This is only a preview of the September 2017 issue of Silicon Chip. You can view 59 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 "Fully adjustable, 3-way active loudspeaker crossover Pt.1":
Items relevant to "Dead simple radio IF alignment with DDS":
Items relevant to "LTspice Tutorial Part 3: Modelling an NTC Thermistor":
Articles in this series:
Items relevant to "Arduino Data Logger Part 2":
Items relevant to "Arduino “ThingSpeak.com” ESP8266 data logger":
Items relevant to "El Cheapo modules Part 9: AD9850 DDS module":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
Dead-easy Superhet
IF Alignment using
Direct Digital Synthesis
• Touch-screen convenience
• Really quick and easy IF alignment!
This project is based on the
touch-screen Micromite DDS
Signal Generator project and makes aligning the IF stage
of superhet sets a snap, whether they are valve or transistor-based.
It also lets you examine the IF stage bandwidth, which gives a good
indication of the set’s selectivity, as well as the shape of the IF curve.
I
n the simplest terms, a superheterodyne AM radio works by mixing (ie, heterodyning) the radio
station signal with a tracking oscillator
signal that has a fixed frequency offset above (ie, super) that of the tuned
station.
The output of the mixer includes
components at the sum and difference
frequencies of the two input signals.
The following stages reject all but the
difference frequency and this carries
the same audio (amplitude) modulation as the incoming signal from the
radio station.
The difference frequency is known
as the Intermediate Frequency (IF)
and the IF circuitry normally comprises two stages with tuned resonant
circuits, each involving a transformer
with adjustable cores (slugs).
In more detail, the primary and secondary windings of each transformer
have parallel capacitors and their
cores need to be adjusted so that their
resonant frequency matches the IF, eg,
455kHz or 450kHz.
66 Silicon Chip
Adjusting the transformers in this
way maximises the gain of the radio
and the whole process is referred to
as IF alignment. IF alignment also
optimises the Q of each stage and this
increases the rejection of unwanted
signals (outside the tuned circuit’s
resonant range).
This has the effect of increasing the
selectivity of the radio which means
that it is easier to tune when stations
are crowded together on the dial.
Normal alignment also involves adjusting the antenna input circuits so
that stations at the top and bottom of
the dial (ie, the full timing range) are
actually received at the marked points
(ie, the station call sign or the transmitter frequency on the dial).
Note that some sets with a wide audio bandwidth (say 10kHz or more)
may have the IF transformer cores adjusted to slightly different frequencies,
say 447kHz and 463kHz, in the case of
by Nicholas Vinen
a 455kHz IF. This “staggered tuning”
gives a wider audio bandwidth but
slightly lower gain.
For more information on how a
superhet set works, see the AM Radio
Trainer project in the June 1993 issue;
it’s available as a PDF download from
our online shop at www.siliconchip.
com.au/Shop/5/3435
We also published a detailed description of the operation of the IF
stage in the December 2002 issue; see
www.siliconchip.com.au/Article/
6698
Aligning the IF stages
There are a number of methods by
which you can do alignment on an
AM radio but the simplest approach
involves injecting a signal into the set
which can be set to the intermediate
frequency.
If this signal is modulated (typically at 400Hz), you can easily judge
the effect of your adjustments by the
loudness of the tone in the radio’s
loudspeaker. That means you need a
siliconchip.com.au
It’s all housed in a small Jiffy Box . . . and if
you’re into restoring vintage radios, for example, you’ll
find this the best thing you’ve ever seen since sliced bread!
modulated RF oscillator which can be
set to precisely 450 or 455kHz.
It is also desirable that its output
is a clean sinewave, ie, with few harmonics to cause problems in the alignment results.
Unfortunately, the output waveform
of most old valve and transistor RF
oscillators is surprisingly distorted
and their output amplitude can also
vary significantly as the frequency is
changed.
But there is a much easier and more
elegant way and here is where modern
technology comes to the rescue.
Sweep oscillator
What we would really like is to plot
of the set’s detector output against the
injected frequency so we can actually see what the IF stage frequency
response looks like.
That’s just what this project does.
It produces a signal which is swept
over a range of frequencies around the
nominal IF and it measures the output
of the voltage detector (usually a diode
just preceding the volume control).
The varying DC output can then be
SWEEP
OSCILLATOR
plotted on an LCD screen.
You can set the centre frequency and
span and it automatically scales the
vertical axis and adds cursors showing the peak frequency and (if visible)
-3dB points.
That makes doing the IF alignment,
and even setting the IF bandwidth,
easy!
But we are getting ahead of ourselves. Fig.1 shows the concept. The
sweep oscillator can be thought of as
an oscillator which can be set to vary
in a linear fashion from say, 440kHz
to 470kHz, repeatedly.
This signal is connected to the input of the IF stages and the output of
the detector is connected to an oscilloscope.
But we have combined the sweep
oscillator and the oscilloscope screen
into the one unit.
For the sweep oscillator, we’re using a Direct Digital Synthesis (DDS)
module based on the Analog Devices
AD9833 IC.
Then we’re using the Micromite
LCD BackPack to provide the oscilloscope function, to display the result.
AM RADIO
Because the Micromite is controlling the DDS, it can synchronise the
plotted result on the screen with the
frequency of the sweep oscillator.
The hardware used in this project is
pretty much the same as that in the Micromite BackPack Touchscreen DDS
Signal Generator that was published
in the April 2017 issue.
The main changes are to the software, to provide the sweep and plotting function. There’s just a slight
change hardware, to provide the required analog voltage measurements.
Circuit operation
The circuit diagram for the DDS IF
Alignment unit is shown in Fig.2. Most
of the work is done by the Micromite
software running on the BackPack and
the arbitrary waveform generator module which contains the AD9833 IC.
If you compare this diagram to the
one from the Touchscreen DDS Function Generator in the April issue (on
page 70), you will see a few minor
changes.
Firstly, we have changed the coupling capacitors from the PGA (pro-
DETECTOR
OUTPUT
IF
Fig.1: an overview of how this unit can be used to plot the frequency response of the IF stage in a radio. A sinewave
signal is produced which sweeps from just below the intermediate frequency to just above and this is injected into the
set via its antenna. The detector voltage is then plotted against the sweep frequency on an LCD screen to produce a
frequency response plot. Note that the sweep oscillator’s output is not amplitude modulated.
siliconchip.com.au
September 2017 67
Fig.2: circuit diagram for the DDS IF Alignment
unit. It consists primarily of the Micromite LCD
BackPack at left, wired to an AD9833-based DDS
module at centre. The DDS module produces
the sweep signal at the output connector and the
resulting DC detector voltage is applied to the input
connector and then fed back to the Micromite, to be
measured and plotted on the touchscreen.
grammable gain amplifier) output of
the DDS module to the output connectors to a single 10nF 630V type,
primarily to provide protection for the
DDS module from accidental connections to HT voltages in valve radios.
We have also added a 10kΩ resistor
in series, to limit inrush current in the
case of a short circuit.
This offers the possibility of inject-
ing the signal into HT-biased parts of
the circuit but as we will see later, that
is generally not necessary.
We’ve omitted the attenuated output terminal since you can adjust the
sinewave amplitude output of the DDS
via the touchscreen and you can also
control the amount of signal coupling
into the radio antenna by how closely
you place the leads (more on that later).
Fig.3: the modified main screen from Geoff Graham’s DDS
Signal Generator. Note the new “IF Align” button at centre
left. You can still use the unit as a signal generator, with all
the same functions of the original unit. We simply added
the extra functions required for IF alignment, accessed via
this new button.
68 Silicon Chip
We haven’t bothered with any DC
biasing of the output since that will
generally be accomplished in the set
if you are using direct signal injection.
In place of the trigger output used in
the original DDS Generator project, we
have an analog input that’s intended to
monitor the DC output of the detector
or AGC (automatic gain control) signal.
This gives the unit direct feed-
Fig.4: we hooked our test unit up to an HMV 64-52 “Little
Nipper” valve superhet and this is the result. The plot
shows that the IF stage needs some re-alignment as its peak
response is not at 455kHz. Note the cursors indicating the
peak and (approximate) -3dB points. The output lead was
simply placed near the ferrite rod antenna while the output
of the detector was taken from the top of volume control
pot VR1 (which doubles as the AGC signal, fed to R4).
siliconchip.com.au
back on the amount of signal passing
through the IF stage. This goes back
to pin 24 on the BackPack since this
is an analog input.
It’s protected from accidental high
voltage application via a 4.7MΩ series
resistor and this also forms a divider
with the 1MΩ resistor to pin 22, if pin
22 is actively driven.
If pin 22 is left floating by the software, it has little effect on the voltage
at pin 24.
For radios which have a negative
AGC/detector output (the majority),
pin 22 is driven high, to +3.3V. This allows pin 24 to measure voltages down
to -15.5V (3.3V x -1 x [4.7MΩ ÷ 1MΩ]).
To measure positive voltages, pin 22
can be left floating for high sensitivity
(0-3.3V) or driven low for low sensitivity (0-18.8V) measurements. This is
all under the control of the software.
We won’t go into a great deal of detail on the operation of the AD9833
DDS module.
This was covered in a dedicated article in the April 2017 issue, starting
on page 18 (see www.siliconchip.com.
au/Article/10608).
It was also explained in the article
on the DDS Signal Generator in the
same issue.
In brief, software running on the
LCD BackPack sends commands to the
DDS module over a three-wire SPI (serial peripheral interface) bus comprising pins SCLK (clock), SDATA (data)
and FSY (module select).
The same SPI bus is used to communicate with a digital attenuator in the
same module, except that the CS (chip
select) line is pulled low when communicating with it, rather than FSY.
By sending serial commands to the
AD9833, the PIC32 in the BackPack
can set the output waveform type (sine,
triangle, square), the frequency (from
0.1Hz to 12.5MHz), the phase and it
can also put the AD9833 IC into lowpower sleep mode, or wake it up.
By sending commands to the digital attenuator, the output level can be
changed in 255 steps, over a range of
about 4mV to 1V RMS.
Software operation
The software for this project is based
directly on the software for the DDS
Signal Generator from April 2017 and
retains all the original features of that
project.
We’ve simply added an “IF Align”
button to the main screen (see Fig.3).
siliconchip.com.au
Parts list – DDS IF Alignment
1 2.8-inch Micromite LCD BackPack kit with microcontroller programmed for
DDS IF Alignment (DDSIFAlign.HEX), laser-cut lid and mounting hardware
(SILICON CHIP online shop Cat SC4021)
1 DDS Function Generator module with AD9833, AD8051 and MCP41010 ICs
(SILICON CHIP online shop Cat SC4205)
1 UB3 plastic Jiffy Box
4 M3 x 10mm Nylon machine screws
12 M3 Nylon hex nuts
11 short single pin female-female DuPoint jumper leads
(Jaycar WC6026; set of 40)
1 USB charger with USB-to-DC-plug cable (see Fig.7)
1 chassis-mount DC barrel socket, to suit cable
2 chassis-mount BNC sockets
1 10nF 630V polyester capacitor
1 4.7MΩ 1W resistor
1 1MΩ 0.25W resistor
1 10kΩ 1W resistor
Once you’ve set up the generator to
produce a sinewave at the expected intermediate frequency, press this button
and the unit will go into sweep mode.
By default, it will sweep from 10kHz
below the current centre frequency
to 10kHz above (ie, a span of 20kHz).
Each sweep takes a couple of seconds.
To do a sweep, the unit first sets the
DDS output frequency to the lower
end of the sweep range, then after a
short delay, measures the voltage at
the detector input. It then increases
the output frequency by 1/80th of the
span and measures the detector input
voltage again.
Once it has at least two measurements, it updates the display with a
short line segment, forming that portion of the IF curve plot.
This process is repeated until the
frequency is at the top of the span (ie,
after 80 steps) and the curve plot is
complete.
The unit then repeats this process
forever, so that the plot is constantly
being updated.
Each time a sweep is completed, it
analyses the data and finds the maximum value, then draws a cursor,
which includes text that shows the
peak frequency and voltage reading,
plus a vertical line down to that part
of the curve.
It then looks for the -3dB points on
either side of this peak and if found,
draws cursors for them too, including
the frequency readings.
The mode buttons that are normally
at the bottom of the screen in the DDS
Signal Generator are still present in
sweep mode, so pressing any of these
will take you out of sweep mode and
back into one of the normal signal generator modes.
Other areas of the screen can be
touched to change the sweep parameters.
You can press on the centre frequency, at the bottom of the plot, to change
it (a keyboard will appear). Similarly,
touching either the lowest or highest
sweep frequency in the bottom corners
will let you set the frequency span.
If you press on one of the cursors at
the top of the screen, you will change
the cursor update interval.
Normally they are updated each
time a sweep is completed but you
can set them to change on every second or fourth sweep, to give you more
time to read them off, by pressing on
the cursors.
The first number in the top-right
hand corner of the plot (before the
comma) indicates the current cursor
sweep update interval.
The second of these two numbers
indicates the detector voltage input
mode. The default mode is “1” which
inverts the voltage measured and gives
a maximum input reading of around
-16V.
In this mode, the pin 22 output is
driven high, in order to shift negative
input voltages up into the range of
0-3.3V, so the micro can measure them.
Pressing on the middle of the screen
will change this mode to “2”, which
sets the pin 22 output low.
Thus, the unit measures positive
voltages, from 0V up to around +19V.
Pressing again will change the mode to
“0”, which causes pin 22 to float and so
September 2017 69
Here’s how it all fits inside
a UB3 Jiffy Box, albeit
with a new laser-cut
acrylic front panel. The
10kΩ 1W resistor attached
to the upper BNC socket
appears to go to nowhere
in this photo; in fact it
is soldered to the 10nF
capacitor immediately
below it. Similarly the
orange cable connecting
to the BackPack solders
direct to the end of the
4.7MΩ 1W resistor. Note
also the small piece of
strip board attached to
the MicroMite BackPack
PCB – we used this to
more firmly anchor the
1MΩ 1W resistor which
connects between pins 22
and 24 of the BackPack.
Incidentally, 1W resistors
were chosen not for their
power dissipation but
instead for their voltage
ratings, assuming the DDS
module will be used with
the higher voltages of
valve radios.
the input voltage measurement range
is 0-3.3V. Another press will take you
back to mode 1.
The input impedance is around
5MΩ, regardless of mode.
Note that current does flow into pin
24 when making analog measurements
and the high source impedance of
4.7MΩ, due to the series resistor, will
cause errors in the readings.
But the whole measurement process
is quite approximate, due to various
factors such as AGC operation, imperfect coupling of the test signal into
the set, non-linearity in the detector,
background noise being picked up by
the set’s antenna (unless it is disconnected), etc.
In general, the measurements are
close enough to get a pretty good plot
of the IF stage’s response and make
any necessary adjustments.
online shop.
You can use the plain BackPack kit (www.siliconchip.com.au/
Shop/20/3321) and load the BASIC
code for the DDS IF Alignment yourself, using a USB/serial adaptor and
the free MMEdit software.
Or for the same price, you can pur-
chase a kit with the software pre-loaded on the microcontroller from www.
siliconchip.com.au/Shop/20/4021
Both kits are supplied with a lasercut lid to replace the UB3 jiffy box lid,
with the required cut-out and holes
already drilled. The kits also come
with the hardware needed to attach
Construction
The majority of the assembly required for this project is to build the
LCD BackPack module. This is available as a kit from the SILICON CHIP
70 Silicon Chip
Fig.5: this diagram shows how the LCD BackPack is attached to the
underside of the 3mm laser cut lid, while the DDS module is mounted in
the bottom of the jiffy box.
siliconchip.com.au
103K
630V
Fig.6: follow this diagram to make the connections between the LCD
BackPack, DDS module and input/output sockets. The components
between the PGA output on the DDS module and the output connector
can be made as shown here while you may prefer to mount the other
two components on a small piece of prototyping board, as we did for
our prototype.
the module to the lid.
Assembly is quite straightforward,
simply fit all the parts where indicated
on the PCB silkscreen label.
For full details, see the February
2016 article describing the BackPack
(www.siliconchip.com.au/Article/
9812) but most constructors won’t
have any trouble figuring it out.
Make sure the 28-pin socket goes in
with its notch in the position shown
and when you plug the micro into its
socket, its pin 1 dot needs to go near
the notch.
The female header for the LCD and
6-pin right-angle in-circuit serial programming (ICSP) header both go on
the same side as the micro and related
components, while the two vertical
male pin headers for the input/output
connections are soldered on the back.
Regarding the three 10µF/47µF capacitors, note that they were shown
as through-hole tantalum types in the
February 2016 article, and you can use
these, but we prefer to use SMD ceramics as they are more reliable and this
is what is supplied in the kit.
The ceramic capacitors are not polarised and the PCB has pads to suit
either type.
The kit is normally supplied with
two SMD capacitors in one pack and
one in another; the one by itself is the
47µF type.
However, it doesn’t actually matter
where you solder them since we only
specified 47µF for VCAP in case tantalum capacitors are used.
When ceramic capacitors are used,
10µF is sufficient for all three. This
has been a point of confusion for some
constructors who have ordered kits.
Once the module is complete, power
it up to make sure it works and then
attach it to the underside of the lid
with the supplied 1mm thick Nylon
washers as spacers.
The touchscreen is held onto the
main board by screws which pass
through the lid, these spacers, the
LCD module and then into the spacers
mounted on the main board. The overall arrangement is shown in Fig.5.
Final assembly
The next job is to place the DDS
module in the bottom of the case and
mark and drill mounting four 3mm
holes, then attach it to the inside of
the case using Nylon machine screws
and nuts, as shown in Fig.5.
This module should be mounted to-
wards the right-hand end of the case,
around 60mm from the end, with the
output connector to the right.
The only other holes you need to
drill are two in the right side of the
case for the BNC sockets (10mm) and
one in the left side for the DC power
socket (8mm).
You can then mount those sockets
and solder the extra components as
shown in the wiring diagram, Fig.6.
The easiest way to do this is to trim
the leads of the 10kΩ resistor short
and solder one to the central pin of
the output socket.
One end of the 630V capacitor can
be soldered to the PGA output of the
AD9833 module before that module
is installed in the case, then trim the
remaining lead and solder it to the free
end of the 10kΩ resistor.
The 4.7MΩ resistor can also be soldered directly to the centre pin of the
input socket and then a short wire
run back to pin 24 on the BackPack
I/O header.
We made up a little plug-in board
out of a piece of prototyping board,
with the 1MΩ resistor onboard and a
header for this wire to plug into so that
we could easily remove it later if we
Fig.7: this power supply cable is made from a USB cable cut short, with a DC plug soldered onto the end. It plugs into a
USB charger, which is a cheap and readily available source of regulated 5V. The unit can also be run from a USB power
bank or the USB port of a computer. The wires inside the USB cable should be colour coded; solder the red wire to the
inner conductor, the black wire to the outer barrel and cut short and insulate the white and green (USB signal) wires.
siliconchip.com.au
September 2017 71
There’s the old way, using a 455kHz generator and
a ’scope to monitor the waveform (and lots of time!)
. . . and the new way, using the touch-screen DDS to
perform the alignment much more easily. Note that
while the oscilloscope’s vertical scale is showing
peak voltage, the display on the DDS Alignment Unit
has a logarithmic vertical scale (ie, it reads in dB) so
the shape of the curve is different. However, they are
effectively displaying the same thing.
needed to. You could solder the 1MΩ
resistor directly between the pins to
save time.
With the four extra components in
place, all that’s left to do is wire up the
various connections using the jumper
leads, as shown in Fig.6, plus the two
wires to the DC socket.
Where you need to go from a header
pin to a soldered connection, you can
simply cut the DuPont socket off one
end of the wire, strip it back and then
solder it in place.
The other end can then just be
plugged in; see the internal photo for
more details.
Now double-check that you have
wired up the DC socket with the correct polarity before powering the
unit up because there’s no protection
against reverse polarity!
The easiest way to do this is to unplug the +5V connection from
the BackPack board (check the
silkscreen labelling to see which
one this is) while leaving the earth
connection attached.
Apply power, then measure between the disconnected pin and the
outer shield of one of the BNC sockets with your DMM, with the black
lead to the BNC socket shields.
If you get a positive reading on the
DMM, close to +5V, plug the cable back
in and the unit should spring into life.
72 Silicon Chip
Once you’ve verified that it’s all
working, you can attach the lasercut lid to the case with the supplied
self-tapping screws and the unit is
complete.
Note that as the lid is slightly thicker
than the one originally supplied with
the case, and doesn’t have recesses
for the screw heads, it’s possible you
may need to substitute longer screws;
we find the ones supplied with UB3
boxes from Jaycar are just long enough.
That’s it, you are ready to start alignSC
ing radios.
Reprinted from the April 2017
feature on the AD9833 module
(siliconchip.com.au/Article/10608)
this shows the circuit of the AD9833-based DDS module
used in this project, The output is taken from the socket
labelled PGA and AGND (lower right).
siliconchip.com.au
|