Eric Dollard Novel Audio Amplifier and McIntosh Unity Coupled Circuit

One of the things that I'm here for is
0:07
to work with Eric on this amplifier,
0:11
which is not the so-called commercial
0:13
one. This is to amplify the seismic
0:16
signals.
0:17
No, this is the
0:18
the
0:20
how would you say? This is the
0:23
the most complete form of the amplifier.
0:26
The full embodiment.
0:27
Yeah, full embodiment.
0:29
Okay. So, this is the one that we're
0:31
going to patent. Um, this this is just
0:34
for our own internal records, but
0:37
this is going to go in the cable TV
0:38
shack.
0:40
Okay.
0:40
So, this thing was intended to amplify
0:42
the earth signals.
0:44
So, is this going to amplify the the
0:48
signals coming off the long the long
0:50
line?
0:50
Yeah.
0:51
So,
0:52
so that this is this is basically the
0:54
line amplifier and then these are going
0:56
to be the input amplifiers. These are
0:58
already made.
1:02
So these go
1:03
So that's a previously manufactured
1:05
western electric.
1:07
Yeah. This these were a real
1:08
And what were those typically used for?
1:10
This was used in Korean War um army
1:15
frequency division multiplexing
1:18
equipment.
1:21
12 telephone channels on one pair of
1:23
wires.
1:26
And we went over a lot of the
1:27
multiplexing stuff in the musical
1:30
seismograph where we had
1:32
Yeah.
1:40
Okay. And so what's
1:44
kind of the overview of the novelty of
1:46
this particular full embodiment
1:50
uh amplifier? The tubes are not being
1:52
used for voltage gain
1:56
and like you said they just add power
1:57
and all
1:58
power gain. So basically this amplifier
2:01
you know looking at the diagram here is
2:03
a transmission line
2:06
made of a series of transformer windings
2:15
and then the tubes are part of the
2:17
transmission line. And you said the
2:20
tubes add power but all the
2:21
amplification is in the transformer.
2:23
Transformer the transformer windings.
2:25
That's that's that's what's doing the
2:26
voltage multiplication.
2:28
The tubes are basically just negative
2:30
transconductance devices. They just take
2:33
power from the power supply and deliver
2:35
it to the signal as the signal
2:37
propagates through this series of
2:39
transformer windings.
2:42
So you can see here there's a lot of you
2:45
know pretty heavy interconnection
2:47
between all these transformers
2:54
and those are not your ordinary
2:56
transformers. Those are the western
2:58
electric transformers from the Bell
3:00
system,
3:03
which are nearly impossible to find
3:05
anymore.
3:08
And if you do, they're very expensive.
3:13
Let that keep running for a second. I
3:16
get some little more light.
3:19
[Music]
3:31
So provisional needs to be filed on this
3:33
for
3:35
this can be scaled up. The idea I have
3:39
when I developed this at Landers
3:42
was to scale it up to a 300 W audio
3:45
amplifier.
3:45
Mhm. using uh the 811 and 812 vacuum
3:50
tubes which are still available.
3:56
They still make those. Not to say that
3:58
any of the stuff that's made is very
4:01
reliable or accurate, but
4:04
there's still enough surplus around.
4:09
So there's a there's a Macintosh version
4:11
and this would be like an advancement of
4:14
the Macintosh design.
4:16
Mhm.
4:17
So there's a there is a YouTube video.
4:21
I forget the title of it, but it's a
4:22
Macintosh 300 watt amplifier.
4:26
So that's kind of this advancement on
4:29
the Macintosh design. And that's u kind
4:32
of the end aim with this is to build a a
4:36
stereo 300 watt amplifier that's all
4:38
vacuum tube. So you said the um
4:42
Macintosh amps they brought it up to a
4:44
certain point
4:46
but they didn't actually take it all the
4:48
way where you have and where you've
4:51
taken it is um a new novel way to do it
4:56
that's patentable.
4:58
Yeah. Well, the whole the whole idea is
5:00
is this is the thing is is the
5:03
amplifiers that are in current use
5:05
really not are not adapted to driving
5:07
speakers.
5:08
Mhm. Okay.
5:09
Cuz the speakers don't have a you know
5:12
they can say that you have an 8 ohm
5:14
characteristic impedance speaker.
5:17
But that's that's like a a nominal
5:20
value. The impedance varies wide wildly.
5:23
And because all the amplifier designs
5:25
are pretty much constant potential,
5:29
that means that the power going into the
5:31
speaker varies with frequency. So it
5:33
screws your frequency response up and
5:36
makes for you know like dead spots and
5:39
uh you know the speaker kind of mucking
5:42
the signal up because it's not getting
5:44
driven by the impedance that it that it
5:46
wants at that frequency. This amplifier
5:49
design is a constant wattage amplifier.
5:54
So
5:55
if the impedance of the speaker changes
5:57
the amplifier works in a way such that
6:01
it will try to maintain a constant power
6:04
for for a given you know voltage at the
6:08
input from whatever you know CD player
6:11
or whatever you're putting into it. It
6:13
tends to want to deliver constant power.
6:15
That's the one one of the main reasons
6:16
why vacuum tube amplifiers sound better
6:18
than transistor amplifiers.
6:23
because the vacuum tube design other
6:26
than how it's modified by feedback but
6:29
in intrinsically
6:32
it tends more for a constant wattage
6:33
output.
6:36
So if you use transistors and you made
6:40
the same circuit configuration and all
6:42
that you you could get similar results.
6:45
But
6:48
my point is, it's not the tubes that
6:51
sound better. It's the tube circuit that
6:53
sounds better.
6:59
But in general, transistors do have a
7:02
funkier distortion and and it's a very
7:05
difficult distortion to get rid of.
7:08
So, how does this this one differ from
7:11
the one that we're going to um put into
7:14
commercial production?
7:15
The one we're putting in commercial is a
7:18
a simplified version of this.
7:20
Uh-huh.
7:22
It uses less transformers,
7:25
you know, and um
7:28
and transformers that that can be
7:30
readily manufactured or may already be
7:32
available. These type of transformers
7:34
here, I mean, you look at all the taps
7:35
on there.
7:36
Yeah.
7:37
You know, now I don't know if a
7:40
transformer company can duplicate those
7:42
or
7:46
but but at any rate if this thing is
7:48
scaled up then these transformers would
7:51
have to be scaled up which is possible
7:54
because their impedance you know and all
7:56
those things can be identified
7:58
then the transformer manufacturer can
8:00
just you know wind the transformers but
8:03
that but as you see you know it's not
8:04
just a you know primary secondary in a
8:07
simple form. There's all kinds of taps
8:09
involved on there.
8:13
And then there's certain networks that
8:15
that stabilize the uh
8:18
the frequency response for
8:21
non-scinosidal waves like square waves
8:23
or pulses or any of that type of stuff.
8:26
So this this was set up here to be an
8:29
amplifier for the the seismic
8:32
system.
8:33
Mhm.
8:35
So because the seismic system after it
8:38
goes through the modulator, the signal
8:40
is a square wave
8:42
and then the output modulator converts
8:45
that back into the steady state or you
8:48
know slow stuff is uh the amplifier has
8:51
to have excellent fidelity
8:53
characteristics and make sure that that
8:55
square wave goes all the way through
8:56
without any distortion. Otherwise when
8:58
the two modulators do you know their
9:00
conversion back and forth they'll get
9:02
you know inconsistencies in relationship
9:06
between the input and output. So the
9:09
commercial amp is it possible to make a
9:12
solid state version that's
9:15
you know it's not going to be as good
9:17
but for like a lower tier more easily
9:19
manufacturable
9:22
one incorporating the
9:24
possibly but but I don't know. Well, I
9:26
mean, that's what we're going through
9:27
now with these damn MOSFETs.
9:29
Yeah.
9:30
And obviously they're they seem to be
9:32
more trouble than they're worth.
9:34
So, like for the, you know, you're
9:36
you're building a radio frequency
9:38
amplifier for 50 watts driving a
9:41
moderate impedance load. It's actually a
9:43
lot simpler and more efficient to use
9:45
vacuum tubes.
9:47
Mhm. So in that case rather than the
9:49
MOSFETs there's tubes available that are
9:52
manufactured the 6L6 that would produce
9:55
that 50 watts and wouldn't need it uh
9:58
another stage of amplification to go
10:00
from the master oscillator. The 6L6s
10:04
would amplify it straight
10:06
but it would be it would be twice the
10:08
volume of the transistors. But the
10:10
problem is the transistors are too low
10:12
of an impedance to feed things like you
10:16
know what you're using is you know a
10:18
gaseous discharge tube which requires
10:20
like you know three 300 to 500 volts.
10:24
Mhm.
10:29
Is this related?
10:30
That's the these here.
10:32
Okay.
10:32
So these are the amplifiers for the um
10:35
the input to the uh toic receiving
10:38
equipment.
10:41
So one there's two types. One is uh an
10:44
input amplifier called the receiving
10:46
amplifier.
10:49
So
10:51
these uh will accept the western
10:53
electric tubes that have high voltage
10:55
filaments. So we can run the filaments
10:57
right off the plate supply. So there's
10:59
only one voltage
11:01
because the s the power to this has to
11:03
be phantom through a
11:06
um a neutralizing coil or a phantoming
11:09
coil. So you can only get one voltage.
11:13
So these tube filaments will be this
11:14
will be modified to hook the filaments
11:16
in series and then because the antenna
11:19
is a high impedance it will be delivered
11:23
to this point and the input transformer
11:26
will just be used as a uh an input choke
11:31
and then the output of this goes to the
11:33
input of the line amplifier and then
11:37
this sends a signal up to the shack.
11:39
[Music]
11:43
And in between the input amplifier and
11:45
the line amplifier is the um is the
11:48
network that uh takes the interference
11:52
out. So basically there'll be a lowass
11:55
filter.
11:56
So you can't hook the lowass filter to
11:59
an antenna and you can't hook it to a
12:00
transmission line because of the
12:02
interaction. So because these amplifiers
12:04
have a constant impedance,
12:07
then the filter will work perfectly in
12:09
between the two.
12:11
And because this is vacuum tube stuff,
12:13
it won't be damaged by, you know,
12:16
lightning discharges and things that
12:18
would screw the transistors up. The
12:20
front end has got to be vacuum tube.
12:23
And these are this is very high quality
12:25
and all made by Western Electric. I
12:28
mean, this you couldn't actually like
12:31
works of art.
12:32
They're beautiful inside.
12:36
So, basically, I got to get in there and
12:39
series the filaments. And then there I
12:42
have to take out this uh voltage
12:44
dropping resistor and that's no big
12:45
deal. I just jumper it. Then the
12:47
amplifier is modified.
12:50
So have to make a p make a panel and you
12:54
know fit the connectors and
12:57
put in the related compo components the
13:01
the current regulator for the filaments
13:04
and the phantom transformer and whatever
13:07
protective devices.
13:09
Okay.
13:15
Okay. This is a commercial amp layout.
13:17
So,
13:19
so basically these are your two tube
13:21
centers. So, this one's on the X. This
13:23
one's on the line.
13:24
This is on the center line.
13:26
Yeah. And there's another line here. I
13:27
didn't draw it, but you know, it's in
13:29
the center of the X where you work out
13:31
of. So,
13:32
Mhm.
13:32
So, then surrounding that is spaced 45
13:35
mm are the two strips.
13:37
Center to center.
13:39
Yeah. It's 45 mm from here to here.
13:42
Center.
13:42
Same on that one.
13:44
Yeah.
13:44
Okay. And these we'll put uh we'll put
13:47
these little posts in.
13:50
That's the single little terminals.
13:51
Yeah, we'll put those in. And then the
13:53
capacitors will lay off to the side.
13:55
They won't be actually be here. They'll
13:56
probably be more like in the center. The
13:58
transformer. We'll do the same thing up
14:00
here. Uh this
14:02
But the these two marks are for those
14:04
individuals.
14:05
Yeah. Yeah. The these have been
14:06
measured.
14:07
Okay. So that just is halfway between
14:09
here and here.
14:10
Yeah. It'll sit
14:11
there and there. Well, no. It'll sit off
14:13
to the side because this is a void.
14:15
Okay.
14:16
Okay. It'll sit off to the side.
14:19
So, this it'll uh it that's it'll work.
14:23
So, and then this kind of mounts in this
14:26
little rectangle here. And then these
14:30
these mount on the centers of the
14:32
terminal strips.
14:37
In other words, these are spaced along
14:38
with this. And this is centered on this.
14:42
Okay. So these two pots are centered
14:44
with the terminal strips.
14:45
Terminal strips. Yeah. And uh and this
14:48
is on the you know this particular line
14:51
here.
14:53
Okay.
14:54
And uh
14:57
I don't know. It looks like that worked
14:58
out pretty good.
15:00
Okay.
15:03
Then later on, I mean, you know, if
15:04
you're whatever, you know, we have to
15:07
make this compact the circuit board, you
15:10
know, and all that kind of
15:12
This is just uh
15:14
Okay.
15:15
I mean, this will be nice when it's
15:16
done, but you know, it's going to stand
15:17
about that high in the rack.
15:23
What do you mean? All three panels or
15:24
whatever, right?
15:25
All three panels. The power supply and
15:26
then, you know,
15:28
and power supply will be on this size,
15:29
right?
15:29
That size panel. Yeah.
15:30
And then a three and a half for this.
15:31
and half will be for the input amplifier
15:34
for the 5670.
15:37
And then we're still, you know, we still
15:39
don't know what the preamp
15:41
that's going to be. Mhm.
15:42
Maybe maybe we'll put the preamp on if
15:44
it fits, we'll put the preamp on the
15:49
Well, I don't know. This we got to do
15:51
one step at a time here.
15:52
Okay. We'll drill and mount this then.
15:54
Yeah.
15:55
Okay. Well, we have uh eight panels on
15:58
the bench. Um the four panels on the
16:01
right side is a full embodiment amp
16:04
which is um
16:06
patent pending. Um there's some
16:08
proprietary parts to that. And then the
16:12
one on the left is a simplified
16:16
version which is um intended to be the
16:19
uh commercially available amp. Um I took
16:23
it as far as I could with doing all the
16:25
pan. Um, I basically did the drilling
16:27
and and spacing and everything for the
16:31
parts on all eight panels. Um, Eric
16:35
started by laying out the parts on the
16:37
panel, taping it off and designing the
16:40
layout. So, I did the metal work on that
16:42
and then, uh, these are assembled as
16:45
much as possible. The full embodiment
16:48
amp on the right side.
16:50
Yeah, there's actually there's two parts
16:52
here. So let me
16:53
So this is the um
16:56
this is the input amplifier.
16:59
So it's next step is the wiring and the
17:03
um
17:06
basically putting the whatever resistors
17:08
in there and what have you that that
17:10
goes in last.
17:12
And then
17:14
this is the output amp. This is where
17:16
the the no this is actually this is the
17:19
output amp. This is kind of where the
17:21
novelty exists. It's similar to the
17:23
Macintosh uh amplifier design except
17:26
it's been extended
17:28
to like the next level. So basically all
17:32
of this stuff is Can you turn the camera
17:34
around to these racks here?
17:38
All of this stuff basically is meant to
17:41
be
17:43
bring it over here.
17:46
This is the basic architecture. So this
17:49
is the uh western electric multiplexing
17:52
system and this is going to be kind of
17:55
the design criteria for for everything
17:58
else. So all the auxiliary equipment
18:00
this this amplifier here is intended for
18:03
the earth signal reception. So that
18:06
stuff all mounts in like this.
18:09
And in a bell system installation,
18:12
there's no attempt to
18:15
basically miniaturize or what have you
18:19
to any great extent in that uh when it's
18:23
like this, it's easier to maintain,
18:25
easier to test, replace parts, and they
18:29
typically used 11 and 1/2t high racks in
18:32
high ceiling spaces. So, they had all
18:36
the room in the world to do this kind of
18:38
stuff. So that's basically how this of
18:40
course these are all western electric
18:42
components for the most part
18:47
and we did a lot of pictures and videos
18:49
in the past and so the amplifier that
18:51
Eric is talking about here with these
18:53
panels. This is so related to this get
18:56
this bench cleaned off.
18:59
And um this this is the amplifier in its
19:03
prototype form
19:08
and this will be
19:10
the finished product.
19:14
And of course the power supplies are all
19:15
external and all that. So in this case
19:18
this stuff runs off a station battery.
19:21
Here I have the test power supply. So
19:23
the important thing is to get this bench
19:24
all cleared off to make way for this
19:27
amplifier here.
19:30
And these this is the
19:33
modulator for the um XYZ seismic
19:37
transducer in the mine.
19:41
So these special high-speed relays
19:47
go in here.
19:49
These are probably the last ones on
19:51
Earth.
19:54
They plug into these sockets.
19:58
And then I still have to
20:01
I'm not going to force it right now. See
20:04
what's going on here. These sockets is
20:06
still
20:08
need to be broken in a bit.
20:16
There it goes.
20:18
So this is the drive coil. There'll be I
20:21
have to make a separate hole for the
20:23
wires for that. And what that does is
20:26
that reverses the polarity of the signal
20:28
coming out of the seismic graph at a
20:31
particular musical note.
20:34
So this channel would be a
20:38
which is 426 and 2/3.
20:41
And then uh then this panel is the
20:46
will be the oscillator that drives these
20:50
choppers as they're called. So there'll
20:52
be three Pythagorean notes F A and C and
20:57
this will drive the choppers and then
21:00
the ELF out of the seismograph will be
21:02
turned into a voice frequency tone group
21:05
where you can actually hear in stereo
21:08
the various channels. So that's what
21:09
this is for.
21:13
This goes out at the mine.
21:16
That goes up to the shack.
21:18
And this is the one that is
21:20
developmental. So, it's pretty much the
21:23
same. This is starts with an input
21:24
amplifier.
21:28
And uh that is not particularly of novel
21:30
design, but but it's not customary
21:35
design. This here is not not done yet.
21:37
That's going to be the voltage regulator
21:40
for the voltages and the filters to
21:44
distribute the things and decouple them.
21:46
But this is the power amp.
21:50
Of course, it's pretty blank at this
21:51
point. We're waiting for these
21:52
transformers.
21:54
So, obviously, this this will be
21:57
minimized down to a great extent. This
21:59
this is just developmental,
22:02
but uh but again, it's designed to
22:03
operate in those racks over there. So
22:06
basically when this thing is undergoing
22:09
test, there'll be like a little short
22:11
rack that holds it. And it'd be easy
22:13
with this type of design, it's easy to
22:15
get in and replace parts and do, you
22:18
know, experiments and stuff and make
22:20
sure everything works out okay. And then
22:22
this is pretty much just a conventional
22:24
power supply to run all this stuff. So
22:27
those papers on there, that's for um uh
22:31
not just waiting for offtheshelf little
22:33
transformers and inductors to show up.
22:36
We have to get those custom made. And uh
22:40
some of them I have to get five at a
22:42
time made, some I have to do 10, but but
22:46
they said that we can get one of each so
22:48
we can at least test them. Um but this
22:51
proof of concept right here is it is
22:54
kind of experimental. Um, but once this
22:56
gets put together, then we'll figure
22:58
out, you know, what the final form of
23:01
the whole thing is going to be, like
23:02
what a uh what the end user is going to
23:05
wind up with. It may be something like
23:08
this
23:11
uh rack mount style or maybe it won't.
23:13
So, that's
23:15
to be Yeah, that's got to evolve. But
23:17
the the thing here is is the whole
23:20
engineering concept here is completely
23:24
diametrically opposed to what is
23:28
customary which is cheap Chinese stuff.
23:32
Everything all jammed together on little
23:34
circuit boards doesn't really use the
23:37
proper power supply techniques. None of
23:39
the proper power supply techniques are
23:41
used in anything anymore because nobody
23:44
wants to pay the money for the filter
23:46
choke. They took that out of the system
23:48
and it's just playing absolute hell with
23:50
the power grid. And uh and for a
23:53
highfidelity amp, you know, that that
23:56
creates a situation where you always got
23:57
some kind of buzzing, you know, going on
24:00
or if you have the um these little
24:03
switching supplies and they're whining
24:05
away. This is the old style uh
24:10
filter choke uh condenser
24:13
L network arrangement that uh goes from
24:15
the rectifier
24:17
and the output is pure DC and it does
24:20
not deliver any harmonics back into the
24:22
power line. So this is basically
24:26
this would be something that uh that
24:29
Bell Telephone Laboratories would be
24:31
putting together to go in these racks.
24:34
But even modern um high-end amplifiers,
24:38
most of them are still using
24:40
100-year-old
24:41
methods of voltage gain. And we're not
24:45
going to go into the proprietary stuff,
24:46
but this is amplifying on a different
24:48
principle that almost nobody is
24:52
Yeah. basically all all the amplifiers,
24:56
all of them
24:58
with uh maybe there's an exception.
25:00
They're all still the basic uh uh radio
25:04
trust patents
25:06
from Bell Labs and RCA that go all the
25:08
way back to about 1920.
25:12
And the only thing that's changed is the
25:14
vacuum tubes are much superior than they
25:16
were then. You were lucky if you had a
25:18
transconductance of a thousand. Now you
25:20
can get a transconductance of 10,000.
25:24
So there's vastly improved performance
25:26
and better transformer designs and what
25:28
have but the circuitry and everything's
25:30
all is all the same. There's nothing
25:32
really. The only real break came with
25:34
Macintosh
25:36
and they came up with a a unique type of
25:38
power amplifier.
25:40
But for some reason they failed to
25:42
extend that concept and used all this
25:45
exotic circuitry which was absolutely
25:47
unnecessary. there was a way to do it
25:49
just with transformers and it I tried
25:52
that at my laboratory at Landers and it
25:54
worked out quite well. So,
25:58
and then there's another amplifier
26:01
is uh is a direct copy of a western
26:06
electric
26:07
uh PA amplifier design which uh uses
26:12
some novel circuit uh configurations
26:16
and all standard tubes that are
26:18
available still you know because of the
26:21
guitar people and you know what have you
26:24
musical instruments that's kind of where
26:26
This thing is gravitating. It's more
26:27
towards a musical instrument amplifier
26:30
rather than hi-fi amplifier.
26:33
I don't know how small we can get this
26:35
thing because the power supply is uh you
26:39
know there's another big transformer
26:40
underneath. Another one here. It's hard
26:42
to miniaturaturize that. You can't have
26:44
all this stuff around the sensitive
26:46
electronics. So the power supply is
26:48
usually something that sits on the floor
26:51
or you know in some corner and then the
26:53
amplifier. This this can be
26:55
miniaturaturized down rather
26:57
extensively, but it's still going to be,
27:00
you know, for a low power amplifier.
27:03
It's still going to occupy quite a bit
27:05
of space, but it'll be uh
27:10
the thing about this particular type of
27:11
amplifier is it allows for overloading
27:16
without distortion.
27:18
So music comes in these peaks that uh
27:23
some of the peaks could be 10 dB
27:25
stronger than the average of the signal.
27:27
You don't see those on the meter. It it
27:30
requires enormous power expenditure of
27:32
the amplifier. So the thing is is to
27:34
make sure that the amplifier doesn't
27:36
distort when it gets hit with those
27:38
peaks other than the peak will be
27:41
basically somewhat distorted.
27:45
So basically the peaks
27:48
can pretty much
27:51
it's kind of hard to describe because
27:52
it's a nonlinear process but um but the
27:55
amplifier can continue to deliver power
27:58
past its overload point without really
28:01
going into heavy distortion mode.
28:04
So that's the one advantage of this
28:06
particular type amplifier design and
28:08
that's why it's being used for the earth
28:10
signal reception because the peaks there
28:12
you know can be a lightning strike a
28:14
mile away from the antenna compared to
28:17
something you know coming out of uh the
28:20
ionosphere 5,000 mi away. So amplifiers
28:24
have to accommodate that without making
28:25
a mess.
28:27
So that's basically
28:31
that's pretty much what's going on here.
28:32
So those are the three things. So this
28:35
commercial amplifier, then the uh
28:37
seismic
28:39
uh modulator, and then the um the full
28:44
amplifier design, which this one I want
28:48
to scale up to about 300 watts using u
28:52
the 811A and the 812A, which are still
28:56
available. That's always been my dream.
28:58
There actually is a Macintosh version, a
29:01
300 watt Macintosh vacuum tube amp video
29:05
on YouTube
29:07
that uh it's absolutely remarkable piece
29:09
of equipment. Takes up about half of a
29:13
rack. That's that's what I'm really
29:15
after is making one of those with some
29:17
of these improved circuit techniques
29:20
because I have a set of Bose 901
29:22
speakers and I'd like them to be driven
29:23
by vacuum tubes.
29:26
So other than that I think all the
29:28
various details and what have you are in
29:30
other videos and presentations.