This is only a preview of the March 2018 issue of Silicon Chip. You can view 37 of the 104 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 "Arduino-based Earthquake Early Warning Alarm":
Items relevant to "Full Wave, 230V Universal Motor Speed Controller":
Items relevant to "An AM Radio Transmitter to build":
Articles in this series:
Items relevant to "Analog Audio/Video Modulator for Vintage TV sets":
Purchase a printed copy of this issue for $10.00. |
Using Cheap Asian Electronic Modules Part 14: by Jim Rowe
Banggood’s
RF Detector
This nifty RF Detector module from Banggood can measure the power of
RF signals from 1MHz to 8GHz, over a range of 60dB. It is on a tiny PCB
measuring 33 x 24.5mm and has an SMA RF input connector attached
to one end. It’s based on the Analog Devices AD8318 chip, which is an
enhanced version of the AD8307.
A
s a matter of interest, we used the
Analog Devices AD8307 chip in
the RF Level and Power Meter project
of October 2008 (siliconchip.com.au/
Article/1971) and also in the Arduino Multifunction Measuring Shield
of April-May 2016 (siliconchip.com.
au/Series/299).
Both the AD8318 and AD8307 are
logarithmic amplifier/detectors which
provide a DC output voltage proportional to the RF input power level.
But the AD8318 has a much wider
bandwidth of 1MHz to 8GHz, compared with the DC-500MHz range of
the AD8307.
While the AD8307 has a range of just
over 90dB, the AD8318 has a smaller
dynamic range of about 60dB (necessary to get the improved frequency
range).
Unlike the AD8307, which operates from a nominal supply voltage of
3V, the AD8318 is designed to operate
from 5V. It also has a typical supply
current of 68mA, compared with the
7.5mA drawn by the AD8307.
But perhaps the most important
functional difference between the two
devices is in terms of the output circuitry. The AD8307 has a current mirror in the output circuit which provides a positive slope to the DC output voltage. So the output voltage is
directly proportional to the RF input
level, with a slope of 25mV/dB.
In contrast, the AD8318 has a different output circuit designed to alsiliconchip.com.au
low it to be used for power amplifier
gain control. As a result, it provides
an output voltage which is inversely
proportional to the RF input, with a
slope of -25mV/dB. Is this a problem?
Not when you are going to use it in
conjunction with an Arduino or other
microcontroller.
Fig.1 shows a simplified version
of the circuitry inside the AD8318. It
has nine detector stages, interspersed
with eight cascaded gain stages. The
nine detector outputs are fed to an adder which drives a current-to-voltage
converter to produce the output voltage, Vout.
The V-I (voltage-to-current) converter at upper right allows adjustment
of the slope of Vout in measurement
mode. For example, the output slope
of -25mV/dB is achieved when the Vset
pin and the Vout pin are tied together.
Higher output slopes can be obtained by connecting a voltage divider
between the Vout pin and ground, and
feeding a fraction of Vout back to the
Vset pin. So if the voltage fed back to
Vset is Vout ÷ 2, this changes the output slope to -50mV/dB.
However, the output voltage is always in the range of 0.5-4.6V, so beyond -55mV/dB, the dynamic range
will be reduced as the output at lower
RF levels will be pegged at 4.6V.
Note that the AD8318 includes an
internal temperature sensor as well
as bias stabilisation circuitry for the
cascaded gain stages so that changes
Fig.1: simplified block diagram of the AD8318 logarithmic detector/controller. It
has nine detector stages interspersed with eight gain stages.
Celebrating 30 Years
March 2018 73
Fig.2: circuit diagram of the log detector module. Clpf and Cobp are optional
capacitors used to filter ripple from IC1’s output. Suitable values are 1nF for
Clpf and 10nF for Cobp with pads provided for mounting on the PCB.
in ambient temperature do not unduly
affect accuracy.
All this is squeezed into a tiny 4 x
4mm 16-lead LFCSP (SMD) package;
much smaller than the 8-pin SOIC/
PDIP packages used for the aforementioned AD8307.
Now have a look at the circuit for the
Banggood log detector module shown
in Fig.2. Apart from the AD8318 chip
itself (IC1), there is not much to it. The
only other IC is REG1, a 78L05 regulator in a SOT-89 3-pin package with tab.
This provides a regulated 5V rail for
IC1. But the 78L05 has a nominal dropout voltage of 2V, so the module needs
a power supply (Vcc) of at least 7.5V.
As with many modules, there’s one
LED to indicate when power is applied. LED1 is connected directly be-
tween the Vcc input and ground with
a 10kW series resistor.
CON1 is the RF input, an SMA edgemount socket. This is terminated via
a 51W resistor and then coupled to
the INhi input (pin 14) of IC1 via a
1nF capacitor, with a second 1nF cap
coupling the INlo pin of IC1 (pin 15)
to ground.
As the input resistance of IC1 between pins 14 and 15 is close to 1200W,
this gives the input circuit a low-frequency cutoff of around 300kHz. The
effective input resistance at frequencies below about 100MHz is around
49W (51W || 1200W).
Pin 16 of IC1 is the enable input,
which can be used to switch the device into a low-current standby mode
if desired, by pulling it to ground.
However, in the Banggood module, it’s
connected to the +5V line, so the chip
always functions while the module is
powered up.
But what’s the purpose of that 510W
resistor connected between pin 10
(Tadj) of IC1 and ground? It allows adjustment of the chip’s internal temperature compensation, to optimise
its operation at different frequencies.
A value of 510W apparently gives
very close to optimum compensation
at frequencies up to 2.2GHz, and also
at 8GHz, while optimum operation at
3.6GHz and 5.8GHz can be achieved
by changing RTadj to 51W or 1kW, respectively.
Even so, a value of 510W apparently
gives acceptable performance over the
whole range.
The two capacitors shown in red,
Clpf and Cobp, are used for filtering
any ripple in the output from IC1. If
both capacitors are omitted, the nominal output video bandwidth of the
AD8318 is around 45MHz, making it
suitable for demodulating pulse signals.
But if you’re using it purely for
measuring unmodulated RF, this wide
bandwidth can allow significant second-harmonic ripple to appear in the
output for input signals below 22MHz.
Since this ripple can cause measurement jitter, the simplest way to reduce its effect is to add either Clpf or
Cobp, or both. A suitable value for Clpf
is 1nF, while that for Cobp is around
10nF and these values give an output
bandwidth of around 100kHz.
By the way, neither of these capacitors are fitted to the module board
(even though pads are provided for fitting them as 0603 SMD components)
Banggood’s logarithmic RF Detector module detector module is based on the Analog Devices AD8318 chip. It has an
RF bandwidth of 1MHz to 8GHz with a range of -65dBm to +5dBm and an input impedance of 50W. These photos are
almost twice actual size.
74
Silicon Chip
Celebrating 30 Years
siliconchip.com.au
which is why we’ve shown them in
red in Fig.2.
Trying it out
To check out this module, I hooked
it up to a suitable 9V DC power supply and connected its RF input up to
a VHF/UHF signal generator.
Then I monitored its output using
a 4.5-digit bench DMM while varying
the RF input level over the range from
+10dBm to -70dBm, for four different
frequencies: 100MHz, 1GHz, 2GHz
and 4GHz.
I wasn’t able to go above 4GHz because that’s the highest frequency my
signal generator provides.
These measurement runs were used
to plot the module’s transfer characteristic at each of the four sample frequencies and the results are shown in
Fig.3. The four plots are very close to
linear between RF input levels from
-5dBm down to -60dBm and only
curve gently away at the upper and
lower extremes.
Although the truly linear part of
the module’s transfer characteristic
only covers about 55dB, the curved
sections at each end give it a useful
range of about 70dB as claimed in the
data sheet.
The linear sections of all four plots
are well within ±1dB of each other
and have a slope of -24.33mV/dB; very
close to the expected -25mV/dB.
Note that we’ve mentioned a 60dB
range before, as this is the practical
range over which you can expect to
get an accurate result.
Connecting to an Arduino or
Micromite
Interfacing this module to a micro is
straightforward. Just feed the module
with 7.5-9V DC and connect its Vout
to either one of the micro’s own ADC
inputs directly, or to a higher-resolution ADC coupled to the micro via an
SPI or I2C interface.
Then it’s just a matter of writing a
firmware sketch or MMBasic program
to read the analog Vout signal and convert it into an RF power level.
So this module should be suitable
for use as the sensor section of a homebrew VHF/UHF level and power meter.
You could even use it as an RF sensor
head for our Arduino Multifunction
Measuring Shield (MFM), although
its negative-slope transfer characteristic would require some changes to
the MFM’s firmware sketch.
Other uses would be in an RSSI (received signal strength indicator) for
UHF base station receivers and WLAN
routers. In short, it seems to represent
good value at around $16.50.
SC
Fig.3: plot of
the transfer
characteristic
for the AD8318
at four
different input
frequencies.
Note the
excellent
linearity.
siliconchip.com.au
Celebrating 30 Years
March 2018 75
|