Recording

Vintage Microphones, Part 1: The Neumann U47

Note: This article is a segment of a multi-part series detailing the history and technology behind some of the industry's most beloved and treasured vintage

Note: This article is a segment of a multi-part series detailing
the history and technology behind some of the industry's most beloved
and treasured vintage microphones. The author—the late Stephen
Paul—founded Stephen Paul Audio (www.spaudio.com), a firm that continues to offer
specialized services in the restoration and hot-rodding of classic
microphones. One of Paul's most enduring contributions to modern
microphone technology was his research and application of ultrathin
diaphragms (sometimes less than a single micron thick), which spawned a
renaissance in the development of improved microphones from companies
worldwide.

First published in 1989, this article series was one of Mix's
most popular back-issue requests, and 15 years later, we are pleased to
bring this work to a new generation of microphone fans.

Also, as a bonus, brochures of all the microphones in this article
and other classic mics (in .pdf format) are available on this Website.—George Petersen


Ah, the great vocal mics. What music they make. But how come
it’s so hard to choose the right mic for the singer at hand? One
performer’s meat is another’s ugghh! This is certainly a
question that has troubled us since the dawn of time. Since we humans
first wandered out of the cave, we have gazed at the night sky and
wondered, “Do I use a C-12 or a 47?”

We will desperately try to answer this age-old question using a
combination of high technology, modern measurement methods and sheer
guts. No hearsay in this corner.

Okay, so what’s the first thing we have to consider? These
days, “How much does it cost?” might come to mind, what
with C-12s and the like going for $2,800 up into the $3,500 bracket.
[At 1989 prices!—Ed.] But we will pretend we didn’t
hear that crass question and simply concentrate on what it is that
makes these damn things so expensive. My old friend Dick Rosmini is
fond of saying, “The only thing that don’t wear out on a
Rolls-Royce is the charisma.” I suppose the same thing might be
said of a U47. Actually, in this case, I’m happy to say there are
other things that don’t wear out in a classic, vintage tube
microphone. Not the least of which is the thrill from the shock you can
get if you pull the mic off the stand before the supply has been shut
down and your hand strays into the wrong grotto. Whew!

Now then, what was the question? Ah yes. Things to consider. I
suppose the thing to do would be to go on a little excursion into The
Unknown. (That is, for people who don’t build mics, it’s
unknown.) Let’s look at some of these old classics and get
acquainted with the antique wonders of their secret interiors.

The first microphone we will investigate is that grizzled old King
of the Kondensers, the Neumann U47. Ol’ grandpa’s a bit
timeworn and his condensers are dusty, but he’s still in the
saddle, by gum. Ee-hah! This is truly one of the great vocal mics of
all time. The neat subjective things about these U47s is that they go
off like a Howitzer. I mean, we are talking about serious
pressure-gradient proximity effect here; only the true
“size” junkies need apply. If that last statement sounds
odd coming from a more-or-less scientific kind of guy, I think a little
elucidation is in order.

TWO DOORS
What is a “pressure gradient microphone,” anyway?
That’s a darn good question. Glad I asked it. A pressure
gradient-operated mic is one in which the difference in pressure
between two physical points provides the driving force for the
diaphragm. Or so the textbooks tell us. It’s actually a hell of a
lot more complex than that, but initially this explanation will do.

For instance, say we are in a house on top of a hill, and a slight
breeze is blowing broadside. If we stand outdoors off to the side, we
feel only this gentle movement. Now, let’s imagine there are two
doors in the house that are on the front and rear broadside,
respectively. If we open both doors and stand inside the house, we find
that the breeze seems to be much more active. This is caused by several
factors—one is that the house is an obstacle to the wind, and the
pressure on the side of the house that faces the wind is greater than
the pressure on the rear side, which is in the shadow of the house.

Have you ever driven behind a semi to save fuel? Dead calm right in
back, ain’t it? Same principle. This difference in pressure from
front to back (this is also a phase difference, as it involves a
transit time delay) creates a driving gradient across the house that
can be felt strongly inside. The wind coming in is accelerated into the
lower pressure area vented by the rear door, just as the high pressure
in a helium tank floods into the relative void of your lungs and makes
your voice go all funny. This effect (the flooding in, not the funny
voice), coupled with the increase in velocity that occurs when the
stream encounters a constricting area such as our door, drives the
accelerated wind through the house at a higher velocity than in the
open area around it. The condition holds true for wind moving in either
direction, as this is what constitutes the AC component of the sound
pressure. Essentially, this is how the pressure gradient microphone
works in principle. Vents in the back plate approximate the doors in
our example, and the capsule itself is the house. The wind is sound
pressure, and the gradient drives the little engine of the stretched
membrane. Now that we’ve got that under our belts, we can look
with a little more understanding at the U47, the granddaddy of them
all. (There was a great-granddaddy, but that’s really getting
esoteric.)

SONIC FOOTPRINTS ON THE SANDS OF TIME
The first thing our delicate little
sounds encounter when they hit this massive edifice is the famous 47
grille. Famous because the steaks are huge and the drinks are great.
Whoops, that’s the 47th Street Grill. Sorry. Good place for
lunch, though. Anyway, lest any of you doubt that the grille is a major
influence on the 47’s sound, see Fig. 1, which shows the
never-before-published difference between the 47 as measured with (top)
and without (bottom) its grille. Astonishing, isn’t it? It is
important to note that not all mics are affected the same way by their
grilles, and therefore, this particular change in the curve only
applies to the 47 design.

Imagine that. And all this time we though it was the emblem that was
the problem! There are people who think the 47 that says
“Telefunken” on it is different than the one that says
“Neumann.” Well, they’re right. One says Telefunken
and one says Neumann. Other than that, I’m afraid, the
differences could only be detected with a scanning-tunneling electron
microscope most likely, and even I’m not crazy enough to think
that would change the sound.

Let’s look closely at some of these differences. This graph
has been printed on a linear chart to expand the range above 1 kHz.
This means that 20 Hz is at the extreme left of the scale, and the
dotted divisions represent 1kHz increments. The dot’s vertical
value is 1 dB. It is also important to note that the U47 shown has been
modified with a 3-micron diaphragm. As a result, the extreme high
frequencies above 15 kHz are supported, unlike the 7-micron stock model
where response falls rapidly above this point. It’s interesting
that with the grille on we find some frequencies actually higher in
amplitude, as well as the decreases we would intuitively expect.

The first thing we can see is that the grille allows the 5kHz
presence area to sail on through, but depresses the sibilance range in
the critical 9kHz slot. Now look at the 10kHz area and notice the
additional energy—up to 1.5 dB of increase at this frequency.
When we remove the grille, the response of the capsule itself can be
seen. This is represented by the lower sweep in Fig. 1. Also, the
frequencies above approximately 11.5 kHz are depressed by the action of
the fine mesh in the grille. And a fine mesh it is, too. The capillary
activity of this very closely spaced wire cloth reaches maximum
attenuation as we near and exceed 20 kHz.

Let’s pause here briefly to consider the fact that all
microphones currently in existence are highly complex echo chambers
inside. Some, just like other acoustic echo chambers (including all
acoustic musical instruments), have a wonderful character, while some
are inherently annoying, or less useful for certain jobs. This is
certainly one thing that has a tremendous influence on our reaction to
a given microphone. The interior reflections that occur as complex
sound waves hit the surface of the capsule and the capsule holder, and
are diffracted (bent) and dispersed around the physical objects in the
microphone head, create a fantastically complex set of interference
patterns. Some of these are measurable in the frequency domain. These
characteristics are part of what give the 47 its “sonic
footprint” and make it desirable for its presence.

By modern design criteria these grosser aberrations are considered
“defects.” Incredible isn’t it? In the '40s this
system was hailed as the ultimate solution to all our miking problems;
it is now worth up to five times its original selling price in stock
trim, and these are defects! This is why life is beautiful. This is not
to say they aren’t defects. Indeed, they are. But they are
beautiful defects that we love, and so they transcend the normal
connotations of the term.

THOSE GOLDEN DIAPHRAGMS
So much for the grille. Once the sound gets past the grille, it moves
on to that temple of ultimate mysteries, the capsule. Fig. 2 is a
photograph of the back plate of the U47 capsule. This little baby
contains more than 100 holes per side, located with great precision.
The body is made of brass, and the diaphragm fits over it with a minute
spacing of less than two-thousandths of an inch. The diaphragm is made
of polyester film that is places in a vacuum chamber where a thin layer
of gold is evaporated onto its surface in a masked-off circular
pattern. This places a conductive layer on top of the plastic so a
voltage can be impressed thereon, thus forming one of the plates of a
capacitor, which after all is the principle of operation here. The now-conductive film is
stretched to a specified tension and glued to a ring, which is then
attached to the back plate with screws. Imagine doing this for a
living!

The photograph to the right shows a Stephen Paul Audio 3-micron
diaphragm after mounting on its ring, but before being trimmed. Note
the masked-off circle of gold in the center.

DON’T STAND, DON’T STAND, DON’T STAND SO CLOSE
TO ME

Another contributor to the 47’s characteristic sound is proximity
effect. Most of us know that “proximity effect” is the
boost in bass response that occurs when a sound source comes within
inches of any gradient-operated mic, be it condenser, dynamic, ribbon
or what-have-you. So far, so good. “But…why?” comes
the inevitable, wide-eyed question.

Here’s my best shot at an explanation. It gets a little like
the cabinet of Dr. Caligari in here, so leave a trail of bread crumbs
to keep from getting lost. Remember that our microphone is driven by
the gradient, or the difference in pressure caused by the time delay
formed by the physical distance between the front and rear of the
capsule. (If you don’t remember this, please see a memory
specialist.) The gradient’s effect is straightforward when we are
dealing with a plane wave. Waves emitted by a point source, such as a
humanoid, generally become plane in shape when we are some distance
from the source. Plane means like a flat surface. The entire wave hits
at the same moment in time.

Now, sound pressure decreases as the reciprocal of the distance form
the source, or in other words, 1/d, where “d” is distance.
When we are in proximity to the source, however, the waves start out
with a spherical geometry. This means that not only will there be a
pressure difference resulting from the phase delay of the front-to-rear
capsule distance, but also a pressure gradient created by the physical
separation in space of the protruding, curved front of one wave, and
the bent-away side of the next one, which is slightly closer to the
source at the moment. Because one wave is higher pressure due to its
lesser distance from the source, a difference in pressure—or a
gradient—is created between the waves.

The reason the low end is affected most is that the phase-related
gradient is frequency dependent (small at low frequencies and rising
with rising frequency, until the length of the path around the capsule
equals one-half wavelength [Theta = 180 degrees] and it starts to
fall), while the distance-related gradient is non-frequency dependent;
as we approach the mic, the phase-driven gradient (which is a function
of the particle velocity) changes very little, while the distance
gradient increases with decreasing distance from the microphone. As the
distance-driven gradient increases, so does the low-end response where
the phase-gradient drive is weak. Woof! I don’t think I could say
that again! (Where are those bread crumbs?) The degree of boost in
cardioid can become quite extreme when we are very close, up to 15 dB
at 100 Hz! Some mics built recently try to control this effect, but
have not won a great following with the vocalist crowd.

Proximity is one of the U47’s greatest defects. Or should I
say, one of the most popular defects. The 47 has an earthshaking boost
in the bass region when it’s switched to the cardioid, or
unidirectional, characteristic. When we switch to omni, this effect is
greatly diminished since we have electrically shorted out most of the
gradient effect by connecting the rear capsule in phase (in polarity,
actually) and summing its energy into the amplifier. If, as is possible
on the U48 model, we switch to figure-8—essentially two cardioid
mics back to back—we get an even larger amount of proximity, as
we have summed the output of both sides in their directional mode.

OF TUBES AND SHIPS AND SEALING WAX
So there’s a basic look at the U47’s operation principle.
The only mysteries left are the tube and the variations on the tube.
Many people ask me, “That big black thing in there, is that the
tube? How come it doesn’t look like a tube?” For the
answer, let’s refer to Fig. 4. It shows the original electronics
in the 47. Just for the archives, the tube in the 47 is famous first
for the reason just quoted—that it looks odd—and second for
the price of a replacement, which can run up to $500 depending on who
you ask. $500 for a tube? It does sound kind of unbelievable. Yet some
people pay more than that for a bottle of wine. And you can’t
even cut a vocal with a bottle of wine. The VF-14 (that’s the
name of our little black friend) was once a very common tube used in
field radios during World War II. After the war, not much else was
left, and it happened to sound good, so Neumann decided to use them in
the 47. Of course, the factory selected them for low noise, and you can
tell you have a factory tube if it has a little white “M”
stamped on it. If it doesn’t, then at some point in the
mic’s history someone pulled the old switcheroo.

The sad thing is that the tube was never manufactured after the war,
although a strange recent rumor has some Italian company tooling up to
make them again. But I have seen neither hide nor hair of one, so who
knows. For a replacement, Neumann offered a “nuvistor” kit,
which plugs into the old socket, but the sound has never really caught
on. The nuvistor is a super-miniature tube in a metal jacket, and a
triode to boot, which is why they sound questionable. The 47’s
output transformer was designed for the VF-14, which is a pentode and
has a plate impedance of only 8,000 ohms or so. The little triode has
an impedance that goes up into the 30k-ohm region, and the bass
response rolls off. I can just hear the fans moaning, “Oh no! Not
the bass response… isn’t that the reason I bought this
thing in the first place?” Well, yes. We can effect a healthy
improvement just by going to a more correctly matched transformer.

In addition, my company developed a complete, contemporary upgrade
that uses a modern tube (shown in Fig. 5) and a Jensen output
transformer dragged out of Deane at great personal peril. This brings
back the 47 sound that we all know and love, and it offers the benefits
of lower noise and distortion, and greater dynamic range, than the
original.

Well, I think we’ve exhausted all the useful information on
the U47, and it’s time to move on to some other old friends: the
M49, U67 and U87. Until then, may your vocal sounds be forever
well-colored.


Stephen Paul is the founder of Stephen Paul Audio, located in
Burbank, Calif.

Want to read more stories like this?
Get our Free Newsletter Here!