This is only a preview of the June 1996 issue of Silicon Chip. You can view 23 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. Items relevant to "A High-Performance Stereo Simulator":
Items relevant to "A Low Ohms Tester For Your DMM":
Articles in this series:
Items relevant to "Automatic 10-Amp Battery Charger":
|
• Realistic stereo performance
• Low noise and distortion
• Adjustable stereo effects
• Runs from a 12V plugpack
A high-performance
stereo simulator
This high performance stereo simulator uses
a digital delay chip to convert any mono
signal source into stereo. You can use it to
enhance the sound from mono VCRs, AM
tuners or electronic musical instruments.
By JOHN CLARKE
If you compare the sound from a
mono source to that of stereo the difference is easily perceived. Instead of
appearing to come from a single point,
the sound is dispersed over a wide
field between the stereo speakers. Very
few recordings these days are made
14 Silicon Chip
with ping pong ball effects whereby
the sounds bounce from one channel
to another and back again. Because
of this, attempting to simulate stereo
sound is a reasonably straightforward
design exercise.
In the past, the usual approach to
producing a simulated stereo effect
was to divide the mono signal into
separate fre
quency bands and distribute these into the left and right
channels.
This frequency division was done
by an array of filters which rejected
certain bands in the audio spectrum
for one chan
nel but allowed them
through for the other channel. The real
drawback to this approach is that you
need a fair few filters for a good result.
Another method was to use bucket brigade delay chips but these are
quite expensive, have high noise and
distortion and the overall result is
mediocre.
So how have we gone about it? Our
task has been made easier by Dolby
surround decoders which require
high-performance digital delay chips.
We have used one of these devices and
the results are very good.
How it works
In essence, we use the delay chip
to provide a “comb filter” effect. This
chops the incoming audio signal into
lots of very narrow frequency bands.
The narrow frequency bands are
subtracted from the mono signal and
the result becomes the left simulated
channel. The difference between the
simulated left channel and the incoming mono signal then becomes the right
simulated channel.
Fig.1 shows the general arrangement
of our stereo simulator. It has an input
buffer IC1a and this feeds the delay
chip IC2. It also drives one input of
mixer IC1d while the delay chip drives
the other input. The output of mixer
IC1d becomes the right channel.
For the left channel, the delay chip
drives inverter IC1c and its output
is mixed with the input mono signal
before mixing with the buffered output
mono signal in IC1b.
This process of mixing a signal
with an identical delayed version
results in some frequencies being “in
phase” and these pass through without attenuation. Other frequencies are
cancelled out because they are “out
of phase”.
If the delay chip is set at 1.5 milliseconds, for example, the input signal
will be in phase with the delayed
output at 666Hz (1/1.5ms), 1.333kHz,
1.999kHz and so on. Thus, these fre
quencies will pass through to the
right channel. For the left channel,
the inverted signals are out of phase
at 666Hz, 1.333kHz and so on and
these frequencies will coincide with
a dip in the response.
Conversely, signals at 333Hz,
999Hz, 1.666kHz, etc will pass through
to the left channel but will have dips
in the right channel. Fig.2 shows the
frequency response for the left and
right channels. The solid curve is the
right channel while the dotted curve
is the left channel.
Note that the notches at the lower
frequencies (333Hz, 666Hz, 999Hz,
etc) are very deep while at higher
frequencies the notch depth becomes
progressively less.
Looking at the responses of Fig.2, it
is easy to see where the term “comb
Fig.1: the stereo simulator has an input buffer (IC1a) and this feeds the
delay chip (IC2). The delayed signal is then mixed with the buffered
input signal to produce the right channel. The left channel is produced
by mixing an inverted delay signal with the buffered input signal.
filter” came from – all those notches
look like the teeth of a comb.
Mind you, Fig.2 shows just one
possible set of frequency responses.
It corresponds to a delay setting of
1.5ms. You can also select delays
anywhere between 0.5ms and 4ms, in
steps of 0.5ms, and each of these settings will have its own characteristic
“comb filter” effect.
Circuit description
We have used the M65830P digital
delay IC from Mitsubishi as the heart
of the circuit. This is the same delay
chip as used in the Dolby Pro Logic
Surround Sound Decoder, as published in the November & December
1995 issue of SILICON CHIP.
The delay chip works by first converting the incoming analog signal to
a digital format which is then clocked
into memory. This digital signal is
then clocked out at the end of the
delay period and converted back to
an analog form. The chip is timed by
a 2MHz crystal oscillator to provide
a 500kHz sampling rate. In the Dolby
Surround Sound Decoder, we used a
microprocessor to control the delay
AUDIO PRECISION STEREO AMPL(dBr) & AMPL(dBr) vs FREQ(Hz)
0.0
21 MAR 96 12:18:53
0.0
-5.000
-5.00
-10.00
-10.0
-15.00
-15.0
-20.00
-20.0
-25.00
-25.0
-30.00
-30.0
-35.00
-35.0
-40.00
-40.0
20
100
1k
10k
20k
Fig.2: these “comb filter” effects are the frequency response curves for the left
and right channels. The solid curve is the right channel while the dotted curve
is the left channel.
June 1996 15
16 Silicon Chip
chip but in this circuit we use three
low cost CMOS ICs. These are required
because the M65830P gets its delay
setting instructions each time it is
powered up.
The full circuit of the Stereo Simulator is shown in Fig.3. The mono
input signal is AC-coupled into unity
gain buffer IC1a via a 2.2µF capacitor.
IC1a then drives mixers IC1b & IC1d
and the delay chip, IC2. The signal to
IC2 is AC-coupled to its pin 23 via a
low-pass filter comprising the 39kΩ
and 18kΩ resistors and the 560pF and
150pF capacitors. This filter rolls off
signals above about 15kHz to prevent
higher frequencies affecting the digital conversion and causing spurious
effects in the output.
The capacitors at pins 17, 18 and 20
control the rate of delta modulation
which is the type of analog to digital
conversion used in IC2. Similarly, the
.068µF capacitor at pin 16 controls the
digital to analog conversion output
signal appearing at pin 15. This output is applied to another 15kHz filter
comprising two 39kΩ resistors, an
18kΩ resistor and 560pF and 150pF
capacitors.
The output of IC2 is then AC-coupled to inverter IC1c and mixer IC1d
via a 4.7µF capacitor.
IC1b & IC1d mix the signals applied
to their inverting inputs via 10kΩ
resistors. Their outputs at pins 1 & 7
become the left and right simulated
stereo channels.
All four op amps in IC1 are biased
to +6V by a voltage divider consisting
of two 10kΩ resistors across the 12V
supply rail.
Delay selection
IC2’s delay is controlled by CMOS
chips IC3-IC6. Each time IC2 is powered up it automatically resets itself to
provide a 20ms delay. This is much too
long for this application so we need to
set it by feeding a serial data stream
to the Data input at pin 6. This data is
Fig.4: taken from a Tektronix TDS360 200MHz digital scope, this printout
shows the timing of the SCK (top), Data, (serial clock) and REQ (request) lines
to IC2. This data is sent once to the delay chip each time it is powered up.
clocked in at each negative transition
of the SCK input and accepted on the
rising edge of the REQ input. The serial data stream must include various
mute, sleep and address codes as well
as the delay information before IC2
will respond.
Fig.4 shows the timing of the Data,
SCK (serial clock) and REQ (request)
lines to IC2. What happens is that
when the REQ line goes low (lower
trace), the serial data block (centre
trace) can be clocked in. In the time
that the REQ line is low, there are 12
clock pulses and these clock in the
respective data levels. Our data line
shows two positive pulses in the data
line but this is not the case as the data
stream actually contains 12 separate
codes which can be high or low.
On the first clock pulse, the sleep
data is fed in and this must be a low.
The following six codes are for delay
selection while the next three are the
low mute, ID1 and ID2 (identification
codes). The last two codes are high for
the ID3 and ID4 identification signals.
For the ID4 code to be valid, pin 7 of
IC2 must be also high.
With all this data complexity, it is
easy to see why a microprocessor is
the most elegant solution in Dolby
Prologic decoders, particularly when
it can provide a lot of other functions
as well.
IC5, a 74HC165 serial shift register
with parallel load inputs, is used to
supply the first eight bits of data. This
has the advantage that the data can
be initially set by parallel load inputs
(inputs A-H). The E, F and G inputs
are connected to a DIP switch to allow
Performance
Frequency Response................. (see graphs)
Fig.3 (left): the heart of this circuit is
the Mitsubishi M65830P digital delay
chip. Each time it is powered up it
needs a stream of serial data to set its
delay time. It can be set for between
0.5ms and 4.0ms using a DIP switch
(see Table 1). Once data has been sent
to IC2, CMOS chips IC3, IC4 and IC5
are effectively out of circuit.
Signal-to-Noise Ratio................. 96dB unweighted (22Hz to 22kHz); -100dB
................................................... A-weighted, with respect to 1V RMS.
Harmonic Distortion................... <0.5% at 1kHz and 1V RMS
Maximum Input Signal................ 1.2V RMS
Output Level............................... 0-1V RMS
Delay Options............................. 0.5-4ms in 0.5ms steps
June 1996 17
Above: bird’s eye view of the Stereo Simulator – there is not much wiring to be
done. Note that shielded cable is used for the connections between the board
and the RCA sockets.
Another view of the assembled PC board, prior to installation in the case. Take
care to ensure that all ICs are correctly oriented.
18 Silicon Chip
the delay to be selected. IC3 and IC4
are used to control IC5.
IC3 is a 4060 binary counter which
has its own oscillator, set by the components connected between pins 9, 10
& 11. IC3 supplies the clock signal for
IC2 at its Q4 output. Its Q5 output at
pin 5 is inverted by IC6a to drive IC4,
a 4022 divide-by-8 counter which has
eight outputs, O0 to O7. We use the O1
output to drive the serial input of IC5.
Finally, we use the QH output of IC5
to drive the data input of IC2.
When the “6” output of IC4 goes
high after 12 counts of SCK, IC3 is
reset, the REQ line goes high and IC5
is set into the load position with a
low pin 1.
At power up, the 47µF capacitor at
the input of Schmitt NAND gate IC6b
is high and its output is low. When
the capacitor charges, the pin 3 output
goes high to apply a short positive
pulse to the reset input of IC4 via the
.001µF capacitor. This resets IC4 and
the code is fed to IC2.
The resultant waveforms are shown
PARTS LIST
1 PC board, code 01406961, 100
x 100mm
1 plastic case, 111 x 45 x
140mm, Arista UB14
1 front panel label, 95 x 33mm
1 rear panel label, 95 x 33mm
1 12VAC 300mA plugpack
1 SPDT toggle switch (S1)
1 4-way DIP switch (DIP1-DIP3)
1 2MHz crystal (X1)
3 panel mount RCA sockets
1 insulated panel mount DC
socket
1 5mm ID rubber grommet
1 400mm length of hook-up wire
1 150mm length of shielded cable
1 250mm length of tinned copper
wire
7 PC stakes
Semiconductors
1 TL074, LF347 quad op amp
(IC1)
1 M65830P digital delay (IC2)
1 4060 binary counter (IC3)
1 4022 divide by-8 counter (IC4)
1 74HC165 8-bit shift register
(IC5)
1 4093 quad 2-input Schmitt
NAND gate (IC6)
1 7805 5V regulator (REG1)
1 7812 12V regulator (REG2)
1 1B04 bridge rectifier (BR1)
1 1N914, 1N4148 signal diode
(D1)
1 3mm red LED (LED1)
Fig.5: the parts layout and wiring diagram for the Stereo Simulator. Note
that the input and output leads are wired in shielded cable.
in Fig.4, as previously discussed.
Note that once the “6” output of IC4
goes high, it also pulls the reset line
of IC3 high and this stops any further
data being sent. Thus, IC3, IC4 and
IC5 serve no further purpose until the
circuit is powered up the next time.
That completes the circuit description except for the power supply. This
uses an AC plugpack fed to a bridge
rectifier (BR1) and a 470µF filter
capacitor. A 12V regulator supplies
power for the op amps in IC1 while
a 5V regulator supplies the rest of the
circuit.
Construction
Capacitors
1 470µF 16VW PC electrolytic
3 100µF 16VW PC electrolytic
2 47µF 16VW PC electrolytic
5 10µF 16VW PC electrolytic
2 4.7µF 16VW PC electrolytic
1 2.2µF 16VW PC electrolytic
3 0.1µF MKT polyester
2 .068µF MKT polyester
1 .012µF MKT polyester
1 .001µF MKT polyester
3 560pF MKT polyester or
ceramic
2 150pF ceramic
2 100pF ceramic
Resistors (0.25W 1%)
1 1MΩ
11 10kΩ
1 100kΩ
3 4.7kΩ
4 47kΩ
1 2.2kΩ
4 39kΩ
3 100Ω
1 22kΩ
1 33Ω
2 18kΩ
1 10Ω
The Stereo Simulator is assembled
June 1996 19
A view of the Stereo
Simulator with the
top removed and
showing the inside
of the rear panel.
onto a PC measuring 100 x 100mm
and coded 01406961. Our prototype
was housed in an Arista UB14 plastic
case measuring 111 x 45 x 140mm.
Self-adhesive labels, were fitted to the
front and rear panels.
The full wiring details and component overlay for the PC board are
shown in Fig.5.
You can start construction by
checking the PC board against the
published pattern of Fig.6. Fix any
broken tracks or shorts that may be
evident. Now insert the ICs, diode,
resistors and links in the locations
shown. Take care with the orientation
of the ICs, noting that IC1 is oriented
differently to the others.
DIP Switch Settings
DIP1
DIP2
DIP3
Delay
Freq.
on
on
on
0.5ms
2kHz
on
on
off
1ms
1kHz
on
off
on
1.5ms
666Hz
on
off
off
2ms
500Hz
off
on
on
2.5ms
400Hz
off
on
off
3ms
333Hz
off
off
on
3.5ms
285Hz
off
off
off
4ms
250Hz
The accompanying resistor colour
code chart should be used when selecting each resistor value. Alternatively,
This DIP switch can be used to change
the delay chip’s setting and thus the
stereo effect. Use Table 1 at left to set
the DIP switches.
use a digital multimeter to measure
each resistor before it is fitted into
the board.
RESISTOR COLOUR CODES
❏
No.
❏ 1
❏ 1
❏ 4
❏ 4
❏ 1
❏ 2
❏
11
❏ 3
❏ 1
❏ 3
❏ 1
❏ 1
20 Silicon Chip
Value
1MΩ
100kΩ
47kΩ
39kΩ
22kΩ
18kΩ
10kΩ
4.7kΩ
2.2kΩ
100Ω
33Ω
10Ω
4-Band Code (1%)
brown black green brown
brown black yellow brown
yellow violet orange brown
orange white orange brown
red red orange brown
brown grey orange brown
brown black orange brown
yellow violet red brown
red red red brown
brown black brown brown
orange orange black brown
brown black black brown
5-Band Code (1%)
brown black black yellow brown
brown black black orange brown
yellow violet black red brown
orange white black red brown
red red black red brown
brown grey black red brown
brown black black red brown
yellow violet black brown brown
red red black brown brown
brown black black black brown
orange orange black gold brown
brown black black gold brown
CAPACITOR CODES
❏
❏
❏
❏
❏
❏
❏
❏
Value
IEC Code EIA Code
0.1µF 100n 104
.068µF 68n 683
.012µF 12n 123
.001µF 1n 102
560pF 560p 561
150pF 150p 151
100pF 100p 101
Seven PC stakes will need to be
fitted to the board. This done, insert
and solder in the capacitors taking
care to orient the electrolytics with
correct polarity. Next, fit the 3-terminal
regulators and make sure you insert
the 7812 (REG2) into the location
nearest LED1. Insert the DIP switch,
crystal and bridge rectifier. The LED is
mounted without shortening its leads
and is bent over at right angles to insert
into the front panel hole.
The PC board is fitted into the case
and secured with four self-tapping
screws into integral standoffs in the
base. Affix the adhesive labels to the
front and rear panels and drill out
the holes for the power switch and
LED on the front panel and for the
RCA sockets and DC socket on the
rear panel. A 3mm hole is required
for the LED.
Note that the DC socket must be
insulated from the metal rear panel to
prevent shorting the AC plugpack to
ground. On our prototype, we fitted the
DC socket inside a 5mm ID grommet
and then secured it with a nut after
shaving the grommet thinner with a
sharp utility knife.
After fitting the front and rear panels, the final wiring can be done. Use
short lengths of shielded cable for the
input and output connections. The
connections from the DC socket to the
switch and PC board are made with
hook-up wire.
Apply power and use a multimeter
to check that pin 4 of IC1 is at +12V.
Pin 16 of IC3, IC4 & IC5, pin 14 of IC6
and pin 24 of IC2 should all be at +5V.
If the LED does not light, it is probably
connected the wrong way around.
Testing
To test the Stereo Simulator, connect
the mono input to the mono output
of your VCR and the stereo outputs
Fig.6: actual size artwork for the PC board.
+
STEREO
SIMULATOR
POWER
+
+
+
+
+
12VAC
INPUT
MONO
INPUT
LEFT
OUT
RIGHT
OUT
Fig.7: these full-size artworks can be used as drilling templates for
the front and rear panels.
to the left and right inputs on your
amplifier. This done, set DIP1, DIP2
and DIP3 on, apply power and listen
to the stereo effect.
Now set DIP 1 off and switch off
the power. Reapply power after about
10 seconds and check that the stereo
effect has changed. If that is the case,
the circuit is working correctly and
you can experiment with the delay
settings. Table 1 table shows the delay
versus frequency bands for various
settings of DIP1-DIP3. We found that
the most satisfying stereo effect was
obtained with the delay set to either
2ms or 2.5ms.
SC
June 1996 21
|