color=blue>the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Experiment #53 RF Peak Detector
From June 2007 QST � ARRL
H. Ward Silver, N�AX PO Box 927, Vashon, WA 98070 n0ax@arrl.org
Experiment #53 RF Peak Detector
N�AX
HANDS-ON RADIO
This experiment is really a three-fer.
Not only do you get the peak detector, but
also a dummy load! And wait, theres more
measuring RF power with an oscil-
loscope! Youll have a useful instrument
and a new shack accessory, and youll learn
some valuable techniques by the time the
dust clears.
Term to Learn
Detect
Recover modulating informa-
tion from a waveform.
The Envelope Detector
A detector is a circuit that recovers infor-
mation from any type of modulated wave-
form. Different types of detectors are used for
AM, FM, PM, SSB and other modes. Most
hams use the term to mean envelope detector,
a circuit whose output is the envelope of an
AM signal. A typical envelope detector is
shown in Figure 1.
This envelope detector is basically a
half-wave rectifier. The input signal source
develops a voltage across R1. (R1 can also be
the output impedance of the signal source.) If
the voltage is greater than that across C1, cur-
rent flows through diode D1 increasing the
voltage across C1 until the voltages are equal.
Once C1 is charged, it discharges through
R2, which can be the input impedance of a
following circuit, such as an audio amplifier.
The voltage drop across D1 depends on both
the semiconductor material and the current
when the diode is conducting. The forward
voltage of a silicon diode such as a 1N4148
is close to 0.6 V when fully on, while a ger-
manium diode, such as a 1N34A will have a
lower drop, typically 0.3 V.
The input signal to a typical envelope de-
tector is an AM waveform whose carrier, f
C
,
is many hundreds of times higher in frequency
than the highest modulating signal frequency,
f
Max
. For example, the carrier of an AM
broadcast station on 1000 kHz is 200 times
higher than a 5 kHz modulating frequency.
This means C1 has to charge very quickly and
discharge very slowly to separate the RF and
AF components of the AM signal.
The discharge time constant
2 = C1 �
R2 should be chosen so that C1 discharges
just slowly enough to reproduce the high-
est modulating frequency, f
Max
. An ap-
proximation for the minimum value of
2 =
1 / (4 � f
Max
). To recover human voice audio
(f
Max
of 3 kHz),
2 = 83 �s. If R2 = 10 k,
then C1 = 83
�s / 10 k = 0.00833 �F and a
0.01
�F capacitor will do nicely.
The combination of R1, R
D
and C1 form
a low-pass filter with f
C
= 1 / [2
� C1 �
(R1+R
D
)]. This removes the carrier com-
ponent from the output. R
D
is the forward
resistance of D1 and depends on the amount
of current flowing through the diode. R
D
can
be estimated as
V
f
/
I
f
, for values of I
f
that
will be encountered in operation. For example,
from a 1N4152 data sheet (enter
1N4152 DATA
SHEET
into an Internet search engine), V
f
for
the 1N4152 is about 0.52 V for I
f
= 0.1 mA and
0.62 V for I
f
= 1.0 mA so R
D
= 0.1 V / 0.9 mA
= 111
. If R1 = 50 , R
D
= 100
, and C1 =
0.01
�F, the low-pass filters cutoff frequency
is approximately 106 kHz, attenuating carrier
components above that frequency.
The Peak Detector
An envelope detector does not make a
very good power measuring device because
its output changes too quickly. Whats needed
is a peak detector whose output corresponds
to the peak value of the envelope instead of
individual modulating waveform cycles.
Theres no need to change the input time
constant,
1. The carrier, after all, still has to
be removed. Whats needed is to lengthen
2
so that the output stays at or near the peak
value of the envelope long enough to be mea-
sured. If R2 is removed completely, then C1
will discharge only through its own and D1s
leakage current. The voltage across C1 can
be read by either a built-in voltmeter or by an
external voltmeter such as a DVM or VOM.
Figure 2A shows a workbench peak detec-
tor for low-power signals up to 10 W or so.
D1 is a 1N34A germanium diode to increase
the sensitivity of the detector and R1 is 50

to present a standard load to the circuit under
test. R2 is replaced by the very high imped-
ance of an external voltmeter. C2 is increased
to 0.1
�F to increase 2 and hold the peak
voltage steady for a stable reading. If the
voltmeter has a 10 M
input impedance, 2 =
0.1
�F � 10 M = 1 s. (Remember meg-
ohms times microfarads equals seconds!)
Dummy Loads and Power
Measurement
Well use a transceiver as a signal source.
To do so, youll need a dummy load to which
you can connect the peak detector circuit, so
in the true ham spirit, well make our own.
(Check your rigs manual for instructions on
reducing output power below 5 W. You may
have to use the ALC input.)
A good option for a single-resistor dummy
load is an Ohmite TCH35P51R0J; a 51
,
35 W resistor in a TO-220 transistor package,
available from Mouser Electronics (www.
mouser.com/ohmite) for less than $6. The
Figure 1 The basic envelope detector
circuit lters out the carrier signal and
outputs only the modulating signal that
creates the envelope. The time constant of
C1 and R2 must be low enough to track the
highest modulating frequency.
Figure 2 At A is a peak detector circuit
that uses an external voltmeter in place of
R2. The input resistor is 50 to present
a good match to most generators and
transmitters. At B is one way to construct
a dummy load out of multiple low-power
resistors in parallel. From June 2007 QST � ARRL
HANDS-ON RADIO
case of the resistor is electrically isolated so
you can bolt it directly to a metal heat sink.
If your junk box is well stocked, you can
construct a dummy load from multiple high
value noninductive resistors whose com-
bined resistance is 50
. For example, 10
510
, 2 W resistors in parallel can make
a 51
, 20 W resistor, if theres enough air
between them. Be sure to use noninductive
resistors not wirewound or film resistors.
Use two strips of solderable metal such as
brass or copper or PC board stock as shown in
Figure 2B. Drill holes spaced to allow some
airflow between the resistors and solder the
resistors to the strips. Attach an SO-239 or
BNC coaxial connector at one end as shown.
Keep all leads short so that the impedance
stays close to 50
at high frequencies.
When you calibrate your peak detector,
youll need to measure the RF power from the
transceiver accurately. This requires an oscil-
loscope. To avoid any impedance changes
due to transmission line effects, use a 10�
probe connected directly to the dummy load
as shown in Figure 3.
Calculate peak envelope power from the
voltage measurements as follows:
PEP (watts)
= V
RMS
2
/ 50
PEP (watts)
= V
peak
2
/ (2 � 50)

= V
pk-pk
2
/ (8 � 50)
Building a Peak Detector
Start by building the peak detector circuit
on a solderless prototyping board. Set your
transceiver to output a low-power AM or
SSB signal (5 W or less) at the bottom of the
160 meter band. Attach the oscilloscope probe
directly to the dummy load at the coax con-
nector. Use short wires to connect the dummy
load to the detector circuit on the prototyping
board.
Verify that the circuit works by measuring
the output voltage at several different power
levels. Youll notice that the output voltage
falls rapidly once below 0.5 V. This is due
to the 0.3 V forward drop of D1. If the input
signal is not greater than 0.3 V
pk
, the diode
does not turn on very strongly and little current
is available to charge C1. When you speak
into the microphone, you should see the peak
reading jump to a higher level as the waveform
envelope tracks voice peaks. Change the value
of C1 to higher (add more capacitors in paral-
lel) and lower values and observe the effect on
how the detector responds to your voice.
Once youve verified that the circuit works,
build it permanently on a terminal strip with
three to five terminals. Use the mounting lug
as ground. A BNC connector is a good choice
for the input. You can use binding posts, or
just a pair of wires with tinned ends, as your
contact points for the voltmeter be creative
and use whatever is handy to make the volt-
meter connection. Now find a metal enclosure
big enough for your peak detector, including
the dummy load. (Hint the enclosure can
also act as a heat sink!) Assemble the dummy
load and detector circuit inside the enclosure.
Youre ready to calibrate!
The 1N34A diode can withstand a maxi-
mum of 65 V
pk
representing a power of
42.3 W and the absolute maximum your detec-
tor can withstand. Set your voltmeter to the 10
V scale and attach it to the detector. Attach the
transceiver to the detector input and set power
so that the voltmeter reads full-scale. (10 V
pk

across 50
is 1 W.) Record the oscilloscopes
peak (or peak-to-peak) voltage reading and
convert to watts. Reduce power and make
ano