This is only a preview of the November 1989 issue of Silicon Chip. You can view 46 of the 112 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 "The Radfax Decoder":
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
Articles in this series:
|
2-Chip
Stereo Radio
This month, we complete construction of the
AM Stereo Radio by giving the alignment
details. Provided you follow the instructions
carefully, you will finish up with a receiver
capable of true hifi performance.
By STEVE PAYOR
Precise alignment of this receiver
is a challenging task, requiring care
and patience. The circuit may appear to be working well even before
alignment has been attempted, but
this is more a testimony to the sensitivity of the MC13024 than
anything else. When it is fully aligned, it will work much better.
32
SILICON CHIP
To achieve the maximum stereo
separation ()25dB) and minimum
distortion (<1 % ), the RF and IF
response curves must be perfectly
symmetrical in both amplitude and
phase, either side of the carrier
frequency.
Not only does the alignment have
to be perfect but the tuning must
also be spot on - within a kHz or
so. Normally, such precision can
only be obtained with a digitally
synthesised front end.
However, this is where the phase
locked tuning of the MC13024
comes into its own. Provided the IF
circuits are all centred on 450kHz
and the reference oscillator is tuned to exactly 8 times this frequency,
the PLL will automatically centre
the signal carrier on 450kHz
whenever it is within the ± 3kHz
capture range.
Block diagram
Before we start randomly twiddling the coil slugs, it would be a good
idea to review the way all the tuned
circuits relate to one another, so we
VOLTAGE CONTROLLED
LOCAL OSCILLATOR
··rn
~
Jf,
r----1
I
I
c4
osc
I
PHASE
COMPARATOR
TRIMMER
450kHz
981-2052kHz
(:t 3kHz-VCO SHIFT)
%
llffi
C3
,fo
6MHz REFERENCE
OSCILLATOR
LS
3688
OSCILLATOR TUNED CIRCUIT
AERIAL TUNED CIRCUIT
f1
PHASE LOCKED LOOP '
CONTROL VOLTAGE
L2
T1342
GANGED TUNING
(CONSTANT 450kHz
DIFFERENCE)
11
450kHz REF
531-1602kHz
AERIAL
TRIMMER
,fo
2ND IF
TUNED CIRCUIT
1ST IF
TUNED CIRCUIT
MIXER
II
Ll
FERRITE ROD
AERIAL COIL
450kHz
CERAMIC FILTER
SFP-450D
L3
T1341
IF
AMPLIFIER
ro
450kHz
IF OUTPUT TO
DETECTOR ANO
STEREO DECODER
L4
T1340
I
--- - ----- ~----WIDE
-12kHz
t
+12kHz
-12kHz
SIGNAL CARRIER
FREQUENCY
t
+12kHz
450kHz
-12kHz
t
+12kHz
450kHz
Fig.12: this block diagram shows the tuned circuits that have to be adjusted during the alignment procedure. This
involves setting the IF stages to 450kHz, adjusting the aerial and oscillator circuits, and setting the 3.6MHz reference
oscillator.
can formulate a suitable plan of
attack.
Fig.12 is a block diagram showing just those circuit elements of importance to the alignment procedure. The response curves of the
RF & IF tuned circuits are shown
approximately to scale, to help you
visualise what is happening with
each adjustment.
The alignment process has four
main aims:
(1). Setting the centre frequency of
both IF tuned circuits to the centre
of the ceramic filter passband
(4g0kHz ± 2kHz).
(2). Setting the 3.6MHz reference
oscillator to exactly 8 times the IF
centre frequency.
(3). Adjusting the oscillator coil and
trimmer capacitor so that the dial
calibrations are correct over the
whole tuning range.
(4). Adjusting the aerial coil and
trimmer so that the aerial tuned circuit "tracks" the oscillator over the
whole tuning range.
Unfortunately, these four objectives cannot be achieved simultaneously by any instant "one shot"
alignment process - at least not
without a lot of test equipment. Instead, you need to repeat a series of
adjustments, with each round bringing you closer to the desired
result.
Tuning meter
The minimum requirement in the
way of test equipment is an analog
moving coil multimeter (remember
those'?). If you don't have one, buy
or borrow one now! Digital multimeters are almost useless here as
they make finding the optimum adjustment unbearably tedious.
If you have a 10M0 FET input
type [ie, the equivalent of the good
old VTVM), simply connect it between the AGC test point and
ground. Alternatively, for a more
expanded scale (useful if you are
trying to make do with a bargraph
display on a digital multimeter), try
connecting it between the AGC test
point and the + 1.0V reference. The
ACG voltage dips slightly below
+ 1.0V on strong signals and will
rise to around + 1.3V when the tuning is way off.
On the other hand, connecting an
ordinary moving coil multimeter
directly between the AGC line and
ground will kill the reception. This
is because the low input impedance
of the meter will pull the AGC
voltage down and turn off the
receiver gain completely.
To overcome this problem, we
have provided a "Buffered AGC"
test point. So if you are using an ordinary multimeter as a tuning indicator, connect it between the Buffered AGC test point and ground.
But regardless of what sort of
tuning indicator you are using,
remember this: optimum tuning
means minimum AGC.voltage.
Getting started
The following settings are ''ball
park" adjustments for all the coils
and trimmers, based on our experience with two prototypes:
• Slide the aerial coil (L1) along
the ferrite rod until it is flush with
the end of the rod.
• Set the oscillator coil slug (L2)
about one turn out of the top of the
can.
NOVEMBER 1989
33
Fig.13: the 'Sharp', 'Medium' or 'Wide' bandwidth option is
selected by bridging solder pads on the PCB. We recommend that
you select the 'Wide' option as shown at right. (Note: repeated
from last month's issue - see footnote).
• Set the aerial trimmer (associated with C3) and the oscillator
trimmer (across C4) to half mesh.
The aerial trimmer adjustment
screw is adjacent to the C3 marking
on the back of the tuning gang. The
oscillator trimmer adjustment is
below the C4 marking.
• Set the slugs of both IF coils (13
& 14) level with the top of the can.
Don't worry too much about the
3.6MHz reference oscillator at this
stage. In fact, it is better left way
off frequency to start with. This
will disable the phase locked loop
and make the initial alignment procedure the same as for any ordinary superhet receiver. (As long
as the tuning indicator is NOT lit,
you are in control of the tuning).
IF alignment
This is not a bad place to start
but remember that you will have to
come back and re-align this part
later on, when the other tuned circuits are a bit closer to their final
adjustment.
Tune in a clear, un-crowded station near the top end of the band
and peak the IF coils (13 & 14) for
minimum AGC voltage. Temporarily
peak the aerial trimmer as well and
re-check the IF coils.
Now swing the tuning knob a little to either side of the tuned station
and, using your ears, locate each
edge of the ± 12kHz passband of
the ceramic filter. The distortion
caused by the steep cutoff at each
edge is unmistakable. Temporarily
affix a slip of paper under the
perspex cursor and carefully mark
each edge of the passband, then put
the cursor line exactly in the
middle.
Before starting alignment, the tuning shaft must be rotated fully anti-clockwise
and the cursor aligned with the horizontal line on the dial. Don't forget to preload the Teflon washer and tuning gang bearings before tightening the
grubscrew.
34
SILICON CHIP
Don't worry at this stage that the
station frequency doesn't match its
location on the dial - this adjustment comes later. Also, if the tuning
indicator LED comes on, get rid of it
by de-tuning the reference oscillator.
Having found the middle of the
ceramic filter passband, go back
and re-peak the IF coils (13 & 14)
and the aerial trimmer again. The
reason for peaking the aerial tuned
circuit here is to prevent its
response peak from masking the
true centre frequency of the IF
coils.
As you can see from Fig.12, the
aerial tuned circuit is quite sharp in
its own right. This gives the
receiver excellent image rejection
and cross modulation performance
but is a bit of a nuisance during the
IF alignment.
Getting the dial
markings right
Contrary to what you may first
think, it is the oscillator frequency
which determines the receiver tuning and not the aerial tuned circuit.
Of course, the aerial circuit is tuned to the signal frequency but it has
nothing to do with positioning the
signal on the dial. It simply maximises the signal at a given position.
To get the cursor to line up with
the correct dial marking for a given
station, you have to adjust the
oscillator coil (12) and the oscillator
trimmer capacitor (across C4). The
trick is to always adjust the coil at
the low frequency end of the band
and the trimmer at the high frequency end.
First, set the dial to the exact frequency of a station at the low end of
the band and peak the oscillator
coil for maximum signal. This done,
move the cursor to a station frequency at the top end of the band,
then move the station to this position using the oscillator trimmer.
If you now go back to the low end
of the band you will find that the
first station has moved slightly (but
not very much), so re-adjust the
oscillator coil. Now you will find
that the station at the top end of the
band has moved a bit too, so readjust the oscillator trimmer.
After about the third pass, both
stations should be spot on.
This photo will allow you to quickly locate the various coils and trimmers during alignment. The 3.6MHz reference
oscillator (L5) should initially be left way off frequency to disable the phase locked loop (see text). ·
Getting the tracking right
This step can be done simultaneously with the oscillator adjustments but we are describing it
here separately for clarity. Basically the procedure is the same: you
adjust the aerial coil at the bottom
end of the band and the aerial trimmer at the too end.
The trimmer adjustment (at the
top end of the band) is straightforward but the aerial coil adjustment
is a little tricky, since you will have
to slide the coil along the rod with
your fingers. Now putting your
fingers on the coil will de-tune it
quite drastically, so it's best to
know which way to move it before
you touch it!
Try this simple test: tune to the
exact centre of a station at the low
frequency end of the band and note
the effect of de-tuning very slightly
to either side (by very slightly we
mean staying within the flat region
of the IF passband). If the indicated
signal strength increases to the low
side, then the resonant frequency of
the aerial circuit is a little low. To
correct this, you will need to slide
the coil towards the end of the ferrite rod, and vice-versa.
So now you know which way to
move the coil, although how much is
still a matter of trial and error. You
must remove your fingers from the
coil to check each position.
Of course, all this peaking of tun-
Where to buy the kit
A complete kit of parts for this project has recently been made available
by Dick Smith Electronics. You can buy it at your nearest DSE store or
by mail order from PO Box 321, North Ryde, NSW 2113 . Phone (02)
888 2105 .
The kit comes with all parts & even includes a low-cost pair of headphones to get you going . The front panel is of aluminium & is supplied
pre-punched together with an attractive adhesive label.
The price is $79.95 (does not include batteries) plus postage charges.
Quote Cat. K-5200 when ordering .
ed circuits assumes that the IF coils
are correctly aligned to start with.
Repeat the adjustments for these
once more before making the final
touchups to the oscillator and
aerial tuned circuits.
Remember: the oscillator adjustments set the stations to their
correct positions on the dial, while
the aerial adjustments peak the
reception at these positions.
Using stations at the extreme
ends of the band is helpful for the
initial tracking adjustments, since
there is minimal interaction between the coil and trimmer adjustments.
The final tracking adjustments
should be done using stations about
one quarter to one third of the way
in from each end of the band. This
will give a "best fit" linear approximation to the dial positions. With
careful adjustment, it should be
possible to get the station frequencies within a pointer's width of the
station position over most of the
tuning range.
Setting the
reference oscillator
Tune to a nice clear station and,
as before, mark the approximate
passband edges on a piece of paper
NOVEMBER 1989
35
Fig.14: these tuning scales are for Melbourne, Brisbane, Adelaide, Perth, Hobart and Launceston, and the Northern
Territory. They fit in the blank central area of the tuning dial. One approach is to make a transparency or adhesive
label from the artwork (try your local printer), which can then be affixed to the front panel label. Alternatively, you
can use a combined artwork to produce a single 3M Dynamark (formerly Scotchcal) label.
slid under the cursor. Now set the
cursor as close as possible to the
centre of the passband as your ears
and eyes can detect and adjust the
reference oscillator slug (15) with a
plastic alignment tool until the tuning indicator LED comes on.
Now carefully tune across the
passband and note whether the
"Locked" tuning range is centred
within the audible tuning range. It
will be fairly narrow - about 1/4 of
the IF bandwidth. If it is a little to
the high side of centre, screw the
reference oscillator slug a teeny
weeny bit out, and vice-versa.
Persevere until you have the locking
range exactly centred.
Note: With the reference oscillator coil, maximum inductance occurs with the slug near the top of
the can. The inductance decreases
as you screw the slug in. If you find
the slug needs to be screwed all the
way out to make the frequency correct, try screwing it in instead.
Stereo?
Once the "Locked" indicator is
lit, the receiver should switch to
stereo mode if you leave the tuning
knob alone for more than 300ms (ie,
the LED should jump from half
brilliance to full brilliance). If not,
then something is amiss with the
25Hz pilot filter circuitry or the station is not transmitting in stereo.
36
SILICON CHIP
The stereo mode is much more
sensitive to phase errors than mono
mode. For example, if you have
chosen the Medium or Sharp bandwidth options for the 2nd IF tuned
circuit, the alignment symmetry
will be quite critical.
With the Sharp setting, try detuning 14 ·slightly while listening to
a stereo signal with headphones.
The sound will appear to go in one
ear and out the other, or even
rotate around inside your head! Of
course, the stereo separation will
be thoroughly shot.
For this reason, the Wide setting
is recommended (at least initially)
since it is the least critical with
regard to alignment.
Other test equipment?
If you have a digital frequency
meter and are thinking of hooking it
up to the reference oscillator, then
forget it. Even the SILICON CHIP
1GHz DFM (which has a lower input capacitance than most commercial units) can't get a proper
reading, even with a 3pF divider
probe.
In any case, the MC13024 has a
habit of shutting down the reference oscillator at the first sign of
any disturbance.
The best approach is to set the
reference frequency indirectly.
Hook the DFM up to an RF signal
generator and adjust the signal fre-
quency to exactly 450kHz (unmodulated). Run a length of insulated wire from the signal
generator and lay it near the
MC13024 chip to inject a little of
this signal into the IF circuitry.
As you tune through a station,
you will hear a faint whistle which
decreases in frequency until the
LED comes on. After that, the whistle frequency will remain constant.
This constant audible frequency is
the error in the PLL setting so,
while the LED is on, carefully adjust
15 until the whistle becomes a low
growl, or even a rapid flutter if you
are very lucky.
Check the adjustment by tuning
through the station again. When the
whistle frequency becomes less
than a few kHz, it should suddenly
drop to a low frequency growl (as
the LED comes on) and stay there
until the LED goes off again.
Now you can peak all the IF circuits (13 & 14), knowing that when
the LED is on, the IF is exactly
450kHz.
However, while this is an accurate way of setting up the
reference oscillator frequency for a
450kHz IF, it doesn't help if the centre frequency of the ceramic filter
is 2kHz off (worst case). Once
again, you should perform an "eyes
and ears" verification that the locking range is centred within the
filter passband.
Final touches
With the reference oscillator
working, you can now make some
last minute touch-ups to the alignment. In particular, the local
oscillator can now be precisely adjusted so that the locking range is
symmetrical (about one cursor line
thickness) either side of the chosen
station markings.
Also, the IF coils can now be repeaked any time the LED is on, as
this guarantees that the IF signal is
centred correctly. You will appreciate that this is a classic
"chicken before the egg" situation
but that's life, as they say.
Operating hints
If you are listening to " fringe
area" stations with headphones,
the 9kHz inter-carrier whistle can
be annoying, especially at night. In
these cases, try rotating the radio
to minimise reception of the interfering carrier. The same applies
to other sources of man-made
interference.
On local stations, however, inter. carrier interference will not be a
+Trim to outside of board outline +
SILICON CHIP
AM STEREO
Fig.15: the Rev.2 version of the PC board will be available only with the Dick
Smith kit (see panel). The Rev.1 version (published last month) is available
from the usual suppliers.
problem, even if you use the Wide
setting. In any case, we plan to
describe an add-on whistle filter to
improve the audio quality from distant stations. Look for it in a future
issue.
Finally, if you are driving speakers (rather than headphones) and
the stereo mode drops out when you
cO
turn up the volume, it's time to
change the batteries.
Footnote: unfortunately, the colour overlay for Fig.5 in last month's
issue was not precisely positioned.
To eliminate any confusion about
which pads to bridge, the diagram
is reproduced in this article as
Fig.13.
~
AUDIO TRANSFORMED
FUJITSU MEMORY
• DYNAMIC RAM
25 6K X 1 DUAL IN LINE PKG
MB8 1256- 10P
100ns
5.00
1 Mb
1 Mb
2 56K
2 5 6K
MB 81C1000- 10P
MB S 1C 1000- 10PS Z
MB 8 1C 4256 - 1OP
MB 8 1C 4 256-10 PSZ
100ns
100ns
100ns
100ns
19.50
X
X
X
X
1
1
4
4
CMOS
CMOS
CMOS
CMO S
DIP
ZIG
DIP
ZIG
PACKAGE
ZAG PKG
PAC K AGE
ZAG PKG
4 Mb x 1 CMOS DIP / ZIP PK G
1 Mb x 4 CM OS DIP/ ZIP PKG
MBB 1C4000-80P/ PSZ
MBB 1C4400 - 80P/PSZ
BOns
BOns
P .O.A .
P.O.A.
• STATIC RAM
SK X 8 CM OS DIP LOW POWER
BK X 8 C MOS SK INNY DIP LO W PW R
3 2K X 8 CMOS DIP LOW POW ER
32K X O FULL CM OS LOW LO W PW R
128K X 8 CMOS 1 Mb LOW POW ER
MB8484 A- 10LP
MB8464 A - 10LP- SK
MB B4 256-1 0 LP
MB8 4F256 - 20LP
MB 84 1000-10P
1oons
6.00
100ns
100ns
200ns
100ns
15.50
P.O.A.
P.O.A.
• Add 20% Sales Tax / $5.00 P & P
PACIFIC MICROELECTRONICS PTY. LTD.
Suite 7, 1st Floor
Westlelgh Centre, Eucalyptus Dve,
WESTLEIGH
N.S.W.
2120
PHONE : 02 481 0065
FAX
: 02 484 4460
AUSTRALIAN REPRESENTATIVE
FUJITSU MICROELECTRONICS PACIFIC ASIA LTD.
FOR PROFESSIONAL, OEM OR AMATEUR
Broadcast qua lity aud io transformers ex-stock
More than 70 standard types ava ilable
Fast prototyp ing service for non-sta ndard types
Comprehensive data availab le o n request
Local ly manufactured
Compet iti vely price d
HARBUCH ELECTRONICS PTY LTD
90 George St. HORNSBY >!SW 20 77 Phone (021476-5854
N OVEMBER 1989
37
|