once

the mortar is applied enough spaceneeds to be left so that you can turn thehex bolt with a socket. The height ofthe tube spacer is 1cut from inner diameter, clear plastic hose (thick) that I had on hand. The otherspacer is a cylindrical form to stabilizethe tube spacer from the inside. Exceptfor painting the connecting plate withbonding agent, it is ready for assembly.

CONSTRUCTION AND ASSEMBLYKnowing the area of the inner circle inthe connecting plate, you can deducethe circumference. Because the innercircle is cut under a slant, use the larg-er of the two circumferences. Like thebell, you will make the neck of 30 di-ameter paper disks sections. From a30 diameter paper disk mark a sectionwith the circumference length slightlylonger, by about 1, than the totallength of the inner circle circumferenceof the connecting plate.

In addition, use a protractor to drawa circle with a 7 radius in the center ofthe 30 disk. For later sections, as theybecome smaller, use a 9 to 10 radius.Cut out what looks like a washer sec-


PHOTO 3: Close-up of connecting plate with fasteners.

S

T h e U l t i m a t e A c o u s t i c A n a l y s i s S y s t e m .

1 - 8 0 0 - 7 3 6 - 2 6 7 3 s e n c o r e . c o m e - m a i l s a l e s @ s e n c o r e . c o mS u b s c r i b e T o T h e N e w s a t w w w . s e n c o r e . c o m

Power fu l Po r t ab l e A f f o rdab le

S

1/12th Octave Real-Time Analyzer (RTA) Energy-Time Graph (ETG)with octave & 1/2 octave band filters

Sound Pressure Level Meter

Speaker Distortion Meter (THD+N) Signal to Noise Ratio

Speaker Polarity Test

Integrated Audio Signal Generator

Handheld, Portable, Battery Operated

Serial Data Download and Report Software

F o r r e a l - t i m e t e s t i n g o fA u d i o S y s t e m p e r f o r m a n c e !

F o r r e a l - t i m e t e s t i n g o fA u d i o S y s t e m p e r f o r m a n c e !

tion. Mating the linear edges of the sec-tion with two paper clips will produce atruncated cone. For a thorough expla-nation of how to make cones and thetractrix curve for these concrete horns,read the aX Sept. 02 article.

Slip the connecting plate over thecone. It should not slide off at the bot-tom. If it gets hung up with more than remaining under the plate, adjustthe paper overlap. It helps to mark sev-eral radial lines on the paper section tokeep the cone properly aligned whenyou make adjustments. Cut the excess,if any, to leave a glue strip.

This is the initial cone. Before gluingit, copy this paper section on the left-over section of the 30 diameter diskyou just worked with. You can cutabout four or more cone sections fromthe same disk. Each cone is less incircumference than the previous one.So, following the initial cone, the sec-ond cone is less in circumferencelength, the third is 1 less than the ini-tial circumference length, the fourth is1 less than the initial circumferencelength, and so forth.

All cones stacked up produce a shal-low axisymmetrical tractrix curve of theneck of the horn along its 24 length.Prepare the stacked cones with epoxyfor a sturdy and water-repellent mold. Itis now ready for assembly.

The outer retainer mold has a basecone that fits outside the 32 screws andjust inside the eight mounting holes ofthe connecting plate (Photo 5). For theretainer mold a larger diameter paperdisk is used to make the cones. Theyhave at this distance along the axis agreater base circumference. I estimated48 diameter and worked with sectionsof that size. The shape of the outer re-tainer mold is not critical and is builtup in four sections (Photo 6).

Working with cones of 6 in heightallows for adequate though tight spaceto work mortar into the void formed bythe inner tractrix mold and the outer re-tainer mold. Once the base cone hasbeen established, the second cone willfit over the base cone at its very top(Photo 7), producing an overlap of just. The third cone overlaps the secondcones top . And the fourth overlapsthe third (Photo 8).

Whereas building the tractrix curvemold cones are stacked on one another

leaving only a small conical area ex-posed at the bottom of the previouslyplaced cone, the building of the outermold is one where stacking conesleaves the vast majority exposed, over-lapping only or small conical area atthe top of the previously placed cone.

COVERED WITH MUDConsider the substantial dry weight ofthe concrete. The 30 bell weighs 63 lb.The extension weighs 40 lb. This outerset of cones does not have the integralstrength of the inner tractrix mold. Fur-thermore, the cones of the outer moldare not glued to each other, so you needto reinforce them with lateral rings per-pendicular to the cone surface extend-

ing outward (Photo 6). Cut paper ringsand glue them with epoxy onto theouter surface of the individual outermold cones. Paint the cones on the in-side with epoxy to strengthen them andmake them water-repellent.

In addition, applying mortar causesthe outer cones to move upward rela-tive to the inner mold! To counteractthis force, you may choose to apply tabsto the base outer mold, held in place bythe eight hex bolt ensembles. I did nothave these tabs in place, causing themold to move upward when Iworked the mortar into the gap.

Take the prepared connection platewith its screws, bolts, and spacers inplace, and fit it into position onto the


PHOTO 5: Fitting the base retainer mold sections, reinforced by a second ring.

PHOTO 4: Bolt mounting and casting spacer sequence.

epoxied tractrix inner mold, at the base.Just prior to applying mortar, paint theinner surface area and screws on theconnecting plate with a bonding agent.Note the outer mold cones with thepaper reinforcement rings mounted.These are painted with epoxy on the in-side. A plywood disk 6 in diameterwith a 2 circular hole in the middleflattens the mortar level at the throat,once applied.

Start by applying mortar into the nar-row upright wall void formed by theouter base cone tractrix and tampingand thoroughly working the mudaround the screws at the base and allalong the neck upward in a spiral mo-tion. Stack the four outer mold conessequentially as you build up the mortarto form the outer retainer mold. Thegreater the height, the more precariousit is to work the mortar. The height of

each cone just barely worked. I used apalette knife to shove mud into what isessentially a curved slot, poking andmixing, laying and pressing fairly swift-ly, moving around the perimeter. Afterbuilding up the first 5 of horn wall,place the second outer cone, overlap-ping the first by .

Measuring, marking, cutting, andgluing paper are all quite important,just as with the inner tractrix mold.Good accurate craftsmanship pays off.Once you toss a portion of mud into thewall void, work it right away, keep thelayering manageable, and dont forgetwhere you just were. Some of this kindof work is done by feel, as if blindfolded(Photo 9).

FINISHING TOUCHESAt the very upper part of the neck, atthe throat, use the wooden level ring toproduce an even plane. It helps to havemarked the inner mold at the 2 diam-eter level. I overapplied mortar, com-pressing the concrete slightly down tothe proper mark. Stick six tack nailsinto the mortar near the throat on theoutside perpendicular to the outer


PHOTO 6: Retainer mold cone sections.

mold surface, pricking through thepaper cone. Let it dry.

Install the driver mounting platewith T-nuts embedded in the marineplywood as needed. My prototype in-cluded 12 to test the Siare, Pyle, andJBL drivers. Prepare the plate with an-chor screws. Place a paper ribbonaround the perimeter of the plate toproduce a dam to keep the mortar fromspilling off the surface of the plate.

Break the concrete extension out ofboth molds. Let it dry some more. Ifthere are no voids you did a most excel-lent job. Even if there are some big ne-glected spots where the mud was notmixed in well you can fill mortar inlater to repair these blemishes (Photo10). It is essential, however, to have theinside surface finished as smooth aspossible for the sound wave to propa-gate properly. The final horn surface isfinished with a paste used to seal bath-room tile grout.

Place the throat end of the concreteextension on a piece of cardboard(Photo 11). It may be thin walled, so becareful. The best defense against brittlemud is a good mixture and adequate

curing time, three to four days. Checkhow it looks. Does it stay upright orlean? This is why it was important thatthe throat was tapped level.

BONDING THE DRIVER PLATEWith the hole centered, T-nuts embed-ded, wax plugs inserted, and screwsand paper edge sticking up, place theconcrete throat onto the wooden throat.Mix mortar and paint the plate side fac-ing the concrete and the concrete onthe outside of the horn about 4 up with

bonding agent. Place the horn, accu-rately aligning the throat opening.

Apply the mix in manageable steps,working it around the screws using upand down movements with a paletteknife, a dowel, or a piece of wood. Tampand stir, applying the mix around andup the neck. Before the mortar hard-ens, you can finish off the top surface ina number of different ways: leave itrough, smooth it, or cut patterns or let-ters into it. Let it dry. After a day re-move the paper dam.


PHOTO 7: Assembling the horn mold pieces ready for casting, a dry run.

PHOTO 8: Overlapping cones in place, a dryrun. PHOTO 9: Forming the concrete mold.

Shipping andhandling add

$7.25, Canadaadd $11.00

Overseas sur-face mail add

$10.25, air mailadd $22.25.

Order Your Copy

TODAYby calling

1-888-924-9465or order on-line at

www.audioXpress.com

Old Colony Sound Laboratory, PO Box 876 Dept. X3, Peterborough, NH 03458-0876 USA

Phone: 603-924-9464 Fax: 603-924-9467 E-mail: [email protected]

NOW IN THE

E D I T I O N6th More Pages More Graphs More References Brand New Chapters Explanations of New Technology

BKAA60$39.95

When the concrete has cured and itappears solid, fill in gross, moderate,and minor blemishes as needed onthe inside and on the outside of thehorn. Fix the substantial blemishes

with mortar mix. Let it dry and finallycover all of the horn inside with a fin-ishing patch material. A smooth sur-face, with no air leaks, is important.Use a narrow palette knife to strike off


PHOTO 11: Aligning the wooden and con-crete throats.

PHOTO 10: Patch up any blemishes on yourconcrete shell.

$129plus S&H

$99plus S&H

barrel cone truncatedcone

cube cylinder domedcylinder

truncatededge prism

4-sidedpyramid

3-sidedpyramid

truncatedpyramid

(trapezoid)

sphere wedgetruncatedcylinder

ellipsoid squareprism

opt squareprism

reg polygonprism

slantedfront prism

2-chamberBP cylinder

3-chamberBP cylinder

2-chamberBP prism

3-chamberBP prism

Design speaker boxes with Design crossovers withNeed help with your speaker designs?BassBox Pro & Xover Pro can help!TM

Sophisticated Driver Modeling and Advanced Box Design Multiple drivers with isobaric, push-pull and bessel options. 3 dual voice coilwiring options. "Expert Mode" dynamically analyzes driver parameters. Designclosed, vented, bandpass and passive radiator boxes. "Suggest" feature providesfast box design. All box types account for leakage losses and internal absorption. Advanced vent calculation. Bandpass boxes can be single or double-tuned with 2or 3 chambers. 22 box shapes (shown below). Open up to 10 designs at one time. Analyze small-signal performance with: Normalized Amplitude Response, Impedance,Phase and Group Delay graphs. Analyze large-signal performance with: Custom Ampli-tude Response, Max Acoustic Power, Max Electric Input Power, Cone Displacement andVent Air Velocity graphs. Includes a helpful "Design Wizard" for beginners.

Design speaker boxes for any space: car, truck,van, home hi-fi, home theater, pro sound, studio,stage, PA and musical instruments. Import acousticmeasurements. For example, the response of a carcan be imported to simulate the in-car response.

Design 2-way and 3-way passive crossover net-works, high-pass, band-pass or low-pass filters,

impedance equalization, L-pads and series or par-allel notch filters. Its Thiele-Small model provides

professional results without complex testing.

Passive Crossover Filters and Scalable Driver Modeling 1st, 2nd, 3rd and 4th-order "ladder" filter topologies. Parallel crossover

topology. 2-way crossovers offer Bessel, Butterworth, Chebychev, Gaussian, Legendre, Linear-Phase and Linkwitz-Riley filters. 3-way crossovers offer All-

Pass Crossover (APC) and Constant-Power Crossover (CPC) filters. Xover Pro's capabilities scale to fit the amount of driver data. A crossover, filter or L-pad can

be designed with just the nominal impedance of each driver. With full Thiele-Small parameters, impedance equalization can be designed and the performance graphed.

Graphs include: Normalized Amplitude Response, Impedance, Phase & Group Delay. Graphs can include the full speaker response including the box. Estimate component

resistance (ESR & DCR). Calculate the resistance of parallel or series components.

For more information please visitour internet website at:

www.htaudio.comTel: 269-641-5924Fax: [email protected]

Harris Technologies, Inc.P.O. Box 622

Edwardsburg, MI49112-0622

U.S.A.

Copyright 2002 by Harris Technologies, Inc. All rights reserved worldwide. BassBox is a trademark of Harris Tech. Other trademarks belong to their respective companies. Harris Tech reserves the right to make changes without notice. All prices are in U.S. dollars.

Harris Tech Pro software for Microsoft Windows can help you quickly create professional speaker designs. Our software is

easy to use with features like "Welcome" windows, context-sensitive "balloon" help, extensive online manuals with

tutorials and beautifully illustrated printed manuals. We include the world's largest driver database with

the parameters for many thousands of drivers.

any excess humps or bumps. Sand thearea.

CONCLUSIONThis extension project pushed the limitsa bit. All mold materials, although forti-fied with epoxy, are made of paper, soweight plays a major part. The mortarcasting is most exciting and anxiety pro-voking, as about 40 lb of mortar is con-trolled in paper. To see the paper standup to such tensions is quite amazing.

I used two layers of cotton duck ma-terial, such as from old jeans, to make agasket to fit between the neck and thebell. Stitch the two at the edges andmake eyelets for the bolts. The insidediameter should be flush with thewooden rings. Connect the extension tothe bell with eight hex bolts.

Now this is a formidable tractrixhorn approximation. It has a 2 throat,30 mouth, is 37 in length, and weighs100 lb. The weaknesses are that thehorn is slightly foreshortened relativeto the ideal, and it is heavy. Strengthsare firmness, high decibels in the band-pass, on-axis circular propagation ofthe sound-wave front, sound presence,

and punch at a relatively far distance.Table 1 contains 11 response tests I

conducted on six different configura-

tions so far. I performed the tests in anopen field with salal, huckleberry,cedar, and fir at the perimeter 70 or


TABLE 1TEST: 1 2 3 4 5 6 7 8 9 10 11HERTZ dB dB dB dB dB dB dB dB dB dB dB61 50 66 56 54 56 62 67127 65 76 55 75 79 74 54 78 78 79251 73 84 67 80 89 86 71 91 90 92 92499 92 85 81 94 95 93 82 95 95 91 92997 92 93 85 88 93 92 86 93 88 90 1031999 89 90 84 69 88 83 83 85 76 88 974001 83 86 86 81 86 80 84 84 75 76 927993 76 66 85 64 72 73 83 80 64 67 7310007 80 76 77 78 66 73 7012503 56 61 54 63 65 5816001 61 60 64

Throat 2.12 2.12 4.25 2.12 2.12 2.12 2.12 2.12 2.12 9.37 1.5(in. dia.)Mouth 12 9.25 18 30 30 30 30 30 30 30 30(in. dia.)Length 8.75 26 10 37 37 37 37 37 37 11 45(inches)Gap (inches)Back vol 2 2 (quarts)fC(Hz) 358 458 238 150 150 150 150 150 150 150 150Drivers: Test:JBL 104H-2 1, 2, 5, 8SIARE 16-VR 3, 6, 7, 11PYLE W6C200F 4, 9EV 12L 10

more away from all sides. It was asunny, fall afternoon (Photo 12).

Although designed as a tractrix con-tour propagation, the extension mimicsan exponential expansion fairly closely.So I tested it as such (test #2): 2throat, 9 mouth, and 26 long!

For the test I used a CBS test CD (OldColony) and played on an unmodifiedCDB 560 Magnavox CD player. The tubeamp is a 145 Leslie, taken from a tone

cabinet. I measured with a Radio Shackdigital sound level meter (33-2055) on-axis at 1m. Signal potential is held at2.71V.

This extension project spurred me onto experiment further. Test 11 is of ahorn incorporating an additional 6 ex-tension narrowing the throat from 2.12to 1.5 in diameter. I may report on thislater. In the meantime, enjoy buildingand listening to horns.


PHOTO 12: The author tests his finished product.

Sy

st

me

TR

IP

HO

N

S EL E

C TR O

N IC

Detailed information on www.selectronic.fr or by e-mail to : [email protected]>>

by

Probably, the nearest approach to :A STRAIGHT WIRE WITH GAIN.

High end 2 x 100 WRMS / 8 ohms MOS-FET amplifier. SILVER plated PTFE / fiberglass printed circuit boards. SILVER plated connections. Military grade components. Only J-FET and MOS-FET audio transistors. NO CAPACITOR in the sound path . Minimum wiring and easy construction. Outstanding performances. Absolute Transparency and Sound. Extremely reliable. Steady performances

over years. : under certain

conditions

for the DIYer2 Audiophile Masterpieces TRIPHON

The ULTIMATE crossoversystem

GRANDMOSThe SMART amplifier

www.selectronic.frPhone : + 33 - 328 550 328 - Fax : +33 - 328 550 329

BP 513 59022 LILLE - FRANCE

To complete the GRANDMOS (or any good qualityamplifier) we have designed the TRIPHON system :

Musically neutral and transparent. 3-way electronic crossover.

6 or 12 dB/octave slope (true LINNKWITZ-RILEY filter). 4-channel Class A minimalist MOS-FET amplifier

for MID and HIGH frequencies. Choice of cut-off frequency.

PTFE / fiberglass printed circuit boards. Fully discrete component technology.

SILVER plated connections. Minimal wiring.

Afew years ago a friend gave mea Dynaco FM-5 FM tuner,which I cleaned up and tried.The unit worked and provided

mono audio output, but no stereo out-put. I replaced the stereo decoder ICQ205, a Motorola MC1307P, with anNTE Electronics Inc. NTE722, and I re-aligned it. It worked fine and I installedit in my dining room, along with a Dy-naco SCA-80 integrated amplifier andtwo Dynaco A-25 speakers, a true clas-sic all-Dynaco system. I know thissound system is more than 30 years oldin design and production, but it worksand has escaped the landfill.

Later, I noticed that if I turned on thetuner and immediately changed stationtuning, after about five or so minutes itwould start muting occasionally. Afterten more minutes, it would mute per-manently. All I needed to do was to re-tune slightly and it would stay properlytuned and not mute until I turned it off.

I told myself that I probably had beeninattentive when I initially aligned it andI needed to realign it. Then I thoughtthat no misalignment would cause this,but drift would show these types ofsymptoms. Could the power-supply volt-age drift and cause the rest to drift, too?I do not have the equipment to checkthe RF or IF circuits for drift, but I surecan check the power supply.

VERSIONSMy FM-5 is the early model with PC-20and PC-21 printed wire boards (PWBs).There is a later FM-5 version with a dif-ferent set of boards, PC-25 and PC-26,which are the same ones used in theAF-6. The information in this articlecan be used for all versions. The AF-6

has a different front end and an addedassembly, PC-27, to take care of the AMtuning and demodulating functions.

REFERENCED DOCUMENTSI checked back issues of The Audio Amateur for any information on Dynacotuners or FM tuners in general. Up-grade your FM-51 is a two-page articleon improving sonic qualities. There is aKit Report2 on the VSM Audio, MPX-2phase-locked loop stereo demodulatorusing an KB4437 IC, designed for replac-ing the stereo decoder circuit with animproved IC-based stereo decoder. Thiswas followed by a letter3 that describesexperiments done on this kit and othernotes on stereo adjustment for the FM-5.

Finally, I found an article, Upgrad-ing your FM tuner,4 on general modifi-cations. It shows how to add signalbuffers on audiooutputs and cer-tain internal nodeswhere there aresome notable im-provements. Thereis a section onadding a regulatedpower supply, tun-ing indicator, anda muting compara-tor. The regulatoris based on theRaytheon RC4195-TK IC, which is afixed dual 15V pos-itive and negativecircuit. Raytheonsold the rights forthis IC to FairchildSemiconductor,but I could not

find it on their website.The Motorola MC-1468 and the Na-

tional Semiconductor LM325 are equiv-alents. The LM325 will be obsoleteshortly. I could not find the MC1468 onthe On Semiconductor website.

I found nothing on drift or erraticmuting behavior either in the literatureor on the web. I did find the schematicfor the FM-5 with PC-25 and PC-26 and tuner alignment information atthese addresses: http//home.indy.net/

~gregdunn/dynaco/components/FM5/align.html, and http//members.home.net/dunn.greg/fm5/schem.jpg. (Hint: go tothese addresses and save both the tun-ing procedure and the schematics onyour computer for safekeeping, in casethis site disappears.)

JUST MAYBEHaving put forth the hypothesis thatsupply drift was the cause of the misbe-havior, I needed to investigate. Thismeans some measurements and someexperiments. I also needed an explana-tion for drift. Even if the FM-5 is an all-

Discover how to add IC voltage regulators to either a Dynaco model

FM-5 or model AF-6 tuner, and also correct the low-frequency drop in

the demodulated signal of the early model FM-5. By Daniel Dufresne

Improving Dynacos FM-5 and AF-6 Tuners


TO GROUND LINE ONBOARD

22

(1)

14

(2)

B

25

10% W

D207/D60 ANODE

D207/D60 CATHODE

16

29

24

C

MPSU10

21

28

27

15

26

23

E

10% W

+C238/C86250F/25V

R242/R96560

+C237/C85500F/25V

D202/D551N4003

R207/R97

150R 5%, W

Q210/Q60MPSU01/D40D2 D203/D56

1N4003R241/R95560

D207/D601N4743

+ C236/C84500F/25V

D205/D581N4003

+C239/C875F 15V

D206/D50/D591N5244

D204/D571N4003

FIGURE 1: Part of original circuit. A-2137-1

semiconductor tuner, thermal drift iseverywherein resistors, diodes, andtransistors alike. Check some databooks, and you will see. You need to de-termine and evaluate how big and howimportant its impact may be on circuitperformance.

I looked at the tuners schematic andnoticed that it uses a simple shuntzener diode for the negative supply andan amplified zener for the positive sup-ply (Fig. 1). The parts identificationsrefer to FM-5 both PC-21 and PC-26 inthat order, except for D206/D50/D59,which is for FM-5 PC-21/FM-5 PC-26/AF-6 PC-26. The numbered circles repre-sent the PWB eyelets used for intercon-nections. I brought the tuner to my testbench and did some measurements.The results are in Tables 15.

ORIGINAL CIRCUIT DESCRIPTIONS(PART REFERENCES ARE FOR FM-5PC-21 VERSION.)All the power for this tuner comes froma single power transformer with dualsecondary. One winding powers thepanel lamps; the other has a center tapand powers all the electronics. A full-wave rectifier bridge made from fourdiscrete diodes, D202D205, charges uptwo capacitors, C236 and C237, to pro-vide the two raw DC voltages, positiveand negative. A two-stage low-pass filter,made from R241 and C238 and R242 andC239, respectively, powers zener diodeD206, which connects to the pass tran-

sistors, Q210, base. The collector con-nects to the raw positive supply. Theemitter is the output terminal.

The negative supply is made of a se-ries-limiting resistor, R207, and a shuntzener diode, D207. These two circuitsprovide a positive supply of 13.6V and anegative supply of 13.6V.

POWER-SUPPLY MEASUREMENTS ONORIGINAL CIRCUITThe raw positive supply is typically18.8V with about 1.2V peak-to-peak rip-ple. The raw negative supply is 22.6Vwith 250mV peak-to-peak ripple. Allthese values can vary according to thesupply line voltage value.

The positive supply loads are: RFfront end only, 16.7mA18.3mA; PC-20and front end, 63.7mA65.6mA; andtotal positive supply load between93.7mA and 97.4mA. The two lamps,tuned and stereo, are powered fromthe raw DC and consume about 36mAeach. The negative supply load is9.6mA10.2mA (into PC-20) and 9.7mA10.4mA (total).

I measured the warm-up drift, whichcaused the positive supply to change by0.3V after three hours. Power-line regu-lation showed a 0.3V change for a 105Vto 130V input change range. I told my-self that for the tuners cost and year ofdesign, it is reasonable to use simplezener voltage regulator circuits.

Integrated regulators cost very littlethese days and could improve the per-


10% W

D207/D60 ANODE

28

26

25

24

23

22

29

21

16

15

14

27

LM317T

(1)(2)

D207/D60 CATHODE

TO GROUND LINE ONBOARD

25V ALU

25V ALU

ADJOUT

IN

LM337T

ADJIN

OUT

3

2

1

1

2

3

25V TAN

1% w

1% w

MOUNTED ON TERMINAL STRIP

15

14

13

28

23

21

27

25

30

24

29

26

TERMINAL NUMBERS

TERMINAL NUMBERS

FM-5/PC21FM-5 or AFPC 26

FM-5 or AF-6PC 26

FM-5 PC-21

W

+ CNEW 247F

D205/D581N4003

+C239/C875F 15V

RNEW11K2

+C238/C86250F/25V

RNEW2125R

+ C236/C84500F/25V

R NEW 41K2 1% W

R241/R95560

D202/D551N4003

D203/D561N4003

R NEW 3125R 1% W

+C237/C85500F/25V

+

CNEW322F

R207/R97

150R 5%

IC NEWLM337T

+CNEW147F

D204/D571N4003

Q210/Q60LM317T

A-2137-2

SILVER FLUTEMadisound is pleased to introduce

this new line of Wool fiber cone

woofers from Silver Flute. These

drivers feature:

Wool fiber cones Cast frames Rubber surrounds Economical pricing

We currently offer these speakers

in three sizes, 5", 6.5" and 8". The 5"

and 6.5" are also available shielded.

We will continue to expand the line

to include a 10" and 12" version and

4 ohm versions of all drivers.

MADISOUND SPEAKER COMPONENTS, INC.8608 UNIVERSITY GREEN

P.O. BOX 44283

MADISON, WI 53744-4283 U.S.A.TEL: 608-831-3433 FAX: 608-831-3771

e-mail: [email protected]

Web Page: http:/www.madisound.com

W14RC25-S

Fs 43.5 Hz

Qts 0.40

Vas 18.9 ltrs

X-max 3.5mm

Sens. 87dB

Price $15.95

W17RC38

Fs 40.5 Hz

Qts 0.29

Vas 28.2 ltrs

X-max 5mm

Sens. 89dB

Price $19.40

W17RC38-S

Fs 41.3 Hz

Qts 0.31

Vas 24.3 ltrs

X-max 5mm

Sens. 88dB

Price $26.25

W20RC38

Fs 30.8 Hz

Qts 0.23

Vas 83.3 ltrs

X-max 5mm

Sens. 91dB

Price $26.60

W14RC25

Fs 43.5 Hz

Qts 0.35

Vas 18.9 ltrs

X-max 3.5mm

Sens. 87.5dB

Price $14.25

FIGURE 2: Modified circuit.

formance of this tuners power supply. Idecided to replace both the positive andthe negative regulator circuits with ICregulators. Figure 2 shows the schemat-ic of the new voltage regulator section.Whether this change would eliminatethe observed drift and improve mea-sured or audible sonic performance re-mained to be seen.

Should your tuner have a power-supply failure, you might also considerthese modifications, which add littlecost and take two to three hours to exe-cute. The modified circuit performancealong with the original circuit measure-ments are in Tables 15.

NEW REGULATOR CIRCUIT DESCRIPTIONThe positive voltage r

Documents

AudioXpress magazine