[ Pobierz całość w formacie PDF ]
pseudo-logarithmic display
for the microwave
spectrum analyzer
In a recent article I described a microwave spec-
trum analyzer which covered dc to 2.5 GHz with up
to 2 GHz of dispersion, 2 MHz resolution, and 50 dB
of dynamic
ranqe.l
This analyzer was built almost
completely from surplus materials and has been well
received by the amateur microwave community.
However, the instrument has one drawback: the dis-
play graduations are linear rather than logarithmic.
This limitation was discussed in the original article,
and reader suggestions were solicited.
Before my spectrum analyzer article appeared (but
after the manuscript was finalized)
ham radio
pub-
lished a very fine article by Jeff Walker, W3JW, on
the design and
construction
of a high resolution
high-frequency spectrum analyzer.
2
In that article
Walker described a simple and effective circuit for
providing his analyzer with a pseudo-logarithmic dis-
play which allowed him to view 40 dB dynamic range
at one vertical deflection setting. It seemed to me
that this circuit would, with suitable modification,
greatly enhance the performance of my analyzer. I
am pleased to report that it did just that.
This pseudo-logarithmic
circuit for your
home-built microwave
spectrum analyzer
provides good resolution
and 40 dB dynamic range
circuit description
Walker's circuit, shown in fig. 1, consists of an
audio-frequency detector, lowpass filter, and a
unique nonlinear diode limiter arrangement. My ana-
lyzer already included an i-f detector diode, the out-
By H. Paul Shuch, N6TX, Microcomm, 14908
Sandy Lane, San Jose, California 95124
34
fII
july
1978
tudes are "stretched" somewhat, as seen in the
scope photograph. However, it is possible to meas-
ure signal amplitudes to within one or two dB over
the entire 40 dB range, once you get the hang of it. It
is possible to view spectral components as far down
as - 40 dBm, but scale compression at the low end is
so great that you can only guess at the actual am-
plitude.
RI
R2
I
ANAL'r2ER
10k
56k
....
,
OUTPUT
L
TO SCOPE
CR2
e-
CRI
+_'-
C/
AUOIO
RJ
IN91.
"
INPUT
~
~
IN4D04
.e-r-,
0.5
CRJ
.....
Ik
IN9/4
~
fig. 1. Pseudo-logarithmic signal-processing circuit
developed by W3JW for use in a high-frequency spectrum
analyzer.
2
calibration
The display response indicated in the photograph
was achieved on my analyzer with i-f attenuation set
at a
minimum
and video sensitivity at
50
mV / cm. The
put of which I applied to Walker's filter/limiter cir-
cuit. However, I found it necessary to change the
value of
Cl
to achieve the desired video frequency
response at high sweep speeds (a value of 1000 pF is
acceptable for sweep speeds of up to
60
Hz). For the
logarithmic shaper circuit I replaced the 1N914
switch diodes with general-purpose Hewlett-Packard
hot-carrier diodes. The final circuit values are shown
infig.2.
Note that the detector circuit I used in my original
analyzer provides a positive-going video output. If
one of the more common negative-output detectors
were used, it would be necessary to reverse the po-
larity of the Schottky diodes in the logarithmic shap-
er circuit.
VIDEO OETECTOR
(SEE
TO SCOPE
VERT/CAL
AMPLIFIER
REF.
I)
ro»
56'
>-----i-+--I__
---,,---')>--VV'v-....--"V\~-__+---~
I.F
H-P
riSCH-IDOl
(lN626.J)
fig. 2. Signal-processor circuit as modified by N6TX for use
with his microwave spectrum analvzer."
performance
This shaper circuit enabled me to easily view 40 dB
dynamic range
(+
10 to - 40 dBm), with an unusual
response which is very nearly logarithmic at 10
dB/cm at very low (-20 to -30 dBm) and very high
(- 10 to
+
10 dBm) signal levels. Intermediate ampli-
display was calibrated with the aid of a stable 10 mW
signal source and a calibrated step-attenuator, byob-
serving changes in the display amplitude as various
amounts of attenuation were switched in. Since
every analyzer is likely to exhibit its own transfer
characteristics, it's a good idea to perform a similar
calibration yourself if you duplicate this project.
One further point: When I change from low-band
(de to 2 GHz) to high-band (500 MHz to 2.5 GHz)
coverage, the vertical scale calibration changes
con-
siderably.
This is due to the difference in i-f gain with
the i-f amplifiers operating at 2 and 1.5 GHz, respec-
tively. Once the analyzer is recalibrated, however, I
find it possible to easily resolve signal amplitudes
over at least a 40 dB range, with the analyzer operat-
ing in either band.
Any feedback from readers who attempt to apply
this or other signal-processor circuits would be great-
ly appreciated. All correspondence which includes a
stamped, self-addressed envelope will be answered.
Spectrum display of a 450-MHz signal source, as viewed on
the microwave spectrum analyzer with logerithmic video
processing. The desired signal is at
+
10 dBm: second har-
monic is down 23 dB at
-13
dBm. Fourth harmonic is clearly
visible at 40 dB down (- 30 dBm/. Also visible is a third har-
monic component at approximately - 35 dBm. Total display
dynamic range easily excaeds 40 dB. Note the non-uniform
vertical deflection graduations, discussed in the text.
dBm
+10
o
-10
-15
references
1. H. Paul Shuch, WA6UAM, "Low-Cost Microwave Spectrum Analyzer,"
ham radio,
August,
19n.
page 54.
2. Jeff Walker. W3JW, "High-Resolution Spectrum Analyzer for Single
Sideband."
ham radio,
July,
19n,
page 24.
-20
-30
8
1
1.2
1.4
1.6
1.8
ham radio
FREQUENCY
GHZ
july 1978
fIJI
35
[ Pobierz całość w formacie PDF ]