Glass Audio Special 05.pdf

(7204 KB) Pobierz

GA505-whole.indd

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

PHOTO 1: 6336-A all-triode

push-pull amp, front view.

6336-A-BASED

ALL TRIODE

P.P. AMP

This amp design features the 6336-A power triode tube, along with push-pull

class A operation, no overall feedback, low distortion, approximately 20W of

output power, reliability, and high-performance sound.

BY KEES HEUVELMAN

& WIM DE HAAN

is an easy-to-use medium-gain tube

with a very low Rp. Moreover, it’s low-

priced and easy obtainable.

ductance, which is the degree to which

the ﬂow of current through the device

changes as a response to a change in

grid voltage relative to the cathode.

By dividing the idle current by the

transconductance, an input overload

voltage results (Vth). If this Vth volt-

age is exceeded in a negatively charged

way relative to the cathode voltage,

the tube stops conducting any current.

his voltage serves as a rough ﬁgure of

merit for a tube circuit, as the higher

the value, usually the better the sound.

Because this Vth is much higher than

in my standard grounded cathode cir-

cuit, I chose this circuit. he basic idea

for the circuit is shown in Fig. 1 .

In other applications I noticed that

a mu-follower (looks quite similar to

O ne of my favorite small signal

tubes is the 5687, which I have

used in several designs with very good

results. Its specs are outstanding and

the sound is therefore really musical.

he result can be a very dynamic and

evolving/bold sound.

I generally use the 5687 made by

General Electric, but I have also used

the Philips/ECG with success. In pre-

vious applications I used a very basic

grounded cathode circuit with standard

values—Ra=15K and Rk=680E—and

with a decoupled cathode resistor. I

achieved the best results with the two

systems in parallel. he 5687 tube

INPUT STAGE

In this design ( Photos 1 and 2 ) I wished

to use a diﬀerent circuit as input stage.

Some time ago I ordered the Tube

Cad Circuit simulation program and

played with it on many occasions. his

tool provides many interesting details,

including circuits, one of which is the

so-called current sourced grounded

cathode ampliﬁer. In this circuit the

plate resistor is replaced with a triode-

based current source, which results in

more gain and better PSRR.

According to Tube Cad, each triode

exhibits a certain amount of transcon-

May 2005 25

235361662.097.png

235361662.108.png

235361662.119.png

235361662.130.png

235361662.001.png

235361662.012.png

235361662.023.png

235361662.034.png

235361662.045.png

235361662.052.png

235361662.053.png

235361662.054.png

235361662.055.png

235361662.056.png

235361662.057.png

235361662.058.png

235361662.059.png

235361662.060.png

235361662.061.png

235361662.062.png

235361662.063.png

235361662.064.png

235361662.065.png

235361662.066.png

235361662.067.png

235361662.068.png

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

26 May 2005

235361662.069.png

235361662.070.png

235361662.071.png

235361662.072.png

235361662.073.png

235361662.074.png

235361662.075.png

235361662.076.png

235361662.077.png

235361662.078.png

235361662.079.png

235361662.080.png

235361662.081.png

235361662.082.png

235361662.083.png

235361662.084.png

235361662.085.png

235361662.086.png

235361662.087.png

235361662.088.png

235361662.089.png

235361662.090.png

235361662.091.png

235361662.092.png

235361662.093.png

235361662.094.png

235361662.095.png

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

May 2005 27

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

235361662.096.png

235361662.098.png

235361662.099.png

235361662.100.png

235361662.101.png

235361662.102.png

235361662.103.png

235361662.104.png

235361662.105.png

235361662.106.png

235361662.107.png

235361662.109.png

235361662.110.png

235361662.111.png

235361662.112.png

235361662.113.png

235361662.114.png

235361662.115.png

235361662.116.png

235361662.117.png

235361662.118.png

235361662.120.png

235361662.121.png

235361662.122.png

235361662.123.png

235361662.124.png

235361662.125.png

235361662.126.png

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

the used circuit) with the signal taken

from bottom triode anode sounded

surprisingly better than if taken from

the cathode of the upper triode—this

was conﬁrmed by several other hobby-

ists. he relatively high output imped-

ance of the circuit is less important,

because the following driver stage will

have a high input-impedance.

hroughout this ampliﬁer, I used

WIMA FKP caps as coupling capaci-

tors. hese low-priced capacitors give

very good sonic results. Feel free to

use more “dedicated” audio capaci-

tors such as Hovland, MIT, or Jensen,

which I’ve used before, but this time I

stayed with the FKPs.

LEDs as voltage reference and added a

potentiometer P2 to adjust the circuit

for current and for minimal distortion.

I used an additional potentiometer P1

to adjust the AC symmetry/balance of

the long tail.

In the original concept the current

source was intended to be connected

to a negative voltage. However, to keep

things simple, and because of good re-

sults with the prototype, I built the

ﬁnal ampliﬁer without using this op-

tion. Reference voltage is taken from

the main high-voltage supply using a

simple resistor.

he 5687 input circuit and the

6SN7-based phase splitter yield enough

gain to make the overall circuit work

without overall feedback. Because of

the relatively low gain, feedback is

not possible using this conﬁguration,

though one of the design goals was no

use of overall feedback. Overall gain/

sensitivity of the design is about 20dB,

which is a good ﬁgure.

he 300B tube is out-of-budget, and,

for example, a 50C-A10 is expensive

and rare. he 6AS7/6080 is an option.

However, after listening to a 6336-A-

based ampliﬁer and knowing that I

had several of them somewhere in a

shoebox, I made my choice just like

that.

he 6336-A tube is intended for

power supply use and contains two

separate triode systems ( Fig. 2 ). Mu is

as low as 2 and power consumption is

as high as 5A at 6.3V; maximum plate

dissipation is 30W. Because plate re-

sistance is very low, the 6336-A tube

can be used in output transformer-

less (OTL) circuits. Another choice is

the 6528, which is very similar to the

6336-A, but mu is 9. his would make

it easier to drive, though this tube is

very rare. Strangely enough, not too

many circuits can be found using the

6336-A—not on the net, in maga-

zines, or other publications. However,

I know that Luxman designed a push-

pull ampliﬁer (the KMQ80) using the

6336-A; unfortunately I was unable to

obtain a copy of the circuit diagram.

I used the 6336-A in full class-A

using cathode bias. It would work

perfectly well with low primary im-

pedance to yield high output power;

however, my goal was for a high 8k

load for low-distortion ﬁgures with

less output power. Output power is

less important.

Because output transformers are

sensitive to DC unbalance, I added a

small meter and a variable resistor to

PHASE SPLITTER

I chose a long-tail circuit as phase

splitter/driver stage. his widely used

circuit yields fair gain and fair bal-

ance—it is perfect for the job. In this

application I chose the 6SN7 in a con-

ﬁguration with a current source, which

I built around a small transistor with

OUTPUT STAGE

In searching for

the perfect out-

put tube, I dis-

covered that ini-

tially the choice

was more diﬃcult

than expected.

he transmitter

tubes are not my

choice; I consider

the very high volt-

ages a problem.

FIGURE 1: Basic circuit of the 6336-A based ampliﬁer.

PHOTO 2: 6336-A all-triode push-pull amp, top view.

FIGURE 2: Power supply 6336 ampliﬁer for two channels.

28 May 2005

235361662.127.png

235361662.128.png

235361662.129.png

235361662.131.png

235361662.132.png

235361662.133.png

235361662.134.png

235361662.135.png

235361662.136.png

235361662.137.png

235361662.138.png

235361662.139.png

235361662.140.png

235361662.002.png

235361662.003.png

235361662.004.png

235361662.005.png

235361662.006.png

235361662.007.png

235361662.008.png

235361662.009.png

235361662.010.png

235361662.011.png

235361662.013.png

235361662.014.png

235361662.015.png

235361662.016.png

235361662.017.png

235361662.018.png

235361662.019.png

235361662.020.png

235361662.021.png

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

the cathode circuit to solve this prob-

lem. Selecting the 6336-A for balance

would be impossible to do. his circuit

is optimized for toroidal output trans-

formers, such as those manufactured

by Amplimo/Plitron.

Due to the high transconductance,

cathode bias is the only option. Nega-

tive bias will not give the 6336-A the

stability it needs. In my attempts to

use negative bias, I had no good re-

sults. With cathode bias, the 6336-A

worked without problems and stability

was excellent.

You must take special care with the

cathode resistors of the 6336-A. he

high bias voltage and current will

produce enormous heat; each resistor

will need to dissipate more than 7W.

Because of the class-A operation, the

6336-A becomes really hot, so pre-

heating the tube for some minutes be-

fore switching the main high-voltage

is necessary.

Using the tube’s Chatham datasheet,

I selected a 1450Ω resistor for biasing

the 6336-A, thus using an eﬀective

anode voltage of 335V (power supply,

cathode bias). In this case, bias voltage

would be around 114V; bias current

would be approximately 70–80mA.

he prototype used two 680Ω resis-

tors in series as cathode resistors; bal-

ancing was done with a tandem 100Ω

potentiometer ( Fig. 3 ). Because these

are hard to ﬁnd, I chose the conﬁgura-

tion as drawn in the ﬁnal circuit ( Fig.

4 ).

In the ﬁnal design, the mains trans-

former has an option for four second-

ary windings: 320V, 330V, 340V, and

350V. I set the high voltage using the

320V winding, yielding a +410V DC

voltage with all loads attached. he

ampliﬁer worked very well and beyond

all expectations. Bias voltage is around

100V, giving a 70mA bias current per

triode.

Because of the high bias voltage and

current, the bias resistors dissipate a

serious amount of heat. Beware of this!

In Photo 3 and Fig. 5 you can see how

to “implant” the 25W resistors. he

2mm aluminum main chassis is used

as a heatsink; the eight 25W resistors

are mounted on an aluminum u-bar

that is aﬃxed to the chassis. Airﬂow

should be more than adequate; but

you might consider installing a small

fan just to keep the air ﬂowing. I chose

BIASING THE 6336-A

In the original concept, the 6336-A

was intended to dissipate around 25W.

FIGURE 3: Alternative output circuit

6336-A using a tandem 100 Ω

potentiometer.

May 2005 29

GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO • GLASS AUDIO

235361662.022.png

235361662.024.png

235361662.025.png

235361662.026.png

235361662.027.png

235361662.028.png

235361662.029.png

235361662.030.png

235361662.031.png

235361662.032.png

235361662.033.png

235361662.035.png

235361662.036.png

235361662.037.png

235361662.038.png

235361662.039.png

235361662.040.png

235361662.041.png

235361662.042.png

235361662.043.png

235361662.044.png

235361662.046.png

235361662.047.png

235361662.048.png

235361662.049.png

235361662.050.png

235361662.051.png

Plik z chomika:

Inne pliki z tego folderu:

Glass Audio Special 05.pdf (7204 KB)
Sound Practices 1992-2000.pdf (77102 KB)

Inne foldery tego chomika:

Zgłoś jeśli naruszono regulamin