Tascam M-35 Manual

TASCAM
TEAC Production Products
Audio Mixer
OPERATION/MAINTENANCE

IN'TRODUCTION T0THE MODEL 35
The Model 35 is an audio mixing console design-
ed to satisfy the requirements of modern multi-
channel recording. Many of the auxiliary mixing
systems needed are built-inand can be re-routed
to do more than one job. Fast, convenient and
complete operation with
4- or 8-track recorders
can usually be accomplishedwithout re-patching.
However, the process of multichannel recording
is constantly changing, growing more complex as
an art with each advance in technology. Your
signal processing needs may require
a unique
arrangement of subsystems. No console has ever
been built so large and complete in
its routing
that it could solve every imaginable problem
with one button. Someone will always be able to
come up with that unusual situation requiring
"just one more mix". Inorder tocopewith thesep
unpredictable requirements, patch points are
provided throughout al1 signal pathways on the
M-35.
As our mixing console becomes more flexible,
the amount of time needed to understand the
available function increases
as well. The main
signal path from mic in to line out is still fairly
straightforward. The requirements have not
changed much since the days of "mono", butthe
routing for effects sends, cue feeds, and stereo
monitoring can be hardtovisualize. 'rhe beginner
often overlooks the significance of connections
that would be immediately obvious to the expe-
rienced recording engineer. Ifyou expect to find
that "extra rnix" quickly, you must be prepared
tostudy the layout of the M-35 thoroughly.
Input module layout including back panel
In most instances, the physical arrangement of As an example, if the controlson an input mod-
the controls on the top pane1 has very little to ule actually followed the order in which they are
do with the sequence of electronic parts inside. wired, the module top pane1 would look Iikethis.
The actual wiring order
is the information you We'll put the jacks on the top, as well as the
need to understandto use the M -35successfully~ switches and faders.

While this arrangernent ofcontrolsrnight helpthe can be used successfully. So along with thedocu-
beginner to understand the flow of signal in the rnentation you will need for service (schernatic
rnodule,
it would be very inconvenient to oper- diagrarns, rnother-board layouts and rnechanical
ate. StiII, thewiringsequence rnustbe understood disassernbly inforarntion), we includeasirnplified
before the more cornplex functions of the M-35P electrical sequence chart called a block diagrarn.
16 15 17
Direct @ ml - P
9 solo
This drawing shows al1 the controls, switches,
arnplifying stages and connectors intheir proper
order. Learning to read itwill providetheanswers
toany questions about what cornes where on the
inside. Even though the block diagrarn can indi-
cate what
is available in the way of extra circuit
flexibility,
it can't explain why a connection or
switch has been included, or suggest a standard
layout. In the following sections of this rnanual,
we will do our best to describe the individua1
functionsand controls of the M-35,and howthey
can be arranged in more than one sequence; but,
your rnixing needs rnay be best served by an ar-
rangernent of inputs and sub-systernconnections
you work out for yourself.
Tobegin, we'll start with some basic inforrnation
about sound and the nurnbering systerns used to
describe levels in and out of the equiprnent and
impedance-what the terrn rneans and how to
dea1 with the details when you rnust connect
frorn our gear to other equiprnent. Many aspects
will be discussed in the rnost basic language we
can use. There is a vast arnount of inforrnation
available to the beginning sound rnixer but rnuch
of it
is not basic enough to be easily understood,
or
it assurnes that the reader has an engineering
or scientific backgroundand wil
l be interested in
"the rnath". Practical "rules of thurnb" for the
novice are not generally available. Sornething be-7'
14
- from subrnaster "Ta~e inn*
tween a picture of the outside of the unitand a
complete rnathernatical analysis of the circuits
inside
is needed. You don't have to builda mixer
frorn scratch, you just need to know how to
operate one.
However, some nurnbers are unavoidable. The
M-35 rnixer does nothing useful without being
connected to quite a lot of sophisticated gear.
Mics, tape recorders, power amps, and loud-
speakers al1 play a part in the process of rnixingl
recording and each piece of gear has its own re-
quirernents and problerns. We have tried to rnake
this rnanual as sirnple as technology will allow.
Each section or topic will give you some basic
instruction in the terrninology used in the pro-
cess of rnixing as well as a list of what pluggoes
into which jack.
Whenever possible, the scientific terrns havebeen
related to understandable cornrnon references.
Understanding what isgoingon insideyourequip-
rnent will help you irnprove your sound. Think
of this rnanual as a reference book. You won't
need al1 of what is here to begin, and itcertainly
is not necessary to rnernorize it, but do try to
find tirne to read
it carefully at least once. That
way you will be farniliar with
its contents. Ifyou
need the nurnbers, they will be there waiting.
Good luck with your sound.

'THE DB; WHO, WHAT, WHY
No matter what happens to the signal while it is
being processed,
it will eventually be heard once
again by
a human ear. So the process of convert-
ing a sound to an electrical quantity and back to
sound again must follow the logic of humanhear-
ing.
The first group of scientists and engineers todea1
with the problems of understanding how the ear
works were telephone company researchers,and
the results of their investigations form the foun-
dation of al1 the measurement systems we use in
audio today. The folks at Bell Laboratories get
the credit for finding out how we judge sound
power, how quiet asound an average person can
hear, and almost al1 of the many other details
about sound you must knowbeforeyoucan work
with it
successfully.
From this basic research, Bell Labs developed a
system of units that could beapplied toal1 phases
of the system. Sound traveling on wires as elec-
trical energy, sound on tape as magnetic energy,
sound in air; anyplace that sound
is, or has been
stored as energy until some future time when it
will again be sound, can be described by using
the human ear-related system of numberscalled
"bels" in honor of Alexander Graham Bell, the
inventor of the telephone.
What
is a bel and what does itstand for?
Itmeans, very simply, twice as loud tothe human
ear. Twice
as loud as what?Anobviousquestion.
The bel is always a comparison between two
things. No matter what system of units of meas-
ure you are working with at the time, you must
always state a value as a referencebeforeyou can
compare another value to it by using bels, volts,
dynes, webers
- it doesn't matter, a bel, or ear-
related statement of "twice as loud" is alwaysa
ratio, not an absolute number. Unless a zero, or
"no difference" point
is placed somewhere, no
comparison is possible.
There are many positive and definite statements
of reference in use today. But beforewe go over
them, we should divide the "bel" into smaller
units. "Twice
as loud" will be a littlecrude to be
used al1 the time. How about one tenth of a bel?
Okay, the decibel
it is, and O means "no differ-
ente, same as the reference". It seldom means
"nothing". Now, if you double the power,
is that
twice
as loud?No, it is only 3dB more sound. If
you double an electrical voltage, is it twice as
loud? No,
it is only 6dB more sound. The unit
quantities must follow nonlinear progressionsto
satisfy the ears' demand.
Remember, decibels follow the ears. All other
quantities of measure must be increased inwhat-
ever units necessary to satisfy the human require-
ments, and may not be easy to visualize. Sound
in air, our beginning reference,
is the least sound
the human ear (young men) can detect
at 1000
to 4000Hz. Bell Labs measured this value to be
.O002 microbar, so we say OdB = .O002 microbars
and work our way up from the bottom, or "no
perceivable sound to humans" point. Here
is a
chart of sounds and their ratings in dB, using
.O002 microbar pressure change in air as our re-
ference for "OdB".
10,000
22 nch bass drum
mic inride drum
pnaredruni
l inch
t140 1 I powr?Ilc voice
at 1 inch Ircream)
10 newtoiir per 100 dvner per
100
100'
-1 newton per 1 dyne per
rquare rneter square cm. l niicrobar Auerage
converratlun
-- -
... Home in citv,cantinuour background
I microbar 1
: noire*(carr, rubwavr. rtreet nolre)
- - - - - - -
t"
" Homencitv at night
.O1 microbar
Isoiared recording or TV studio
Open field, night, no wind
-001 microbar
icricketr. nrect noirer, etc.)

We should also make a point of mentioning that
the maximum number on this chart represents
"peak power" and not average power. The rea-
son? Consider if even some monetary part of
your recording is distorted, it will force a re-re-
cording and
it is wisest to be prepared for the
highest values and pressure even if they only
happen "once in
a while". Onthis point, statistics
are not going to be useful, the average sound
pressure
is not the whole story. 'rhe words them-
selves can be used
as an example. Say the word
"statistics" close to the mic while watching the
meters and the peak LED level detector. Then
say the word "average". What you are likely to
see are two good examples of the problems en-
countered in the "real world" of recording. 'rhe
strong peaks in the "s" and "t" sounds will pro-
bably cause the LED's to flash long before the
VU meter reads anywhere near "zero" while the
vowel sounds that make up the word "average"
will cause no such drastic action.
To allow peaks to pass undistorted through
a
chain of audio parts, the individua1 gain stages
must
al1 have a large reserve capability. If the av-
erage is X than X + 20dB is usually safe for
speech, but extremely percussive sounds may re-
quire as much as 90dB of "reserve" to insuregood
results. Woodblocks, castanets, latin percussion
(guido, afuche) are good examples of this short
term violence that will show
a large difference
between "LED flash" and actual meter movement.
When you are dealing with this kind of sound,
believethe LED,
it is telling you the truth.
Since the reference is assumed to be the lowest
possible audible value, dB spl is almost always
positive, and correctly written should have
a +
sign in front of the number. But it is frequently
omitted. Negative dB spl would indicateso lowan
energy value
as to be of interest to a scientist try-
ing to record one cricket at 1,000 yds. distance,
and is of no significance to the multichannel
recordist. Far more to the point
is the question
"What is
a microbar?" It is a unit of measure-
ment related to atmospheric pressure and al-
though it
is extremely small, it must be divided
down quite
a lot before it will indicate the mini-
mum pressure change in air that we consider
minimum audible sound. This will give you
a
better idea of the sensitivity of the human ear.
One microbar of pressure change is slightly less
than one millionth of an atmosphere, and you
can find it on our chart as 74 dB spl. It is not
terribly loud, but
it is certainly not hard to hear.
As
a matter of fact, it represents the average
power of conversational speech
at 6 feet. This
level is also used by the phonecompany to define
norma1earpiece volume on a standardtelephone.
Nowthink about that minimumaudible threshold
again: .O002 microbar.
That's two ten thousandths of
a millionth partof
one atmosphere
!
This breakdown of one reference is not given
just to amaze you, or even to provide
a feel for
the quantity of power that moderate levels of
sound represent. Rather itis intended to explain
the reason we are saddled with a ratiollogarithm
measurement system for audio. Adding and sub-
tracting multi-digit numbers might be easy in
this age of pocket calculators, but in the 1920's
when the phone company began
its research into
sound and the human ear,
a more easily handled
system of numbersbecame an absolute necessity.
Conveniente for the scientist and practical en-
gineer, however, has left us with a system that
requires
a great dea1 of complex explanation be-
fore you can read and correctly interpret a "spec
sheet" for almost any piece of gear.
Here are the formulae for unit increment, but
they are necessary only for designers. And unless
you build your own gear, you won't have to dea1
with them. For power (watts) increase or loss,
calculate by the following equation:
10 LOG,,
For voltage, current or pressure calculations:
20 LOG,, v2
v1
= N (dB)
One whole atmosphere, 14.70 pounds per square
inch, equals 1.01325 bars. So one whole atmos-
phere in microbars comes out to be 1,013,250.

Once we have this chart, we can see the differ-
ente between the way humans perceive sound
and the amount of force it takes to change air
pressure. Unfortunately, the result is notasimple
"twice
as much pressure" of sound to be heard
as "twice as loud". Ifyou plotdecibels as theeven
divisions on
a graph, the unit increase you need
is a very funny curve.
VOLTAGE, CURRENT /
OF PRESSURE
v1
dB = 20 iog,, E
wav
5 10 15 20 25 30
-
increase
RIS~ in even 1 dB Unit
This is how the ear works, andwe must adapt our
system to it.We have no choice ifwe expect our
loudspeaker to produce
a sound that resembles
the original sound we begin with. The high sensi-
tivity to sound of the human ear produces a
strong "energy" illusion that has confused listen-
ers since early times. How powerful are the loud-
est sounds of music in real power?Can sound be
used
as a source of energy to do useful work,
such
as operating acar? For any normally "loud"
sound the answer
is, regrettably, no! Perhaps not
so regrettably, consider what would happen if
one pound of pressure was applied not to your
head, but directly to your inner ear. One pound
of air pressure variation
is 170dB spl ! This a-
mount of "power" might do some useful work -
but not much, it's still only one pound and to
make use of it you will have to stand one mile
away or you will go deaf immediately.
Ifwe reduce our sound power torealistic musical
values, we will not be injured, but we will have
almost nothing (in real power terms) to run the
mic with! This low available energy is the reason
that high gain amplifiers are required for micro-
phones.
When we take
a microphone and "pick up" the
sound, we do have some leeway indeciding how
much energy we must have in order to operate
the electrical part of our system. Ifwecan decide
that we don't have to truly hear the signal while
we are processing it from point to point and we
can wait until the electronic devices have done L
al1 their routing and switching before we need
audible sound, we can lower the power of the
signal. What is
a good value for a reference here?
Well, we need tohave enough energy so that the
signal
is not obscured by hiss, hum,buzzor other
unpleasant things we don't want, but not so high
that
it costs a fortune in "juice" or electrical
power. This was
a bigconsideration for the tele-
phone company.
They now have the world's biggestaudio mixing
system, and even when they started out, electric-
ity was not free. They set their electrical power
signal reference
as low as was practical at the
time, and it has lowered over the years as elec-
tronic equipment has gotten better. In 1939 the
telephone company, radio broadcasting, and re-
cording industry got together and standardized 1
milliwatt of power
as OdBm, and this is still the
standard of related industries. Thus,
a OdBm sig-
nal at a 600ohm line impedance will present a
voltage of 0.775 volts.
Once again, we owe you an explanation. Why
does
it say ZERO on the meter? What isanohm?
Why 600 of them and not some other value?
''
What's a volt? Let's look at one thing at a time.
1. The logic of ZERO on the meter is another
hangover from the telephone company prac-
tice. When you start
a phone call in Califor-
nia, the significant information to a telephone
company technician in Boston is
- did the
signal level drop? Ifso, how much? When the
meter says ZERO
it indicates (to the phone
company) that there has been no loss in the
transmission, and al1
is well. The reference
level is one milli-wattof power, butthe gainor
loss is in the information the meter was sup-
posed to display, so the logic of ZERO made
good sense, and that's what they put on the
dial. We still use it even though it's not logi-
cal for anything else, and the idea of
a refer-
ence level described as a "no loss" ZERO, no
matter what actual power
is being measured is
so firmly set in the minds of everyone in the
audio world that
it is probably never going to
change.
2. One ohm is a unit of resistance to the passage
of electrical energy. The exact reasons for the
.
choice of 600 ohms as a standard are con-
nected to the demands of the circuits used
.

for long distance transmission and are not
simple or easy to explain. Suffice it to say
that the worst possible thing you can do to
a
piece of electronic equipment is to lower the
resistance
it is expected to work into (the
load). The lower the number of ohms, the
harder it is to design
a stable circuit. When
you think about "load", the truth
is just the
opposite of what you might expect! O ohms
is
a "short circuit", no resistance to the pas-
sage of signal. If this condition occurs before
your signal gets from California to Boston,
you won't be able to talk- the circuit didn't
"get there",
it "shorted out". Once again, tele-
phone company logic has entered the language
on a permanent basis. Unless the value for
ohms
is infinity(nocontact, nopossibleenergy
flow) you will be better off with a higher
value, and many working electronic devices
have input numbers in the millions or billions
of ohms.
3. A volt is
a unit of electrical pressure, and by
itself is not enough to describe the electrical
power available. To give you an analogy -
that may help, you can think of water in a
hose. The pressure is not the amount of water,
and fast flow will depend upon the size of
the hose (impedance or resistance)
as well.
Increase the size of the pipe (lower the resist-
ance, or Z) and pressure (volts) will drop un-
less you make more water (current) available
to keep up the demand. This analogy works
fairly well for DC current and voltage, but
alternating current asks you to imagine the
water running in and out of the nozzle
at
whatever frequency your "circuit" is working
at, and is harder to use a menta1 aid. Water
has never been known to flow out of
a pipe
at 10,000 cycles per second.
This reference level for a starting point has been
used by radio, television, and many other groups
in audio because the telephone company was the
largest buyer for audio equipment. Most of the
companies that builtthe gear started outworking
for the phone company and new audio indus-
trie~,
as they came along, found it economica1to
use
as much of the ready-to-handstuff as they
could, even though they were not routingsignals
from one end of the world to the other.
Must we use this telephone standard for record-
ing? Its use inaudio has been so widespread that
many people tiave assumed that
it was the only
choice for quality audio. Notso.
A 600 ohm, 3-wiretransformer-isolated circuit
is
a necessity for the telephone company, but the
primary reason
it is used has nothingtodo with
audio quality. It is noise, hum and buzz rejec-
tion in really long line operation (hundreds and
hundreds of miles).
Quality audio does not demand 600 ohm, 3-wire
circuitry. In fact, when shielding and isolation
are not the major consideration, there are big ad-
vantages in using the 2-wire system that go well
beyond cost reduction. It is, as a system, in-
herently capable of much better performance
than 3-wiretransformer-isolated circuits.
Since TASCAIVI M-35 mixer is designed to route
a signal from
a mic to a recorder, we think that
the 2-wire system
is a wise choice. The interna-
tionally accepted standard (IEC) for electronics
of this kinduses a voltage referencewithout speci-
fying the exact load it is expected to drive. The
reference
is this:
O = l volt
This is now the preferred reference for al1 elec-
tronic work except for the telephone company
and some parts of the radio and television busi-
ness. Long distance electronic transmission still
is in need of the 600-ohm standard.
Ifyour test gear has provision for inserting a 600
ohm load, be cure the load is not used when
working on TASCAM equipment.
Now that we have given a reference for our "0"
point, we can print the funny curve again, with
numbers on it, and you can read voltages to go
along with the changes in dB.
,"
20 log,,+ curve

All electronic parts, including cables and non- you are making connections in your mixing sys-
powered devices (mics, passive mixersand such), tem. The outputs of circuits have an impedance
have impedance, measurable in ohms (symbol S2 rating and so do inputs. What's good? What
or Z). Impedance
is the total opposition a part values are best? It depends on the direction of
presents to the flow of signal, and
it's important signal flow, and intheory,
it looks like this:
to understand some things about this value when
OUTPUTS
- plug into
- INPUTS
It is generally said that the output imped-
ance (Z)should be
as low as possible. 100
ohms, 10 ohms. The lower, the better, in
theory. A circuit with
a low output imped-
ance willofferalow resistancetothe passage
of signal, and thus will be able to supply
many multiple connections without
a loss
in performance or a voltage drop inany part
of the total signal pathway. Low impedance
values can be achieved economically by us-
ing transistors and integrated circuits, but
other considerations are still a problem in
practice, such as:
1. The practical power supply
is not infi-
nitely large. At some point, even if the
circuit
is capable of supplying more en-
ergy you will run out of "juice".
Inputs should have very high impedance
nurnbers,
as high as possible (100,000 ohms
1million ohms, more, if
it can bearranged).
A high resistance to the flow of signal at
first sounds bad, but you are not going to
build the gear. If the designer
tells you his
input will work properly and has no need
for a large amount of signal, you canassume
that he means what he says. For you, a
high input impedance isanunalloyed virtue.
Itmeans that the circuit will do its job with
a minimurn of electrical energy as a begin-
ning. The most "economical" electronic de-
vices in use today have input impedancesof
many millions of ohms, test gear for ex-
ample, voltrneters of good quality must not
draw signal away from what they are meas-
uring, or they will disturb the proper opera-
2. Long before this happens, you may burn tion of the circuit. A design engineer needs
out other parts of the circuit. Theoutput to
see what is going on inhisdesignwithout
impedance may be close to the theoreti- destroying
it, so he must have an "efficient"
cally ideal "ohms" but many parts in the device to measure with.
practical circuit are not. Passing energy
through a resistance generates heat and
too much current will literally burn parts
right off the circuit card if steps are not
taken to prevent catastrophic failure.
3. Even if the circuit does notdestroy itself,
too high a demand for current may seri-
ously affect the quality of the audio. Dis-
tortion will rise, frequency response will
suffer, and you will get poor results.
The classic rneasurement for output impedance
is to load
a circuit until the voltage drops 6dB
(to half the original power) and note what the
load value is. Intheory, you now have a load im-
pedance that is the same
as the output imped-
ance. If you reduce the load graudally, the dB
reading will return slowly to its original value.
How rnuchdrop
is acceptable?What load will be
left when an acceptable drop
is read on the me-
ter!
When the load value (input Z)
is approxirnately
seven times the output impedance, the needle is
still a little more than 1 dB lower than the origi-
nal reading.
Most technicians say "1dB, not bad, that's ac-
ceptable". We at TEAC must say wedo notagree.
We think that a seven-to-one ratio of input
(7)
to output (1) is not a high enough ratio, and
here's why:

1. The measurement is usually made at a mid-
range frequency and does not show true loss
at the frequency extremes. What about drop
at 20 Hz?
2. All outputs are not measured at the same time.
Most peopledon't have twenty meters, we do.
Remember, everybody plays together when
you record and the circuit demands, in prac-
tice, are simultaneous. All draw power
at the
same time.
Because of the widely misunderstood rule of
thumb
- the seven-to-one ratio - we will give
you the values for outputs in
a complete form.
Even though the true output impedancemay be
low, say 100ohms, for the practical reasons ex-
plained previously, we feel that the 7:1 ratio is
not sufficient. To use this rule of thumb, you
must use
a higher value. We'll call this value the
"output load impedance". For example, in our
model M-35:
ACCESS SEhID 1.4k ohms
X 7 = 10k ohms
LINE OUT 1.4k ohms
X 7 = 10k ohms
This
is a number that will give good results with
the 7:1 method. To go one step further, here are
the actual minimum ohmic values we feel are
wise. Connect to TOTAL INPUT IIVIPEDAIVCE
LOAD higher than:
ACCESS SEhID 10k ohms
LINE OUT 10k ohms etc.
Our specifications usually show 10,000 ohms as
a "Nominal Load Impedance" and you can see
that we arrived at the first column above by
dividing 10,000 by 7. Any number higher than
10,000
is less load.
Input impedance is more straight forward and
requiresonly one number. Load
is load, and here
are'the values for the M-35:
MIC IN 600 ohms
I-INE IN 50k ohms
ACCESS RECEIVE 220k ohms
BUSS IN 12k ohms
If one output is to be "Y" connectedto two in-
puts, the total impedanceof the two inputsmust
not exceed the load impedance, mentioned
above, and if it becomes necessary to increasethe
number of inputs with slight exceeding of the
load specifications, you must check for a drop in
level, a lossof headroom, lowfrequency response,
or
else suffer from a badrecording. Ifone input is
10,000 ohms, another of the same 10,000 ohms
will give you a total input impedance (load) of
5,000 ohms. To avoid calculations you can do
the following when you have two inputs to con-
nect to one output.
Take the lowervalueof the two input impedances
and divide
it in half. If the number you have is
still 7 times the output impedance, you can con-
nect both at the same time. Remember, we are
not using the true output impedance, we are us-
ing the adjusted number ingroup l, output load
impedance.
When you have more than two loads (inputs),
just dividing the lowest impedance by the num-
ber of inputswill not be accurate unless they are
al1 the same size. But if you still get a safe load
(higher than 7 : 1 ratio) by this method, you can
connect without worry.
If you must have exact values, here are the for-
mulae: For more than 2
:
RX=-j 1 1 1
-+-+-+ .... +--
R1 R2 R3 Rn
RX
= Value of Total Load
For 2 loads or inputs
Finding Impedance Values on Other Brands of
Equipment
When you are readingan output impedancespec-
ification, you will occasionally see this kind of
statement:
Minimum load impedance
= X ohms
or
Maximum load impedance
= X ohms
'rhese two statements are trying to say the same
thing, and can be very confusing. The minimum
load impedance says: please don't make the
NUMBER of ohms you connect to this output
any lower than X ohms. That's the lowest IVUM-
BER. The second statement changes the logic,
but says the exact same thing.

Maximum load impedance refers to the idea of
the LOAD instead of the number, and says:
please don't make the LOAD any heavier. How
do you increase the load? Make the number
lower for ohms. Maximum load means minimum
ohms, so read carefully.
When the minimum/maximumstatement is made,
you can safely assume that the manufacturer has
already done the "seven times
is best" ratio cal-
culation. And the number given inohmsdoes not
have to be multiplied. You can MATCHthe value
of your input to this number of ohms successful-
ly; but
as always, higher ohms will be okay (less
load).
Occasionally, a manufacturer will want to show
you that
7 times the output Z is not quite the
right idea and will give the output impedance
and the correct load this way. They will call the
output impedance the true impedance and then
will give the recommended lowest LOAD im-
pedance.
It may be a higher or lower ratio than
7 times and will be whatever the specific circuit
in question requires.
REFERENCE LEVELS
We should talk about one more reference, a prac-
tical one.
Anyone who haseverwatched
a VU meter bounce
around while recording knows that "real sound"
is not a fixed value of energy. Itvaries with time
and can range from "no reading" to "good grief"
in less time than
it takes to blink. Inorder togive
you the numbers for gain, headroom and noise
inthe M-35, we must use a steady signal that will
not jump around. We use
a tone of 1000 cycles
and start itout
at a level of -60dB at the rnic in-
put, our beginning reference level. All levelsafter
the rnic input will be higher than this, showing
that they have been amplified, and eventually we
will come to the
last output of the M-35 - the
line-out and the reference signal there will be
- 10dB, our "line level" reference.
From this you can see that if your sound is
louder than 94dB spl, or your rnic will produce
more electricity from
a sound of 94dB spl than
-60dB, al1 these numbers will be changed. We
have set this referencefor rnic level fairly low. If
you examine the sound power or sound pressure
level (Spl)chart on page 6 you will see that most
musical instruments are louder on the average
than 94dB spl, and most commercial mics will
produce more electricity than the
- 60dB for a
sound pressure of 94dB, so you should have no
problems getting up of "OVU" on your recorder.
If you are going to record very loud sounds you
may produce more electrical power from the rnic
than the M-35can handle
as an input. How can
you estimate this inadvance?Well, the spl chart
and the rnic sensitivity are tied together on
a one-
to-one basis. If 94dB spl gives - 60dB (1mV)out,
104dB spl will give you -50dB out, and so forth.
Use the number, on our chart for sound power
together with your rnic sensitivity ratings to find
out how much level, then check that against the
maximum input levels for the various jacks on
the M-35. Ifyour rnic
is in fact producing - 10dB
or line level, there is nothingwrong with plugging
it into the line level connections on the mixer.
You will need an adaptor, but after that it will
work
!
Most rnic manufacturers give the output of their
mics as
a minus-so-many-dBnumber, but they
don't give the loudness of the test sound in dB,
it's stated as a pressure reference (usually 10 mi-

crobars of pressure). This referencecan be found segments.
'
on our sound chart. It is 94dB spl, 10 microbars,
10 dynes per cm2or 1 Newton per square meter.
For mics, the reference "0"
is 1 volt (dB).So, if
the sound
is 94dB spl, the electrical output of
the rnic
is given as -60dB, meaning so many dB
less than the reference 0=1volt. Inpractice, you
will see levels of -60dB for low level dynamics,
up to about -40dB or slightly higher for the
better grade of condenser mics available today.
TASCAM recorders and mixers work
at a level
of -10dB referenced to 1 volt (0.3 volt) so, for
94dB spl,
a rnicwith a referenceoutputof - 60dB
will need 50dB of amplification from your M-35
or recorder in order to see "OVU"
(- l OdB) on
your meter. Now, if the sound you want to re-
cord
is louder than 94dB spl, the output from
/
the rnic will be more powerfuland you will need
less amplification from your IVI-35to make the
needles on your recorder read "OVU".
THE
BLOCK DIAGRAM AND GAIN BLOCK
DIAGRAM
'
Beforeyou begin reading the next section of this
manual, flip out the extra fold on page 42. On
this page, we have printed the block diagram. It
shows the signal flow through the M-35 and it
represents in simple form, the actual electron
arrangement of
al1 the jacks, controls and gain
stages from mic-into line-out.
The diagram on page 43 indicates the gain of a
"
reference signal, the noise level, and the available
reserve gain or headroom
at any point in the
signal chain. An experienced audio engineer
would be able to operate the M-35 successfully
with just these two diagrams and
a list of input
and output specifications.
Any question about function or gain can be an-
swered by studying the drawings. Will the acces-
sory send signal change in level ifthe input fader
is moved? No, the signal is shown leaving the
main line before the input fader. You read both
diqgrams from left to right, input tooutput.
1. As laid out for convenience.
2. As wired, but knobs and jacks as they appear
on the outside.
3. The block diagram, with the controls num-
bered to correspond to numbers on the first
two drawings.
Even with this "translation system" to help, mul-
tiple sources and outputs can complicate things,
so when necessary; we will also include other
types of drawings to help get the point of
a sub-
system across when we first encounter a source
"point" that will be used in
a specific way. This
may require re-reading if you are not familiar
with subsystems, but we think
it best to advise
you as early as possible.
When printed in
its entirety, a block diagram can
look formidable, and tracing a signal path is not
/
easy, so toaid you in your initial understanding,
we'll continue to use our 3 drawing system first
shown in the introduction, but in slightly smaller
d

INPUT MODULE
6 Input Select
Q Trirn
XLR
@ LINE n
@
frorn Subrnaster Tape in' *
All 8 input modules are identica1 and can be in-
terchanged without modification.
Mic input XLR connector @
Balanced pad circuitcontro1 MIC ATT switch a
3 positions are provided: off, or no effect, set
rightwards one step, a loss insignal of 20dB, set
rightwards two steps,
a loss in signal of 40dB.
Before using the first step, reduce the 8 trim
contro1 tominimumor furthest countercIockwise
rotation. Since the combination of trim and pad
is a maximum of 60dB loss,
it is possible to use
this mic-in jack
as an emergency "line in" if no
tools are availableto convert a3-wirecircuit to
a
2-wire RCA connector - if the line level signal
can be reducedto a max of OdB, (1V).
Inputtransformer
@
Maximum signal to thisinternally mountedtrans-
former is -35dB (17.8mV)without usingthe pad
or MIC ATT switch. At 20dB pad, maximum in-
put is -15dB (17.8mV). At 40dB pad, maxi-
mum i'nput is +5dB (1.8V).
This 3-pinconnector, padcircuit, and transformer
are the only 3-wirecircuits inthe M-35.
We have talked a lotabout the 2-wirecircuit be-
ing a better way to do the audio job, and rnic
lines do not run for "miles and miles" inour sys-
tem. Why do we use this more expensivedesign
to begin with if
it offers no improvementinqual-
ity?The low-power signal that the rnic generates
must be protected and isolated from other low-

power signals in the real world. Radio power line
hum, crackles and switching noise when motors
start up (do you have a refrigerator on your AC
line?)-al1 these unwanted things-must be kept
out of the very high gain amplifiers that are ne-
cessary to raisethe rnic signal to aworking level.
So, the balancedor 3-wire,circuitand input-isola-
tion transformer becomes the only cure way to
dea1 with the problem:
Here's how itworks:
Mic
Any signal will pass to amplifier, no rejection.
Radio Frequency Interference
7~--
Mic
m
Audio signals from rnic have opposite polarity.
Buzz, hum, and
RFI have common polarity.
rv Secondarv
/=%;j'lL-i
RFcancels in transformer
Signals with opposite polarity in the primary coi1
will genera1 current in the secondary coil. Signals
with common polarity will cancel out in the pri-
mary coi1and will not pass to the secondarycoil.
No signal in the secondary coi1 means no signal
in the amplifier.
Input Select Switch
This switch has 3 positions. Left selects the IMIC-
IN XLR. Right selects the LINE IN RCA jack on
the back of the module, and center selects one
of the TAPE IN jacks on the buss master mod-
ules. Since each input module will receive only
one TAPE IN signal, we'll provide
a chart to
show which signal goes towhich module.
Right here we have our first major problem in
comprehension. The connection and its circuit
is drawn plainly on the block diagram, but what
does
it mean in functional terms?Why is the IN-
PUT switch wired to this extra LINE IN when
there
is another LINE-IN on the module?The
answer lies in the requirements of an 8-track sys-
tem in use, and to explain, we'll have to show
the system in
its entirely, even though we have
not reviewed the first path to the recorder at all.
We must assume that a recorder has only one set
of playback outputs. We will have at least three
basic jobs to do that require the playback signal:
1. Simple playback to judge performance, requir-
ing no corrective EQ. In short, what did you
record
?
2. Simple playback into a cueing system so par-
tially completed tapes can be finished. This
function should somehow combine the signals
of simple playback with new rnic signals, so
musicians may hear when overdubbing.
3. Final remix, when the full contro1 capability
of the system (EQeffects mixing, etc.) can be
used to "fine tune" the finished tape.
Three tasks, one output. How do you plug in?
This special input RCA Jack is on the master
module, not the input module, and the
8 sections
tions are laid out on the back like this, part of
our standard "working patch" for 8-track record-
ing.

The numbers on the input module now relateto for "re-mix" and al1 module settings for "mic"
the jacks on the submaster. Track one from tape will have to be disturbed
- every time you play-
playback will now be availableon three separate back. By using the TAPE-IN jack, resetting
is
systems. If only the "line input" on the input avoided. Another drawing may rnake the wiring
module
is used, the signal will only be available more understandable.
for contro1 roorn M-&\
Track 8 playback
To keep the routing clear, we show only the last
submaster modulethat handles track 8 and track
4. The other groups are similar, routing signal to
their respective input module numbers
as shown
in our first drawing (theone with the 8-track).
EFFECTS RCV I
EFFECTS RCV 2
EFFECTS RCV 3
EFFECTS RCV 4
L
AUX IN
R
EXPANDER INPUT
TAPE IN
11-41
CUE OUT
Il Il
EFFECTS SENO
1x11
TAPE IN
15-81
L
CONTROL RDOM
R
STUDIO FEED
L
STUDIO FEEO
IREAR1 l n1
STUDIO FEEO
TALK BACK

Now for thethree requirements oftape playback:
1. Simple record check
To do this, the tape playback signal
is substi-
tuted for the monitor output signal on the
submaster monitor select, and the monitor
mix is now derived from tape playback oneach
section so switched. Any or al1
8 may be
selected individually and the control room
master will then set
a level for the control
room loudspeakers.

-
-Q -0 7. ,llllllll~'
T-., u* z .- I' ;A . :- 7 I, o i' I ,Lo~nan "i, .-v:... Q
l l l.l l l l l , 11 t
i i
*oirin.""-o
2. Cue System
Tape playback plus mic cueing for overdubs.
81155 IN
(.41
LINE
IN @
1.81 LUNE OUT
(x41
nux OUT
1x4)
TAPE IN
1141
CUE OUT
EFFECTS RCY
2 1x1)
EFFECTS SENO
EFFECTS RCV 3 1111
EFFECTS RCV 4 TAPE IN
15-81
t, ' From Tape 4 Tracks
This cue system combines the tapeplayback con- "hear" it, and a meter shared with the effects
trols (x8)on the submaster moduleswith the in- buss allows you to set levels; but caution is ad-
put cue controls (x8)on the input modules to vised. The headphones volume may not relate
form a mono sum of al1 16 possible signals that directly to your contro1 room volume. You have
might be needed for a musicians cue. Since the a master for loudspeakers, but it does notaffect
monitor system for the contro1 room can be set this mix in the headphones. The cue mix has no
to audition this signal "mix" you will be able to master pot.

3. Tape-in for Rough Remix
MIC IN
1x01
EFFECTS RCV
EFFECTS RCV
EFFECTS RCV
EFFECTS RCV
EXPANDER INPUT
I
2
3
4
When the INPUT select switches on the modules
are set to the center position, the submaster
TAPE IN jacks are internally connected to the
input modules. Selecting this remix position on
the input module will not disable the norma1
operation of the monitor. Signal will go to both
circuits
at the same time, allowing separate use
of the monitor outputs as extra mixes. True
stereo echo is an obvious first choice, and the
mixdown machine can be monitored using the
AUX IN position on the contro1 room monitor
module. Since the remix position can be selected
one module
at a time, a single track may be
equalized and monitored without disturbing
AUX OUT
1x41
TAPE l N
11-41
CUE OUT
11 Il
EFFECTS SEND
111 I
TAPE IN
15-81
mix to 2 track
MIC IN settings on the whole mix. A decision
can be made quickly on the artistic success or
failure of an individua1 part without the need to
place the entire console in remix mode just to
see the effect of corrective equalization on
a
single track. Since the TAPE IN jacks on the
buss master moduels actually feeds 3 separate
mixing positions
it will present a more severe
load to anything connected to it than "Line in"
on the module. TAPE IN'S have an input load
impedance of 20k ohms and LINE IN'S(onthe
input module) have an input impedance of 50k
ohms.

Line InJack
An RCA jack on the rear of each input module.
The maximum signal you can apply here is
+15dB. The MIC ATT switch does not affect
this input. The input impedance 50k ohms.
Trim
This control will alter the gain of the first ampli-
fier in the console. It will affect the level of any
signal, MIC, LINE or TAPE. With this potrotated
fully clockwise (rightward), the maximum gain
of the first amplifier is 26dB. In this position,
the maximum input signal before overload is
- 1 l dB (282mV). When the pot is in its mini-
mum setting (fullycounterclockwiseor leftward)
the gain is reduced to 5dB and the maximum
signal that can be handled without overload will
be +10dB (3.2 V). Remember, these overload
figures refer to the input of the amplifier, not
the input plug or connector. Losses occur, and
pads can be inserted before this point. The
maximum signal that can be applied tothe LINE
IN jack is +8dB (2.5 V) with the TRIM rotated
fully leftwards, and +30dB (31.6 V) with the
TRIM rotated fully rightwards. Trim pots con-
trol the gain of an amplifier by adjusting the
amount of output signal returned to asecondary
input control "pot" or input. Because of this
reverse control aspect, we consider it unwise to
adjust the TRIM while signal
is being recorded.
Obviously, you mustadjust when signal is present,
but when serious recording
is in progress the
possible negative side effects on amplifier stabili-
ty and distortion indicate that you should "mix"
with the straight line input, submaster, and
master faders
P only, and adjust "trim" during re-
hearsal.
Cue Pot
At this point in our input module, we derive sig-
nal for a or headphones mix. This function
is best served by
a mix of signals that will remain
constant after being set, so it is drawn off, be-
fore the input fader. To raise the level, rotate
rightwardsor clockwise.
This pre-fader source insures that your mixing
decisions will not interfere with the rehearsal in
the studio. The only thing that
is more annoying
to a player wearing a headphonesthan asudden
change in tone
is losing track of the sound of his
instrument entirely. Remember, al1 music execu-
tion is in large part a response to what is heard,
and if the main source of sound is provided by
this headphones"mix", and you turn itoff inthe
course of some other control room action, you
will deprive the player of the creative guide to
what
is going on. Your session may stop cold
right then. 'rhe cue system routing has been
drawn on page 18 and shows the 16 sources of
signal and the inputs and outputs on the back
panel.
Overload LED
When signals high enough to make the ACCESS
SEND jack output exceed
+l 5dB are applied to
an input module, this LED will light up. The
TRIM, or the MIC ATT pad should be adjusted
until the LED remainsout when signal ispresent.
When recording extremely percussive transient
material, it may require full negative trim and
pad (MICATT) to prevent this LEDfromflicker-
ing on strong peaks. Changing to a less-sensitive
rnic may help.
Access Send - Rcv Jacks
The high gain provided by the rnic preamplifier
allows us to place
a "patch-point" in this more
useful position. The level at the SEND jacks is
-10dB (0.3V) and the output load impedance is
10k ohms. A limiter connected to this point in
the M-35 circuit can now be set to
a range of
compression that will not be altered when either
the equalizer (nextstage) is adjusted or the input
fader
is moved (the stage after the E0 amps).
This pair of jacks is not "normalled" so, when no
device is bridged from SEND to RCV, jumpers
must be in place for signal to flow to the E0
amps and on through theconsole. However, since
al1 the mixing controls lie after the "RCV" jack,
it is possible to use "ACCESS RCV" as an input,
and by-pass the first gain stage. The only func-
tions that will be lost are the trim and overload
indicators. The signal quality will improve slight-
ly but
it will not be possible to switch to "MIC
IIV", "LINE IN" or "TAPE" without repatching.
This unorthodox patch
is suggested for final re-
mix when al1 recording has been completed, and
more time for patching
is available. Maximum
level in will be +15dB. Input impedance is
220k ohms.

PARAMETRIC EOUALIZER SECTION
Before we begin, the label itself will requiresome
explanation. What
is parametric and, equal to
what
? A logical question, because the term does
not describe what you do with the controls. In
multitrackaudio,tonecontrols are almost always
used to "make different" and the concept of
"make the same" doesn't quite fit. How did we
get this label on our tone control?
The telephone companyuses
it. Inthe earlydays,
the system worked well in the lab, and in short
runs of 100yards or so, but
... ..
When two "phones" were 10miles apart the line
between them did not transmit al1 of the sound
representing signal in the same way. Some parts
of the frequency spectrum did not pass down
the line at all, some parts were different in level
or displaced in"time". What came out of the ear-
piece was definitely not what had "gone in" 10
miles away and understanding
a conversation
proved to be difficult. What now? 'rhe Phone
company had to learn how to make the output
"sound like" the input.
If "output equals input" is the concept, an "equal-
izer" is a logical name for the device used to fix
your problem. Just as in many other concepts in
audio, the telephone company language has set
the terms we use today.
The term "parametric" refers to the adjustable
frequency point. The "parameters" or "rules" are
not fixed at any given number, butare continu-
ously variable. Both aspects of the circuit, fre-
quency center point as well as gain or loss are
continuously adjustable without"steps" but there
are limits.
The Model 35 offers a two control, four range
parametric equalizer.The lowergroup of controls
offer +10dB of boostlcut controlatany frequen-
cy between 60Hz and 1.6kHz in two ranges,
selected by the switch below the concentric con-
trols. Set leftwards, the switch selects the range
from 60 to 300Hz. The outer knob when see
fully leftwards or counterclockwise then sets
3
center frequency of 60Hz. As the outer knob is
rotated rightwards (orclockwise), the center fre-
quency is raised in stepless fashion. When fully
right (or clockwise), the center frequency
is
300Hz. The inner knobT the boostlcut control.
Set fully leftwards, the cut is -10dB. Set fully
rightwards, the boost
is + 10dB.
Outside Control,
rotate right to
raise center frequency.
Inner Control
rotate leftwards to cut
rotate rightwards
toWboost"
When the lower switch is set fully rightwards,
the action
is the same, butthe range changes.