Modifying the Studio 200
power amplifier
Back in February 1988 we published the Studio 200
power amplifier which has proved to he a very
reliable and consistent per£ormer. In the
intervening period a number of readers have
written to us describing modifications they have
made to "audibly improve" the performance. Here
are the modifications made by one reader.
By LEO SIMPSON
Since it was published in
February 1988, our Studio 200
power amplifier has proved itself to
be very trouble free and many hundreds have been put together
without problems. Constructors
have been particularly delighted
with its very low noise. In that
respect it is a lot quieter than many
The Studio 200 power
amplifier was actually
developed from this
module which was
published in the December
1987 issue. This circuit
pioneered the use of PTC
thermistors for amplifier
and loudspeaker
protection.
100
SILICON CHIP
commercial power amplifiers,
regardless of their cost.
But just as there exists a small
specialist market for those well
heeled people who want commercial amplifiers to sound better, so
there is also a temptation among
keen audio enthusiasts to modify
the Studio 200. Up till now, we have
been reluctant to encourage the
idea of modifications because of the
considerable risk involved.
First of all, there is the risk that
uninformed modifications may lead
to a costly failure in the amplifier.
This would be particularly frustrating if the amplifier had performed perfectly up until it was
modified. And then there is the risk
that the amplifier may sound no better or may sound worse, after the
mods have been made.
We have also been reluctant to
publish readers' modifications to
our amplifier designs because,
without exception, such modifications have not been accompanied
with any analysis or objective
measurements to justify the work.
This is not to say that the design
of Studio 200 is absolutely right in
every way. As with any commerical
design, the Studio 200 amplifier is a
tradeoff between cost and performance. At the time we produced it,
we believed the Studio 200 to be a
very good compromise. We still do.
But ...
In publishing the accompanying
modifications we are in no way endorsing them. We do not even agree
with them. Why then, are we
publishing them? To stimulate
debate. Here then, is the reader's
letter:
In relatively recent times I have
replaced op amps to gain a small
audible improvement in various
consumer audio products. I was not
happy with the apparent loss of
depth information in NE5534
preamps compared to an earlier
discrete design I had used, so I built
an experimental preamp using an
LM394, BC559 and LH0002 buffer
in each channel (I believe in low
output impedance).
I repeatedly became aware of a
phenomenon whereby after a certain amount of time I realised that
the soundstage had considerably
improved and the sound was
,--_ _ _ _ _ _____ _ _ _ _ __ _ _ __ _ _ _ _ ____,.,__ _ _ _ __ _ _ _ _ _ _ _ v+
5A
r
lN;:_..s
22k
1'/,
22k
1%
OUTPUT
.,.
-;-
*
04
47k
1%
2xBF469
2.2k
1%
2.2k
1%
5A
100()
1°/,
,..___ _ _ _ _ _ _ _ ____,.,__ _ _ _ _ _....,_ _ _ _ _+ - - ---<11-- - - --
..--------------v+
......----v-
0.1+
RIGHT
0.22
* HFE MATCHED TO 10%
240VAC
0.22
10k
1W
V- RIGHT
N
V+ LEFT
0.22
10k
1W
0.22
10k
1W
.
V- LEFT
Fig.1: the modified circuit of the Studio 200 amplifier uses current sources for the second differential pair instead of
the now tried and proven current mirror arrangement which is very common in op amp ICs. Note that quite a few more
components are employed and that the amplifier has a gain of 23 down to DC.
"sweeter". Each time I noticed the
change, I read all DC voltages and
on every occasion, the potential difference between the collectors of
the input differential pair was less
than 5mV!
Unfortunately the improvement
slowly disappeared (dependent on
the surrounding temperature or
ventilation of the preamplifier).
Several versions of the preamp
later on, the BC559 was replaced
with an imported 2N3811A dual
transistor (at greater than $20
each) and the LH0002 with a
discrete version using selected
BC559 and BC549 transistors. The
unused half of the 2N3811A was
given an identical collector load to
the other half and its collector
voltage used to control a shunt tr ansistor across the reference LED of
the adjustable input pair current
source to maintain optimum conditions.
This arrangement also resulted
in ± 5mV offset from the DC coupled preamplifier. I fluked obtaining
a beautiful dual 18-position rotary
switch in a Jaycar switch pack and
fitted 1 % metal film resistors to it,
to make a precision matched dual
volume control. Surprisingly, the
sound was less harsh also.
More recently, I decided to try
similar circuitry in a 100 watt/
channel amplifier based on the
Studio 200. The amplifier (as it now
stands) is not suitable for the
average home constructor as it requires some parts selection and the
dual PNP input transistor. A competent designer could overcome most
of the drawbacks however.
The audible effect of the input
pair collectors being adjusted to
less than 5mV difference is at least
as audible as in the preamplifier
and the amplifier audibly outperformed my reference " Tillbrook
6000" which uses matched pairs
throughout, 1 % resistors and
polystyrene capacitors in lieu of
ceramics wherever possible. The
JANUARY19 90
101
+4!1.5V
t
01
1N4002
2700
02
1N4002
f
48.1V
0.5V
:_J
B
01
BC556
C
~
.,.
18k
33.3V
et,.
07
B0139
.012
--1
B
l
06
BF470
C
5A
0.1i
B
E 0.1
INPUT
22k
B
1k
...
VR1
5000
47
6VW:r
.,.
04
B
l
!
2xBF469
C
B
48.3V
2.2k
~
:-i
--1
:__J
1.9V 2.2k
1.9V
-49 .5V
O.llSTUDIO 200 STEREO POWER AMPLIFIER
SC1-1-288
Fig.2: this is the circuit of the Studio 200 as originally published in the February 1988 issue of SILICON CHIP, although
minus the power supply. Its signal to noise ratio was close to - 120dB with respect to full power. That's very quiet.
equipment used for auditioning was
a much modified Marantz CD65
player, the previously mentioned
preamplifier and B&W 801s in an
almost ideal listening room.
I am not sure what the improvement is caused by, as I expect the
distortion would already be vanishingly low. Perhaps supply rejection is the answer?
I feel that this phenomenon is
worthy of further investigation,
particularly as most designers do
not fully take into account the
loading effect of the succeeding
base emitter junctions across the
collector resistors of the input pair.
The loading transistors must track
each other thermally if the input
pair is not to be unbalanced. Alternatively, if only half the input pair
is loaded, an appropriate value
resistor should be fitted to the collector of the other input transistor.
Food for thought? (A.K., Chatswood, NSW).
Potential problems
Readers should compare the accompanying circuit diagram (Fig.1)
with that for the Studio 200 (Fig.2)
- see also SILICON CHIP, February
1988 - to see how extensive the
modifications are. We do not propose to analyse the effect of these
modifications as we would prefer
Notice to Subscribers
Are you about to change your address? Please advise us by the middle
of the month by filling in the form on your address label and posting (or
faxing) it to us.
If you have any queries about your subscription, call us or write to us
at: SILICON CHIP, PO Box 139, Collaroy Beach, NSW 2097. Telephone
(02) 982 3935. Fax (02) 982 9553.
102
SILICON CHIP
readers to make comments.
Briefly though, we see a number
of potential problems in the
modifications. First, the use of
separate bridge rectifiers and filter
capacitors for each channel, while
superficially an advantage from the
point of view of better channel
separation, may cause circulating
currents in the earth returns. This
could well lead to audible distortion
or low frequency instability.
Second, we can't see how the use
of a LED (light emitting diode) will
lead to improved power supply rejection. At a current of 2mA, a LED
will typically have around the same
dynamic resistance as a 1N4002
power diode which was originally
specified for the Studio 200. In fact,
at 2 milliamps, the dynamic
resistance of 1N914 or 1N4148
small signal diodes is a good deal
better.
There are many other ramifications of the proposed modifications.
For the time being we will not comment on these. We invite readers to
have their say.
~
