Tube CAD Journal Jan 2001
Tube CAD Journal Jan 2001
Tube CAD Journal Jan 2001
January
January
20012001
^^ TUBE
CAD
JOURNAL
^^
The Journal
of Tube
Audio
Circuit Design
This Issue
In This Issue
1Inverted Shunt Regulator
9Portable Tube Headphone Amplifier
22E-mail
Publishing Information
Glossary of Audio Terms
Glass-Ware.com Articles
Classic Magazine articles
< PREVIOUS
Load
NEXT >
+560vdc
+400vdc
1k
+360vdc
Load
output
output
Potentiometer analogy of
shunt and series regulators
Rs
+560vdc
1k
Simple AC
only shunt
regulator
1M
0vdc
pg. 2
NEXT >
Load
The power supply's view of
shunt and series regulators
pg. 3
Regulator Analogies
One way to look at the difference between the
series and the shunt regulator is to view the
series regulator as being analogous to a Class B
amplifier and the shunt regulator as being
analogous to a Class A amplifier. If the regulator
is functioning as an amplifier, what is being
amplified?
NEXT >
+430vdc
+400vdc
+430vdc
0vdc
<PREVIOUS
0vdc
1M
+560vdc
-40vdc
Rs
-160vdc
pg. 4
NEXT >
pg. 5
Antique Electronic
Tubes
Kits & Power Supplies
Sockets
Guitar Items
Transformers
Tools & Chemicals
Capacitors
Books
Other Parts & Supplies
Software
Restoration Products
Over
10,000
square
feet
of:
Tubes, books, transformers, sockets, friendly
folks, capacitors, resistors, literature, cabinet
restoration materials, friendly folks, wire,
grill cloth, gifts, tools, information...and did
we mention the friendly folks?
NEXT >
+400vdc
output
Grounded
grid amplifier
200k
1k
300k
0vdc
5751
+2vdc
input
Rs
200k
+560vdc
1k
1M
+2vdc
0vdc
2k
+400vdc
1M
1M -10
+2vdc
10v
-40vdc
-160vdc
6550
10v
160k
-40vdc
<PREVIOUS
1k
+560vdc
10k
+560vdc
+2vdc 160k
1k
200k
5751
1M
0vdc
-160vdc
0vdc
1k
-40vdc
160k
410k
200k
+560vdc
+400vdc
6550
-20vdc
200k
-160vdc
pg. 6
NEXT >
1k
100k
1k
1M
6BM8
6BM8
260k
0vdc
10v
-12vdc
1M -10
+1vdc
+340vdc
20k
90k
20v
-30vdc
200k
-90vdc
Series
Series
Series Shunting Shunting Shunting Regulator
Raw B+
Resistor Resistor Resistor
Tube
Tube
Tube
Total
Voltage
Volts
mA
Watts
Volts
mA
Watts
Watts
Wall
Voltage
104
130
13.5
400
10
65.5
504
90%
160
200
32
400
80
32
112
560
100%
216
270
58.3
400
150
60
166
616
110%
< PREVIOUS
pg. 7
NEXT >
A Series-Shunt Regulator
Combining the series and the shunt regulator
might prove the best compromise. The benefits
from each type can be added together in one
regulator. The series regulator half provides the
compliant current sourcing that improves the
efficiency and PSRR of the regulator and the
shunt half of regulator provides the net current
canceling function to keep load's current swings
contained to the regulator. Putting both regulator
together requires only a little imagination, as the
regulator will look and function much like a
totem pole push-pull amplifier.
+400vdc
I = (1 + mu) / rp,
Which equals the absolute decrease in current
the top triode experiences:
I = -1 / rp + -1 x Gm
I = -1 / rp + -mu / rp
I = -(1 + mu) / rp.
In other words, 100% feedback from plate to
grind in a grounded cathode amplifier make this
amplifier function identically to a cathode
follower, which also experiences 100%
degenerative feedback.
+560vdc
40v
1k
1M
+360vdc
+50
0vdc
+400vdc
+560vdc
1k
1M
+40
vdc
+360vdc
+80
0vdc
pg. 8
NEXT >
Portable Tube
Headphone Amplifier (part 1)
Battery Concerns
pg. 9
NEXT >
+12vdc
<PREVIOUS
Modes of Operation
Class A output stages, whether push-pull or
single-ended, strive to meat a theoretical limit of
efficiency of 50%, where efficiency is defined as
power delivered into the load divided by the
power dissipated by the output stage. For
example, if an amplifier's output stage consumes
100 watts and puts out 50 watts into a load, its
efficiency would be 50%. Solid-state Class A
amplifiers come close with efficiencies of up to
48%, pentodes usually fare worse, with
efficiencies of only 25 to 40%, and triodes
usually fare the worst, with efficiencies of only
10 to 35%. If the stipulation that the grid never
enter grid current is removed (Class A2), then the
efficiency of tube output stages can move closer
to the 50% theoretical limit, but at the cost of a
much more robust, higher-current driver stage.
pg. 10
NEXT >
output
-10vdc
output
Inductive loaded amplifiers
pg. 11
200vdc
100vdc
output
200vdc
100vdc
output
NEXT >
200vdc
output
0vdc
output
SE Amp CAD
100vdc
RL
RL
-100vdc
200vdc
output
0vdc
Rk
-100vdc
output
100vdc
Rk
RL
GlassWare
PO Box 2
Santa Cruz CA 95063 USA
www.glass-ware.com
pg. 12
NEXT >
200vdc
100vdc
200vdc
where
Rk=RL
T2
T
100vdc
0vdc
100vdc
100vdc
Rk
RL
CS
RL
RL
RL
T1
60 mA
30 mA
0 mA
T2
Rk
CS
T1
60 mA
45 mA
30 mA
15 mA
0 mA
60 mA
45 mA
30 mA
15 mA
0 mA RL
Efficiency:
Ipeak:
RL
RL
RL
12.5%
50%
25%
50%
Ip/2
Ip
Ip
2Ip
< PREVIOUS
pg. 13
NEXT >
NEW
Plate Curves
Tube Substitutions
Datasheets
Tube Outlines
Tube Basing
Complex DB Filtering
GlassWare
PO Box 2
Santa Cruz CA 95063 USA
www.glass-ware.com
Live Curves for over 40 tubes
<PREVIOUS
Database editor
pg. 14
NEXT >
+100vdc to +150vdc
Input
100
B+
Rk
100-400
1m
IRFF220
150k
1k
LF412
-6.3vdc
-6.3vdc
1m
1m
Diodes =1N4001
1f
.1f
100
.1f
1m
1m
-12.6vdc
< PREVIOUS
pg. 15
Output
100
0vdc
1kf
1kf
-12v
-16v to -20v
NEXT >
Audio Gadgets
GlassWare
PO Box 2
Santa Cruz CA 95063 USA
www.glass-ware.com
pg. 16
NEXT >
pg. 17
NEXT >
80vdc
80vdc
CS
60vdc
300
6922
1M
Rk
CS
60vdc
6922
80vdc
3M
CS
60vdc
6922
1vdc
1M
DCR=100
300
1M
300
6922
4.7F
20vdc
1M
Rk
1M
6922
Rk
1M
CS
Rk
CS
70vdc
300
6922
300
6922
1M
Rk
4.7F
71vdc
CS
pg. 18
NEXT >
Push-Pull
Transformer Coupled Amplifiers
Since the music sources, portable CD players,
MP3 player, and pocket FM radios, are all
single-ended output devices, a phase splitter will
be needed to drive a push-pull output stage,
which will greatly increase the complexity and
tube count of a design. Or will it? The White
cathode follower and the SRPP are push-pull
circuits that accept a single-ended, an
unbalanced input. And even a traditional pushpull output stage with two output tubes sharing a
common cathode resistor and plates terminated
into the output transformer's primary, may get
away without a tube phase splitter. An input
transformer with a center-tapped secondary is a
phase splitter. Additionally, a phase splitter can
be from as little as a two resistors! The circuit
below makes the point.
4k
60vdc
6922
6922
20vdc
1M
6922
1M
4k
1M
4k
Ra
60vdc
40vdc
6922
6922
6922
20vdc
6922
1M
4k
Rk
1M
1M
4k
60vdc
80vdc
60vdc
1M
pg. 19
6922
9vdc
10vdc
1M
1k
NEXT >
300
70vdc
6922
300
6922
80vdc
900k
CS
4k
4k
<PREVIOUS
100k
9vdc
6922
CS
1M
RL
Rk
pg. 20
NEXT >
80vdc
CS
300
70vdc
6922
300
1M
1vdc
6922
H
1M
Rk
RL
Your Volume
makes a
Low Noise 24 Position Stepped Attenuator
Precision Control Long Lasting Contacts
10K - 25K - 50K - 100K - 250K
< PREVIOUS
pg. 19
http://www.goldpt.com
NEXT >
CS
300
70vdc
6922
300
6922
1M
6922
1kF
1vdc
6922
1M
OTL single-ended
headphone
amplifier using
3 output triodes
and a choke load
H
DCR=33
Rk
4.7F
10k
15k
10k
103vdc
.22F
73vdc
200
72vdc
70vdc
200
10F
Z's=60v
output
10k
100k
560F
63v 0vdc
300k
OTL
push-pull
headphone
amplifier
200
36vdc
1M
Rk
10k
10k
1M
1kF
10v
<PREVIOUS
pg. 20
NEXT >
< PREVIOUS
pg. 21
NEXT >
<PREVIOUS
E-mail
Subject: printing articles, dropping filament
voltages, balanced phono amps
Hi, I got your address out of the current
Audio Electronics. I have tried printing out 2
articles, so far. I have a Compaq Presario
running IE5 You have stops between each page,
so they are printed out individually, each with
your header! So, the second page comes out
with just one line, listing the title and page 2 or
a short remainder, pushed down by the space
your header takes up. This wastes twice as much
paper! Why don't you fix this? Otherwise, your
magazine is quite informative.
A comment on John Atwood's Ultrapath
circuit. I tried it with the Fender stand alone
reverb CKT - a single ended 6K6 - as I was
building a new guitar amp which included it,
recently. The output hummed loudly and I
traced this to the fact that the capacitor was
coupling power supply noise to the cathode of
the 6K6 and the tube was amplifying it! I had to
add a second choke and cap section to get rid of
the residual hum on the B+.
Hot Wall Voltages
I work on guitar amps. Out here on Cape Cod
they "improved" the power system a while back.
In summer, when it is so hot in Boston that they
send the government workers home... and
predict brown outs, the line voltage down here
is still 123-4 volts! This means that you have
7V on you filaments - perpetually (5.7V on
rectifiers) - in your 115-117V amplifier. What
to do? You mention in-rush current limiters and
there are not many choices. The only applicable
ones are CL-10 (CL-11), CL-20 (CL-22) and - I
believe - CL-100 (Keystone parts - they have renumbered them!) These are the three lowest
resistance values @ 6, 8 & 16 amp ratings. In
practice, they all work about the same. A single
one in a typical BF/SF Fender Pro, Twin, etc.
with non CT filament winding will bring the
voltage down to about 6.3 (or 5) volts. Check it
out!
pg. 22
NEXT >
FET Preamps
Allen Wright goes into this in great detail in
his Preamp Cookbook! Get the transistors - at
reasonable prices - from MCM, 2SK170,
2SK369, single or - I believe - 2SK398? dual
version in the BL? or V (medium or high IdSS
versions) - MCM does not differentiate, so you
can ask them to check, or take your chances.
The 2SK147 is no longer with us but these use
the same die!
John
(USA)
The whole issue of printing would disappear
with the creation of a PDF file for the Tube
CAD Journal. Thanks to readers Johan, Ken,
and Yury the latest PDF renditions look much
better and print beautifully. One problem
remains: the files are too large because they
hold two identical versions of the Tube CAD
Journal issue! Every variable has been tried and
it still comes out with a redundant internal file.
Once this bug is stamped out, I will post a PDF
version of this journal.
As for printing right now, set your browser's
page setup options to .25 inch margins and clear
the header and footer entries.
As for the Ultrapath circuit, the blame does
not belong to John Atwood, as I believe it is
from Jack Elliano of Electra-Print. I have heard
good reviews of his amplifiers and transformers,
but I am sure that it does hum, you see it is only
half wrong, or rather, half right. The complete
UltraUltrapath circuit is covered in the second
issue of this journal in an article titled
"Lowering the Single Ended Amplifiers Output
Noise."
As for the hot line voltages you are
experiencing, where I live on the California
coast, the wall voltage is 122 VAC. But I have a
friend who live 70 miles north in San Francisco
and his wall voltage is only 108 VAC. He lives
in old apartment building that loses 2 volts per
floor. Consequently, his heaters are being under
heated. What to do?
< PREVIOUS
pg. 23
NEXT >
Subject: Circlotron
John I think you need to offer us one more
article on the circlotron amp, this time
addressing offset adjustment. Like the Atmasphere boys I plan to use one stage of gain, but
cascoded with MOSFET to eliminate the need
for a ridiculously high driver stage B+ to obtain
the necessary current draw through the triodes
and also solving the heater to cathode voltage
concern.
Also a pot tying the cathodes together with it's
wiper arm connected to a resistor to ground in
the first stage, followed by a cathode follower
driving twelve 6AS7G's. Global cross coupled
feedback will be introduced into the fully
differential drive stage. Thanks to you I will be
making use of the abundant and cheap supply of
12SN7's or 8SN7's. I find the circlotron to be
absolutely simply and straight forward. I implore
you to compile all your work into a book, we
will buy it in honor of the magnificent tutor you
are. You are not of the average tube specie, your
articles are modern classics, I'll be on your ass
until I see the book.
Rowan
First of al thanks for the compliments. I don't
think this circuit is anywhere near being
exhausted. Now that I own a pair of AtmoSphere MA-1s, definitely expect some more
articles.
MOSFET cascoding of a triode does allow for
big voltage swings. But do not forget that like all
cascode circuits, the cascoded differential
amplifier has virtually no PSRR. So either a
supremely well filtered or regulated power
supply or some of the noise reduction tricks this
journal expounds will be needed.
Feedback can be placed at the cathodes of the
differential pair or at the top of the plate resistors
by cross coupling, which may help preserve the
gain of the driver stage.
As for the book, I am asked monthly to come
out with a book and I should consider it.
<PREVIOUS
NEXT >
< PREVIOUS
Current
25C
125C
Gate-to-source V
Notice that at low currents the positive
relationship between temperature and current
NEXT >
<PREVIOUS
NEXT >
Our Purpose
The Tube CAD Journal is a monthly online magazine
devoted to tube audio circuit design. Each month we will
present some fresh looks at some old tube circuits and
some altogether fresh tube circuits as well (yes, new tube
circuits are possible). Circuits and more circuits. While
we plan on covering complex tube circuits, like phono
preamps or power amplifiers, our focus will be primarily
on elemental circuits. Elemental circuits are the primary
topologies, or part configurations, arrangements that can
stand on their own as recognizable functional circuits
although they may be part of a larger circuit. A power
amplifier circuit, such as the famous Williamson,
comprises several sub-circuits: the Grounded Cathode
amplifier, the Split-Load phase splitter, the Differential
amplifier and finally a push-pull output stage. Just as we
must understand how a resistor or a capacitor functions in
a simple circuit, we must understand the function and
logic of these elemental circuits before we can
understand more complex compound circuits.
Why a Webzine?
The original intent was to print a conventional
magazine. We knew there was a need. A query on our
Tube CAD registration cards that a magazine devoted to
tube circuit design drew an overwhelmingly loud YES.
Still, we knew the difficulty and impracticality of starting
yet another underground tube audio magazine.
The Web offers the publisher some great advantages
over the traditional approach: worldwide distribution,
free subscriptions, no paper (for those who must own a
paper version, the size of the journal has been left small
enough to be printed on A4 or 8.5" by 11" three-hole
punched paper for compilation in a three-ring binder),
live forums, no Post Office, color, motion, a glossary.
Schematics can now evolve, as the web allows for the
easy display of animated GIFs, which display color and
motion. Schematics can now show more than just part
connections, they can reveal voltage potentials, current
flow directions, and possibly, relative impedances.
Math errors and typos will not live indefinitely on a
paper page; once spotted, the Web page can be corrected
quickly.
We look forward to your letters, suggestions and
contributions.
< PREVIOUS
Publisher
Editorial Staff
Editor: John E. R. Broskie
Copy Editor: Anna Russomano Broskie
Mailing Address
P.O. Box 67271
Scotts Valley, CA 95067-7271
E-Mail Address
editor@tubecad.com
Article Submission
A self addressed, stamped envelope must accompany
all mailed editorial submissions. We are not responsible
for unsolicited materials.
Advertising
Please contact us if you are interested in placing an ad
in the Tube CAD Journal.
Printing
While no portion of Tube CAD Journal can be
reproduced for profit without the written permission of
the publisher, we encourage the reader to print a
copy of each Web page for easier reading and personal
archiving. First, click File and then Page Setup
on your browser menu bar and set the left page margin
to inch.
NEXT >