This is only a preview of the April 1988 issue of Silicon Chip. You can view 42 of the 96 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Articles in this series:
Articles in this series:
Articles in this series:
Items relevant to "Walkaround Throttle for Model Railroads":
Articles in this series:
Articles in this series:
Articles in this series:
|
THE WAY I SEE IT
By NEVILLE WILLIAMS
Audio power ratings - right
back to where we started from
Having agreed some years ago that music power
was a highly suspect rating, the world hifi industry
settled on RMS watts as the true measure of
amplifier power output. But since then, music
power has gained totally new respectability and
even been legitimised by the term "headroom"!
I don't deny that there are supportable reasons for this apparent
about-face but, in thinking them
through, I was reminded of many
other technical twists and turns
that have confused the subject of
audio power output ever since
about 1930, when enthusiasts first
began to take note of it as a
parameter in its own right.
Before then, audio output watts
or, more fittingly, audio output
milliwatts hardly rated a mention.
Enthusiasts simply got to know, by
experience or repute, that some
valves were "louder" than others.
The legendary 201-A detector/amplifier triode, for example,
frequently used in the final stage of
old-style battery sets, didn't even
have an official power output
rating.
Figures were published for its
derivatives, the 112-A (285mW
max.) and the 71-A (790mW max.)
but at a time when most receivers
were powered by batteries or
eliminators, the figures that really
mattered were plate voltage and
plate current. If either was unattainable or unaffordable, tough!
For most enthusiasts, awareness
of rated power output followed the
release, in 1929, of the mains type
45 triode, the first substantial
16
SILICO N CHIP
"receiver" type power valve.
Under maximum supply conditions
(330V) it offered a healthy 2 watts
of output power in single-ended
class-A, and up to 18 watts in pushpull class AB2 .
Its dominance was shortlived
however, because the 47 power
pentode, released about eighteen
months later, offered higher gain
and 2.7W in single-ended class-A at
the more manageable supply
voltage of 270 - about twice the
power available from the 45 at the
same figure.
Watts missing?
But the 47 was not without its
critics. Maybe it did have more gain
but they didn't like its "tone" and it
certainly didn't sound anything like
twice as loud as the 45. The
manufacturers were obviously having them on!
But they weren't. Audio enthusiasts had yet to accept the
sobering fact that a 2:1 (or any
other) increment in audio power, or
sound pressure level, is not subjectively apparent as such. Our ears
spontaneously re-scale all such
changes to an approximate
logarithmic base - nature's way of
enabling them to cope with a huge
real-life range of sound pressure
levels, from the merest whisper to
the shattering roar of a not-somodern jet plane (modern ones are
quieter).
It was to quantify this natural
logarithmic relationship that scientists devised the "bel" and its more
practical derivative, the "decibel",
abbreviated to dB.
Taking the above 2:1 power ratio
as an example, conversion to a
decibel relationship involves taking
the log (to the base 10) of 2, which
turns out to be 0.3010; multiply this
by 10 ·and call the result decibels say 3dB.
And therein lies the answer to
the fate of that supposedly missing
watt: to the ear, with its logarithmic
response, a 3dB change in level is
perceptible but certainly not twice
as loud.
Watts twice as loud?
If you're expecting a change in
SPL (sound pressure level) sufficient to be described subjectively
as " twice as loud", you need a
decibel ratio of 10, or a 10-times
boost in output power. To double
the apparent loudness of that ancient 45 triode would call, not just
for a 47 pentode but for a valve
able to deliver 20W from a singleended stage or 180W from a pushpull pair - a rather tall order.
If memory serves me correctly,
no "receiver" type output valve has
ever been marketed which could
meet that requirement. Even today,
to conservatively design a solidstate amplifier with an output of
20W single-ended or 180W in pushpull, you're still likely to end up
with a parallel configuration!
It's rather odd, when you think
about it, that in terms of power output rating, most of the receivers
and amplifiers that hifi enthusiasts
have used during the past 50 years
fall into a l0dB slot; from say, 10W
for time-honoured push-pull 2A3s to
100W, covering virtually all other
valve amplifiers and the vast majority of their solid-state counterparts.
So be careful before you start
boasting that your pet system can
be twice as loud as the one next
door. Unless you have 10 times the
audio power and/or more sensitive
loudspeakers, it can't!
Watts and efficiency
On the strength of that last
reference, I should perhaps interpose the observation that in the
context of loudness, many hifi enthusiasts are still unclear about the
importance of loudspeaker sensitivity - the efficiency with which
loudspeakers convert electrical
drive power to sound.
Loudspeaker sensitivity depends
on design factors too numerous to
catalog here but the intensity of the
magnetic field surrounding the
voice coil looms large as one of
them.
How sensitivity is measured is
also beyond the scope of this article
but it is expressed in decibels
(referred to a sound pressure level
of 12µbar at a distance of one
metre) and curiously, the sensitivity
range of domestic hifi loudspeakers, used over the past 50-odd
years, also happens to be about
10dB wide - from around 85dB for
comparatively insensitive systems
to 95dB for models of reasonable
efficiency.
For a given level of audio drive
(within overload limits), a 95dB
loudspeaker system would be expected to sound twice as loud as
one with a sensitivity of 85dB.
On that basis, a venerable pushpull valve stereo amplifier, delivering 10 + 1OW RMS to a pair of those
large and notably sensitive circa
95dB Wharfedale or Goodmans
drivers of 25 years ago, would
develop about the same SPL as a
100 + 100W solid-state amplifier
feeding a pair of low efficiency
(85dB) compact enclosures.
These days, the "average" hifi
FOUR POWER PENTODE valves from the late 20's and early 30's (from left to
right]: type 33, 42, 45 and 2A3. Power outputs were in the region from 2-3
watts.
enthusiast probably ends up with
an in-between combination: around
40 to 60 watts of amplifier drive, a
loudspeaker sensitivity of around
90dB and a marginal loudness advantage of about 2dB over the
above extremes.
In short, whether visualised in
watts or decibels, audio drive
power, on which we place so much
emphasis, is only half the story.
How efficiently we convert it is just
as decisive!
Watts available!
But back to the 1930s: despite the
levelling effect of ears and decibels,
valve and receiver manufacturers
of the period kept right on doggedly
chasing small power increments,
which would hopefully catch the
buyer's eye, even if their ears were
unmoved.
For engineers at least, it was
visually stimulating to see a larger
pattern on the CRO screen and a
higher reading on the output meter,
especially in the knowledge that, to
calculate the power resultant in
watts, the formula required the
voltage reading to be squared:
P = E 2 /R
That was fine but at an aural
level, instead of squaring something, listeners were lumbered with
logarithms, often ending up with a
thoroughly unexciting 2 or 3dB an exercise that, if nothing else,
served to separate technically in-
dined pessimists from their opposite number.
"It's not worth doing this or
that", they would say, "because the
difference won't be noticeable". To
which the optimists would reply:
"Rubbish! Put enough 2dBs end to
end and, somewhere along the line,
you'll end up with a difference
worth having!"
In a roundabout way, I guess,
they were both right!
Watts impedance?
But valve or solid-state, with or
without an output transformer,
audio amplifiers normally end up
feeding into a loudspeaker load one that is generally more reactive
than resistive, having an impedance that can vary widely with
frequency and which may approximate its rated value across only a
limited portion of the spectrum.
That's common knowledge, but I
wonder how many realise the extent to which it compromises our
oh-so-tidy thinking, based on those
convenient but fictional resistive
loads used for most power
measurements .
Watts or volts?
By definition, any amplifier these
days using adequate negative feedback has a substantially "constant
voltage" output characteristic with
loads of not less than a specified
value. As a corollary, such an
APRIL 1988
17
amplifier will also have a low output impedance and a high damping
factor. .
As such, it should have the potential to combine well with any ordinary hifi loudspeaker system, for
which the normal design objective
has been to achieve an adequate
SPL in the listening room, when fed
from a compatible "constant
voltage" amplifier.
Audio power, as such, does not
appear in the above statement; nor
can it in any precise way because,
while the audio voltage at the
amplifier/loudspeaker interface
can be monitored and measured,
the impedance and power factor of
the load itself is subject to wide instantaneous variations with frequency.
But, with rare exceptions (eg,
SILICON CHIP, Feb. 1988, p.40) we
simply pretend that loudspeakers
are high-wattage resistors of,
typically, 4 or 8 ohms. On that
basis:
(a) We quote their normal and maximum drive levels in terms of fictitious watts rather than the actual
volts with which they are normally
Did you
fed and on which their frequency
rating is based.
(b) We likewise measure and rate
amplifiers on their ability to deliver
power into resistive loads which, in
fact, they rarely encounter.
Maybe I'm being over-pedantic.
Maybe it's a bit like Santa Claus:
easier to go along with the fiction
than to complicate kids' lives by denying it!
Admittedly, there are a few
loudspeaker systems around in
which the reactive components
have been substantially balanced
out but I wouldn't advise holding
your breath until the technology
becomes the rule rather than the
exception.
Watts - the genuine variety
This brings us back to where this
article began - the about-face in
relation to amplifier power ratings.
In its post-war resurgence, the
hifi industry adopted the practice
of quoting the audio power
delivered to a load resistor across
the output-to-voice-coil terminals at
the onset of overload.
The test was done under con-
•
llllSS
tinuous tone conditions, using a
lkHz sine wave and such other frequencies that the designer might
choose. The figure obtained was
commonly ref erred to as the RMS
power output - a term that was
challenged on principle by some
who maintained that there was no
such quantity as RMS power. It has
persisted, however, on the grounds
that it refers to power derived from
a measurement of RMS voltage.
Continuous tone testing was
recognised as a conserve tive
method but justified because it indicated the ability of an amplifier to
handle sustained passages of loud
music as, for example, the heavy
bass pedal notes of a pipe organ.
It was a comfortable enough
rating for generously designed
amplifiers but an embarrassing one
for " budget" models in which sustained loud signals of any kind
caused a reduction in the internal
supply voltage and, with it, a reduction in the measured power output.
Watts -
music and peak
In consequence, some manufacturers ignored the RMS convention
these issues?
Issue Highlights
January 1 988: 4-Bay Bowtie
UHF Antenna; Dual Tracking
Power Supply; Custom Phone
Ringer; Subcarrier Adaptor for
FM Tuners .
Please send me a back issue for
□ November 1987
□ December 1987
□ February 1 988
□ March
□ January 1988
t 988
Enclosed is my cheque or money order for $ ... ..... or please debit my
□ Bankcard
□ Visa
Name .. .. ..... ... ... ... .. ....... .... .... ..... .... .......... .... .. ... ... ...... .... ... ...... ..... ..
Address .. ... ... .. .. ..... ...... ........ ... ........ ... ..... ... .. .. ............... .. ...... ....... .
Suburb/town ...... .... ... ... .. .. ..... ..... .. ... ... ... ....... .... Postcode .. .... ... .... .. .
Card No .. ..... .. ... .. .. ... .. ..... .. .. .. .. ... ........................... .. ..... .. .. .... ... ... .. .
Signature .... ... .... ........ ....... ... ....... ..... Card expiry date .. ... .. / .... ... / ... ... .
February 1 988: 200 Watt
Stereo Power Amplifier ; Deluxe
Car Burglar Alarm; End of File
Indicator for Modems; Simple
Door Minder; Low Ohms
Adapter for Multimeters.
March 1 988: Remote Switch
for Car Alarms ; Telephone Line
Grabber; Low Cost Function
Generator; Endless-Loop Tape
Player.
Price: $5.00 each (incl. p&p). F ill
out the coupon at left (or a
photostat copy) and send it to:
SILICON CHIP , PO Bo x 139,
Collaroy Beach 2097.
~---------------------------------------18
SIUC() N CIIII'
and began rating their amplifiers in
"music power", signifying the output available at the onset of distortion during short musical phrases.
It could typically be from 10-30%
above the RMS rating, so that a
humble 10 + 10W amplifier might
well be rated at 13 + 13W music
power.
To create a still better impression, the two figures could be added
together to yield "26W total music
power" .
But then someone else realised
that the instantaneous power at the
peak of a sine wave was twice that .
of the sine-wave itself, permitting
the aforesaid humble 10 + 10W
amplifier to be re-rated yet again to
26 + 26W peak music power or "a
massive 52 watts of total peak
music power".
Such tactics gave the hifi industry a thoroughly bad name and
consumer pressure in the US eventually led the Federal Trade Commission to rule in 1974 that
amplifiers must be rated in terms of
continuous power outout. Further,
in testing the power output of an
amplifier. it would be subject to an
hour's preconditioning at 33% of
its rated power. This caused great
consternation among amplifier
manufacturers because it meant
that amplifiers had to be designed
more conservatively, particularly
as far as their power supplies and
heatsinks were concerned.
Subsequently, music power was
re-defined and recognised in the
1978 IHF (Institute of High Fidelity)
standards as a supplementary
rating, together with a new term,
headroom, which referred to the
decibel ratio of music power to continuous power. But music power
was still regarded by many as a
Clayton's output: watts you had
when you really didn't!
Watts -
transient peaks?
But, as I indicated right at the
outset, the recording scene has
changed drastically during the past
decade, as also have our expectations of amplifier performance.
Around the mid-70s, audio
engineers became increasingly
aware that, while their analog tape
decks were producing ostensibly
clean master recordings, they were
subtly crushing the high amplitude
transients - a problem of dynamic
range for which there appeared to
be no ready answer.
Then quite suddenly, through the
efforts of resourceful professional
recordists, hifi enthusiasts were
confronted with a sequence of new
"ear popping" audiophile discs,
some direct cut, others dubbed
from digital master tapes.
It so happened, at that opportune
time, that I acquired a then-new
Technics SU-V 4 70 + 70W integrated amplifier, fitted with
fluorescent peak-hold level indicators calibrated to 100 + 100W.
For the first time, in a home
listening situation, I was exposed to
recorded transients that not only
gave pianos, acoustic guitars and
other percussion instruments a
startling sonic presence but which,
from no more than ample room
volume, flicked the level meters to
50 watts or more - well above the
reading for a sustained, subjectively loud, organ or orchestral
fortissimo.
Clearly, a new era had dawned,
which was soon to be perpetuated
by compact discs.
Taken at face value, the above
observation suggests that a hifi enthusiast who likes to listen at a
generous domestic level, using
typical loudspeakers (1 W, lm,
90dB), now needs several hundred
watts per channel to be reasonably
sure of coping, not so much
nowadays with fortissimo passages, but with the high amplitude
transients that characterise some
modern recordings.
Such amplifiers are very expensive and can easily run into many
thousands of dollars, particularly if
they fully conform to the US FTC
standards. There has to be another
way, and there is. In the last few
years a number of manufacturers
have looked at the concept of
headroom. Why not design an
amplifier with far more headroom
than previous designs offered?
Instead of having a headroom
figure of say, ZdB, which is fairly
typical of current designs (such as
the Studio 200 described in the
February 1988 issue of SILCON
CHIP), why not go for a headroom
figure of 6dB? That means a music
conlimwd on pogo
oa
Problems?
... and you
don't have our
112 page
catalogue ...
you've got
real
problems!
ARISTA ... your one-stop
problem solver. Audio leads
... Batteries ... Chargers ...
Battery holders ... Cables
... Car accessories ...
CD accessortes ... Converters
... "Cutec" .. . Earphones .. .
Fuses ... Headphones .. .
Intercoms ... Knobs .. .
Microphones and accessories
... Mixers ... Multimeters ...
Plugs/Sockets, etc ... Plug
adaptors ... Power packs and
leads ... PA ... Disc and Tape
care ... Security equipment ...
Signal modifiers ... Solderless
terminals ... Storage boxes ...
Switches ... Telephone and
TV accessories ... Tools and
Technical aids .. .
Video accessories ... Wiring
accessories ... You name it
and we're bound to have it
... Try us ... NOW!
Get your catalogue ... it'II solve a
whole lot of your problems!
Just send $2 + 50c p&h and
your return address to:
ARIST~
ELECTRONICS PTY LTD
PO BO_X 191, LIDCOMBE, NSW 2141
A l'lllL HJBtl
l9
There is also a large illuminated
rocker switch on the variac which
is a convenient feature.
The price is just $175 which is
something of a bargain in today's
market. For further information
contact Geoff Wood Electronics,
229 Burns Bay Road, Lane Cove
NSW 2066. Phone (02) 427 1676.
Sony's limited edition
Video 8 Camcorder
To mark the Australian
Bicentenary, Sony has released a
limited edition of its Video 8 Camcorder, the model CCD-AU200. This
is dressed up in the green and gold
Bicentenary colours.
The CCD-AU200 uses the standard 8mm Video 8 cassette giving
up to three hours of recording time
(depending on the battery pack) and
has a charge coupled picture sen-
The Way I See It -
Light sensitive switch for outdoor lighting
If you wish to turn a light on at
dusk and off at dawn this unit
from Arista is the solution. It
combines a bimetal switch and
large photocell to provide
automatic switching. It is
energised from 240V AC by connecting it in series with the lamp
to be switched. It is rated for incandescent lamp loads up to
2kW; up to lkW for mercury
discharge lamps with high power
factor; and up to 1.8kW on mercury discharge lamps with low
power factor.
Power dissipation in the
switch itself is two watts. It is
claimed to switch on for light
levels below 20 lux and off for
light levels above 100 lux.
The NLS2 light sensitive
sor which will operate down to a
minimum light level of 19 lux. The
camera has automtic white
balance, automatic iris and
automatic focus so that picture
recording is just a matter of aiming
and squeezing the trigger.
In addition to these features, the
camera has a manual zoom lens
(2.5x) with macro capability for
close-ups. It also has compensation
for backlighting which normally
ctd from page 19
power output four times as high as
the continuous power rating.
Now we are starting to talk about
a really worthwhile increase in
power, or are we? Designing an
amplifier with such a headroom
figure means that the power supply
and heatsink requirements are
quite different from those for an
amplifier expressly designed to
meet FTC standards.
Instead of having a power supply
which is comparatively well
regulated (ie, varying in voltage between no signal and full power by
only 10% or less), we have a power
supply which is designed to deliver
maximum output for a very short
period of time which under the IHF
standards is only 20 milliseconds.
That means you can use a small
power transformer and still come
up with spectacular power outputs.
For example, with a power
transformer, output transistor and
heatsink configuration which might
normally be expected to give 50
watts per channel, you could now
get 200 watts per channel but only
on very short term basis. This is just
what you want for handling those
short term transients which
characterise pianos, guitars and
other percussive instruments.
That is just what a number of
manufacturers have done. By
various means, companies like
Carver, NAD and Proton have come
up with very large figures for music
power and headroom.
And therein lies the basis for
what I described at the outset as a
supportable reason for an apparent
about-face, and a totally new
switch is available at the recommended retail price of $17.95
from Arista outlets.
tends to darken the subject; just
push the back-light control to correct the picture.
The versatile black and white
electronic viewfinder has indicators for function and lighting
conditions and is adjustable to suit
left-handed or right-handed
operation.
Again, to mark the Bicentenary,
the CCD-AU200 is priced at $1988.
(Get it?)
respectability for once-maligned
music power. But, having thus completed one ride on the technical
merry-go-round, we find ourselves
heading for another, this time to do
with the watts/decibels dilemma of
the 1930s. What's the best way to
describe and rate this extra
output?
The options are "music power"
or "dynamic power" in watts,
hopefully supported by the endorsement [IHF-A-202). Or, again,
"headroom" in decibels, arrived at
by similar methodology.
As it happens, the manufacturers
in question are specifying both
music power and headroom but the
tendency is to emphasise the watts
and forget the headroom.
The way I see it, we still prefer
the watts we can boast about to the
decibels we can actually hear! lb
Al'lllL 1988
93
|