Just about every stereo systemuses a preampor line amplifier,if you prefer. This article de-scribes a preamp (Photo 1) that’s

a little different from most, using low-mu power triodes to get good perfor-mance with only one amplifier stage,and no negative feedback.

WHY DO WE NEED A PREAMP?The most visible functions of a preampare to provide source signal selection(e.g., CD player, phono preamp, and soforth) and volume control for youraudio system. A preamp doesn’t reallyneed very much voltage gain; in fact,some designs, called “passive pre-amps,” don’t have an amplifier stage atall. There is much to be said for thesedesignsafter all, with no gain stageand only switches and passive attenua-tors in the signal path, the noise anddistortion performance is hard to beat.So why would we need an amplifier?

The biggest problem (for me) with apassive preamp is that its outputs havea relatively high source impedance,and one that tends to vary substantiallydepending on the position of the vol-ume control and the output impedanceof whatever is driving it. This works OKas long as the load (the amplifier) iswell behaved, with a high load imped-ance, and there isn’t too much capaci-tance in the amp or interconnect cable.

In my system, this isn’t usually thecase. I typically have several poweramps connected, with only one pair inuse at a time. Often I also have a head-phone amp or two plugged in. Each ofthe loads is usually pretty benign, butwhen you parallel severalalong withtheir interconnect cablesyou wind up

with a load that re-ally needs to be dri-ven with a lowsource impedanceto sound good.

Many amplifiersalso aren’t all thatwell behaved. Peo-ple assume that aclass A1 load (suchas the input of theirpower amp) is al-ways a high imped-ance. At DC, this is usually the case; butthe effective input capacitance can bequite high, especially in designs thatuse high-gain triodes in the input.

TUBE PREAMP DESIGNSSo, you want a preamp, and you’regoing to build it with tubes. You canchoose from several common designtopologies used in tube preamps, aswell as some unconventional ones.

The oldest preamp designs are essen-tially small single-ended power ampli-fiers. They use a gain stage with an out-put transformer as the plate load of atube, with the turns ratio set to provide ahigh load impedance to the plate, and alow output impedance. Many people be-lieve that this is the best type of circuit,partly because it doesn’t use any cou-pling capacitors in the signal path. Thedownside is that it’s expensive to imple-menta good line output transformercan cost several hundred dollars, andthe cheap ones don’t perform very well.

A “classic” design that’s typical of1960s hi-fi equipment uses a high gaintriode such as a 12AX7 as a groundedcathode amplifier, coupled to anothertriode acting as a cathode follower. With-

out negative feedback, this type of cir-cuit has very high gain (around 80) anddistortion, and high output impedance.Lots of overall negative feedback is usedto get the gain down to close to 10 or 20,which reduces distortion and output im-pedance. Though it measures well, Idon’t like the way this type of circuitsounds. I suspect that I’m hearing thelarge amount of feedback that increaseshigher-order harmonic distortion.

A similar design uses a medium-mutriode such as the 6SN7, and no feed-back. This type of amplifier, if properlydesigned and not overloaded, soundspretty good to me. It has low output im-pedance and distortion, but cannot de-liver lots of current to the load (currentdelivery is limited by the amount of cur-rent flowing through the tube, regard-less of output impedance).

Cathode followers are not withouttheir problems (and critics), though. I’vehad problems with the heater-cathodeinterface inducing noise, and even insu-lation breakdown, since the cathode isoperated at an elevated voltage.

Another common preamp stage usestwo triodes in an SRPP circuit, with thetriodes connected in series and the

This author describes how to achieve low distortion

from power triodes. By Pete Millett

A Low-Mu Triode Preamp

4 audioXpress 2/04 www.audioXpress.com

PHOTO 1: Top view of low-mu triode preamp.

upper tube acting as both a cathode fol-lower and a plate load for the lowertube. These circuits can achieve verylow distortion and output impedance aswell, but suffer from the same issues asa cathode follower output stage.

This design takes a different pathasingle gain stage using a low-mu triodewith very low plate resistance (Photo 2).This achieves the modest gain requiredalong with very low output impedance.

LOW-MU POWER TRIODESThe output impedance of a grounded-cathode amplifier is roughly equal tothe plate resistance of the tube in paral-lel with the plate load resistor. Youneed to make the plate load resistorlarge relative to the plate resistance toget low distortion. So, to get low outputimpedance and low distortion, youneed a tube with a low plate resistance.

A number of large power triodeshave low plate resistance (Rp), such asthe 6AS7 and 6080, 5998, 7236,WE421A, 6528A, and 6336A. Most weredesigned for voltage regulator service;the 7236 (I believe) was designed to

drive motors for vacuum-tube computertape drives!

Unfortunately, at first glance, thesetubes don’t look linear. In a normal re-sistor-loaded, grounded-cathode ampli-fier, they have high distortion, mostlyeven harmonics. This distortion de-creases with increasing plate load resis-tance, but a larger resistor means that

you need a higher supply voltage. Luck-ily, there is a way to present a very highload to the plate of the tube withoutusing a big resistor and high voltagepower supplya constant-currentsource (CCS) load.

CONSTANT-CURRENT PLATE LOADSA constant-current source (CCS) plate

PHOTO 2: Quarter view.

+C6

330uF 100V

J13

R4

150 2W

R2

1.5K 2W

L

L

R

R5

1K

L

SW1

3 POLE ROTARY SW

R8

10K

PPJ7

X

+C5330uF 100V

V1

EL34

3

5

2 7

1

8

4

R9

1K 1W

R

WW

R

C1

10UF 400V

CC

J9

J11

INPUT 4

7236 / 6AS7

2

1

3

OUTPUT 2

R

Y

WW

J6

MF

EL*X

R10

1K 1W

R

V3B

R1

1.5K 2W

7236 / 6AS7

5

4

6

78

L

C2

10UF 400V

+320V_L

WW

XR3

1K

J1

VOLUME

INPUT 1

CC

C3

10UF 400V

OUTPUT 3

RPP

+320V_R

J4

INPUT

PP

MF

OUTPUT 1

V2

EL34

3

5

2 7

1

8

4

L

J2

J14

R13

100K STEREO POT13

2

6

5

4

WW

J8

MF = METAL FILM RESISTOR

CC

EL*

L

R7

10K

INPUT 3

Y

INPUT 2

PP = POLYPROPYLENE CAP, SOLEN OR EQ

J3

J10R12

330 5W

WW

Y

PP

R6

150 2W

J12

LJ5

EL* = ELECTROLYTIC CAP, AUDIO GRADE

V3A

CC = CARBON COMP\ RESISTOR

WW = WIREWOUND RESISTOR

CC

WW

R11

330 5W

C4

10UF 400VR

FIGURE 1: Amplifierschematic.

G-2300-1

6 audioXpress 2/04 www.audioXpress.com

load uses an active device (usually atube or FET) that’s connected so that italways wants to supply a constant cur-rent to its load, which in this case is theplate of a triode. This provides a veryhigh resistance load to the plateessentially, a near-horizontal load line.With a CCS load, the load line that thetube sees is dominated by the currentthat’s delivered to the load.

There are many possible designs forCCS plate loadssome very simple (justa depletion-mode FET and a resistor),and some very complex, using bothtubes and solid-state devices. If you’re in-terested in CCS plate load designs, thereis a gold mine of info about them on theInternet (see “References” at the end ofthe article for some places to look).

I experimented with simple designsusing an FET and a pentode. They bothperformed well. I opted for a pentode,partly because I had themthey weresimple and performed well in this appli-cationand partly because they don’tneed a heatsink. Using a tube lets youdissipate the power above the chassis!

One of the problems with using aCCSespecially a tube CCSis thatyou need to have substantial voltageacross them, especially if the outputvoltage swing is substantial. With a low-mu power triode preamp, this isn’t toomuch of an issue; the output swing isonly a few volts, and the triode can behappily run at low plate voltage, leavingmore of the available power supply volt-age to drop across the CCSs.

THE PREAMP DESIGNThe preamp design is straightforward,using EL34 pentodes wired as CCSs asthe plate load for the low-mu triode. Thetriode tube has two sections, and one isused for each channel. The schematicfor the amplifier is shown in Fig. 1.

Note that on the EL34 CCSs, thescreen grid is bypassed by a large ca-pacitor to the cathode of the tube. Ascreen resistor is used not so much toreduce the voltage on the screen, but toallow the capacitor to make the screenvoltage constant with respect to thecathode.

The current through the CCS (andthe triode) is set by the resistor in thecathode of the EL34. The 150Ω resistorshown sets the stage current at about50mA. Note that the grid is connected

PHOTO 3: Single-point terminals.

audioXpress February 2004 7

• Access to over 1.5 million electronic parts and related products

• Superior customer service

• Special offers and discounts

• Quotation team, send us your quotes for quick response

• State-of-the-art automated warehouse

or visit

www.mcminone.com/magazine

Call toll free

1-800-543-4330

FREECATALOG

Source Code:AE03

O v e r 4 0 , 0 0 0 P r o d u c t s S t o c k e d

Order #555-2505

250 WattMixer/AmplifierModules

Order #55-12115

12" GuitarLoudspeaker•Mfr. #C12Q

Order #50-280

50 Watt Stereo L-Pad

Order #LS00385

Acoustic Wadding Material

Order #54-325

Heavy Die-Cast Aluminum Horn Lens

to the lower end of this resistor,through a grid stopper resistor. This re-sistor is criticalwithout it, my amplifi-er oscillated at 115MHz, with 100VRMS on the triode plate!

The triodes are cathode biased with a330Ω resistor. With a 7236 tube and a50mA CCS plate load, this sets the platevoltage at about 125V. I experimentedwith different operating points using the7236 tube, but found that this was aboutthe optimum for lowest distortion.

I bypassed the cathode resistors witha good-quality audio electrolytic capaci-tor. At first glance it seems as thoughthe cap is not needed, since the platecurrent is fixed by the CCS; however,

the LOAD current still flows throughthe cathode resistor, and not bypassingit provides some local negative feed-back, which reduces the gain some-what. In this case (I’m not sure why), itincreased the distortion.

The triode grids are connected di-rectly to the wiper of the volume con-trol potentiometer. There is a compro-mise in selecting the volume control re-sistance: You want a large resistance toachieve a high input impedance, but ahigh resistance, in combination withthe Miller capacitance of the tube, caus-es a high-frequency rolloff. I used a250k pot, and as you can see from themeasurements, obtained reasonable

high frequency response. You will getbetter response from a 100k pot, andeven better still with a 50k pot.

I wanted several inputs to be selec-table with a switch, so I used a good-quality rotary switch to select one offour inputs to feed the volume control.The output is taken from the triodeplate through a polypropylene couplingcapacitor. I connected three sets of out-put jacks in parallel, so I could driveseveral loads at the same time.

POWER SUPPLYThe power supply is conventional,using a tube rectifier and capacitor-input filter. I used a 5U4GB rectifier

PHOTO 4: PCB, top view. PHOTO 5: PCB, bottom view.

J16

AC RECEPTACLE

N L

G

+C9

750UF 400V

C7

47UF 400V

J15

IEC INLET

N L

G

C10

47UF 400V

R15

100 3W

+320V_R

DS1

LAMP

Y

+320V_L

R16

30K 5W

TRIAD C-14X ORHAMMOND 159Q

DS2

LAMP

WW = WIREWOUND RESISTOR

V3

5U4GB

42

86

PP

R17

5K 1W

X

SW2

DPDT CENTER OFF

ELECT

R14

100 3W

ELECT = ELECTROLYTIC CAP

RV1

V230LA10

ELECT WW

PP

+C8

47UF 450VF1

3A SLOW

THORDARSON 26R160OR HAMMOND 270FX

PP = POLYPROPYLENE CAP, SOLEN OR EQUIV.

5V 3A

6.3V 8A

270V .26A

270V .26A

115V

T1

8

9

1

2

3

4

5

6

7

L1

6-7H 150-200MA

WW

WW

FIGURE 2: Power-supply schematic.

G-2300-2

8 audioXpress 2/04 www.audioXpress.com

tube, but others would work as well, asthe total current draw is only around100mA. I used a large second capacitorin the filter, 750µF, to provide low rip-ple. The B+ voltage is between 300V and325V, and in bench tests appears to benot very critical.

The audio circuitry is de-coupledfrom the power-supply filter by using100Ω resistors and 47µF polypropylenecapacitors. I believe that this scheme,which keeps almost all of the audio cur-rent out of the power-supply filter and

instead sends it through the higher-quality polypropylene caps, sounds bet-ter than relying on the large electrolyticcap in the filter.

Filaments are powered with 6.3V AC,which is biased up to about 45V DC, de-rived from the power-supply bleeder re-sistor. This bias is needed to stay withinthe 100V maximum heater-cathode rat-ing of the EL34 tubes, whose cathodesare operating at around 140V. It alsohelps reduce any noise coupling fromthe triode tube’s filament, by reverse-

biasing the intrinsic diode formed be-tween the heater and cathode.

I use the preamp power switch to con-trol the power to my amplifiers, but thereare times (such as when using a head-phone amp) that I don’t want to powerthem on but still use the preamp. To ac-complish this, I used a center-off DPDTpower switch. In one position, power isapplied to just the preamp; in the other,power is applied to both the preamp andto an AC receptacle which is used to turnon the power amps. In the center “off”position, both the preamp and poweramps are off. Neon pilot lamps are usedto show what’s turned on.

In the schematic, you’ll see a metal-oxide varistor placed across the powertransformer primary. This device is atransient suppressor, and helps toclamp both the line voltage spikes aswell as the inductive spike that occurswhen you turn the power off. I alsoused a filtered IEC power input connec-tor to help keep noise out.

THE CHASSIS AND BASEI built the preamp using a little differ-ent construction method than usual:PHOTO 6: Wood for base, cut and ready for assembly.

audioXpress February 2004 9

Founded in 1982 as a manufacturer of magneticcores for transformers and inductors. In 1993branched into products minimizing distortion infine audio systems.

In 1988 developed Laminax™, composites of foil conductors and dielectric film, used as winding conductors and for EMI/RF shielding.

In 1995 developed flat magnet wire, now used in motors and power supplies throughout the world.

Milestones in Audio1993GOERTZ™ speaker cables US Patent. 5,393,9331995Air-Core Foil Inductors for crossover circuits.2001Micro Purl & TQ2 interconnects. US Patent 6,225,5632002BP30, Symmetrical, Balanced power supply.

All products are created and manufactured at ourfacility in Bridgeport, Connecticut, USA.

www.alphacore.comAlpha-Core, Inc. 915 Pembroke St., Bridgeport, CT 06608

• 800-836-5920 • 203-335-6805 • Fax 203-384-8120

point-to-point wiring and single-pointsolder terminals. The chassis top ismade up of a sandwich of PCB material

and polished stainless steel sheet. Theuse of single-point terminals, as op-posed to terminal strips, allows you to

FIGURE 3: Measured specifications.

G-2300-3

10 audioXpress 2/04 www.audioXpress.com

Electronic CrossoversTube

Solid State

Free Catalog:Marchand Electronics Inc.

PO Box 473Webster, NY 14580

Phone (585) 872 0980FAX (585) 872 1960

info@marchandelec.comwww.marchandelec.com

All available as kit

Line Level PassiveCrossovers

Custom SolutionsWe can customize our

crossovers to your specificneeds. We can add notch

filters, baffle stepcompensation, etc....

Lundahl Transformersin the U.S.

K&K Audio is now sellingthe premier European audio

transformers in the U.S.

High quality C-core transformers and chokesusing silicon iron, mumetal, and amorphouscores.• Power output transformers – SE or PP• Interstage transformers – SE or PP• Line output transformers – SE, PP, and parallel feed• Input transformers – MC or line• Audio chokes• Power transformers• Power supply chokes

High End Audio Kits using LundahlTransformers• MC Phono Step-up Unit• Phono Preamplifier• Line Stage Preamplifier• Microphone Preamplifier

For more information on our productsand services please contact us at:

www.kandkaudio.cominfo@kandkaudio.com voice/fax 919 387-0911

place each end of each component ex-actly where you want them.

Single-point terminals are used insome mil-spec point-to-point wiring,and are available in a number ofshapes, sizes, and styles. Photo 3 showsan assortment of types. I used termi-

nals with female threads, such as theones at the right side of the picture(refer to the “Parts Sources” section atthe end of the article for information onwhere to get these terminals).

The terminals are fastened to a pieceof bare PCB material, using flathead

screws that are countersunk flush withthe surface. Photo 4 shows the PCBfrom the top, with the flush screwheads. Other components inside thechassis are mounted to the PCB materi-al the same way.

After I mounted all the terminals,tube sockets, and other components tothe underside of the PC board, I at-tached a thin piece of polished stain-less steel over the top, using an adhe-sive called “3M high-performance trans-fer tape.” This stuff is no ordinary dou-ble-stick tapeit’s an acrylic adhesivethat forms a permanent bond. The“squishiness” of this material alsohelps to make the completed chassisacoustically dead.

The stainless steel covers all of thecountersunk screws, giving a neat ap-pearance on the outside. The only holesdrilled or punched in the stainless arefor components that mount on top ofthe chassis, such as the tubes andtransformer. I purchased the polishedstainless steel pre-cut in a 12″ × 12″square from McMaster-Carr.

The copper foil on the inside of thePCB is used as a ground plane. You

audioXpress February 2004 11

A Complete Precision Clock kit is now available for Audio Research CD2!

The kit is specially designed for AR CD2 and includes all you need:Reference Clock (XO3 Ultra High Precision Clock)Wire Harness specially adapted for CD2Easy to follow Step by Step Mounting instructions, with photos how to update the CD2 playerto the highest sound quality, using modern Low Jitter Clock Technology.More Natural Sound - Better Placement in Soundstage - Improved Bass Distinction

It reallyWorks!

NOTE L C Audio Technology is an independent third party vendor, and has no associations with Audio Research

The latest in Advanced Low JitterClock Technology!

LClock XO3

PriceReference Clock kit forAudio Research CD2

US$ 399

USA: 153 Mountain Home Road Port Angeles WA98362 Washington tel 360-452-9373Australia: POBox 307 Surry Hills NSW 2010 tel 02 9660 1228Scandinavia: Christiansgade 1 DK7500 Holstebro tel +45 97410863

DIYCABLESoundlabs Group pty.

L C Audio Technology Inc

www.diycable.comwww.soundlabsgroup.com

www.lcaudio.com

Complete Reference Clock kit with all you need!

Further Information: www.lcaudio.com

PHOTO 7: Gluing and clamping the base.

could argue that a well-thought-out single-point ground scheme is betterthan a ground plane, but I’ve found thata solid copper ground plane works verywell, and is much easier to wire; youcan just drop a component lead downand solder it to the ground plane. Photo5 shows the copper side of the PCB,with the terminals and componentsmounted. Note the bracket for the inputselector switch, clamp for the electrolyt-ic cap, and filter choke are mounted tothe PCB.

The PCB and stainless-steel sand-wich is mounted into a wooden basethat I made from walnut. Photo 6 showsthe individual pieces of wood cut andready to be assembled. Note thecounter bores on the inside of the frontpiece to allow mounting the powerswitch and volume control. I glued thepieces together with a number ofclamps as shown in Photo 7. The clampgoing diagonally across the base pullsthe base into squareyou need tomake sure the base is square, or yourchassis won’t fit together!

You can see the finished wooden basein Photo 8. The chassis top sits into a

step that I routed into the woodpieces before gluing them togeth-er. Similarly, an aluminum plateattaches to the underside to forma bottom.

WIRING AND COMPONENT INSTALLATIONAfter mounting the chassis“sandwich” into the base, youthen install the remaining com-ponents (power transformer,RCA jacks, and so on). I usedsome copper foil tape to providea ground connection to the vol-ume control from the copperground plane (Photo 9).

I finished all of the wiring be-fore installing any of the parts,using Teflon-insulated solidwire. This type of wire, though alittle hard to strip, is nice towork with, because the solidconductor is easy to wrap around theterminals, and the insulation doesn’tmelt from the heat of a soldering iron.It’s also easy to wrap around the single-point terminals and tube sockets with-out fraying as stranded wire would.

I mounted the input selector switchright above the input connectors on abracket, and operated it with an extend-ed shaft, to minimize the length of thesignal wiring. After wiring, install theleaded components, wrapping theirleads around the terminals and solder-ing. Adding the parts after the wiringmakes it easier to change them afterthe fact, if you want to try different com-ponent values. A view of the inside of

the completed chassis is shown inPhoto 10.

MEASUREMENTSI made a number of measurements(Fig. 3), using both a Cetron 7236 tubeand a Russian 6AS7G tube. The mea-surements included THD, noise, fre-quency response, output impedance,and maximum voltage gain. In general,there were only small differences be-tween tubes, other than the expectedgain difference due to the tubes’ differ-ing amplification factors.

I also performed spectral analysis(FFTs) of the output distortion to checkthe harmonic content. I didn’t include

12 audioXpress 2/04 www.audioXpress.com

PHOTO 8: Completed wooden base.

PHOTO 9: Inside chassis, showing volume controlgrounding.

any pictures of them because they’re notvery interestingthe only harmonic thatwas above the noise was the 2nd har-monic. This was true for all of the condi-tions tested, and for both tubes.

Photo 11 shows an oscilloscope pic-ture of the output (top trace) and thedistortion residual (bottom trace). Thedistortion residual, shown at a muchhigher scale than the output signal,was filtered and averaged to remove allthe noise components. You can see thatit’s pretty much a pure 2kHz sine wave.

As discussed, I used a 250k volume

control pot, which is the limit-ing factor for high-frequencyresponse. Using a lower resis-tance control will extend thehigh-frequency response con-siderably.

LISTENING IMPRESSIONSI never thought much aboutwhat sonic impact my old pre-amp (a PAS-type 12AX7 cir-cuit, with lots of negative feed-back) was having on my sys-tem until I replaced it withthis preamp. The biggest dif-ference was in the clarity ofhigh frequency transients. Forexample, the strike of a bellsounded real, not recorded.

I often don’t like liverecordings, because theytend to sound too muddythere’s so much low-levelstuff in the background (suchas the audience). This pre-amp seemed to resolve moreof that detail, and live record-ings I didn’t like much beforecame alive. I was pleasantlysurprised.

Of course, nothing is per-fect. My only complaint is thatthe tubes are microphonic

and a little noisyespecially whilewarming up. It’s understandableafterall, these are big power tubes, and theyweren’t designed as low-noise, low-levelamplifiers. After they’ve heated up for awhile the noise diminishes, and I sus-pect some vibration dampersmaybejust a silicone O-ringwould help, espe-cially on the tall, skinny EL34s.

audioXpress February 2004 13

REFERENCESwww.audioasylum.com − Great Internet bulletin boardfor all things audio. Check out the “Tube DIY” sectionfor lots of discussion about tube circuits.

www.bottlehead.com − The folks at Bottlehead wereone of the first to popularize CCS plate loads. Checkout their kits.

home.pacifier.com/~gpimm/ − The personal page ofGary Pimm, who has done an impressive amount ofwork perfecting CCS plate loads. Gary also sells PCboards of some of his designs.

pmillett.addr.com − The author’s web site, where youcan download full-scale drawings of this project.

PARTS SOURCESMcMaster-Carrwww.mcmaster.com(630) 833-0300Stainless steel and aluminum sheets, hardware, 3Mhigh-performance transfer tape adhesive

Cambionwww.cambion.co.uk(800) 947-1256Single-point terminals(part number 572-4827-01-05-16)

Concord Electronic Hardware www.concord-elex.com(800) 847-4129Single-point terminals (part number 1127-30-0516)

USECO Fastners www.staffall.com(401) 461-5554Single-point terminals (part number 1417)

PHOTO 10: Inside of the chassis.

PHOTO 11: Output (top) and distortion residual (bot-tom, not to scale).

SOLEN4470 Avenue ThibaultSt-Hubert, QC, J3Y 7T9 Canada

Tel: 450.656.2759Fax: 450.443.4949Email: solen@solen.caWeb: www.solen.ca

DRIVERS: ATC AUDAX ETON FOSTEX LPG MAX FIDELITY MOREL PEERLESS SCAN-SPEAK SEAS SILVER FLUTE VIFA VISATON VOLT

CUSTOM COMPUTER AIDEDCROSSOVER ANDCABINET DESIGN

SOLEN CAPACITORSAND INDUCTORS - USED BY THE MOSTDISCRIMINATING LOUDSPEAKER MANUFACTURERS.

HARDWARE

HOW TO BOOKS

FREE!

Contact us for the freeSolen CDROM Catalog.