The DX Prowess of HF Receivers

Document Sample
The DX Prowess of HF Receivers Powered By Docstoc
					                             The DX Prowess of
                               HF Receivers
                     Are you looking for a good receiver for DX hunting?
                      Here are some distilled performance numbers that
                            might point you in the right direction.

                                              By Tadeusz Raczek, SP7HT

        omparing the performance of           their products be evaluated at 50 kHz    results for some HF receivers with
        one receiver to another is quite      or even 100 kHz signal spacings, where   closely spaced signals. The ARRL
        a difficult task. Receiver-           much more optimistic results can be      labor-atory and G3SJX used 20 kHz
performance tests are described in            achieved. Table 1 illustrates IC-765     for wide-spaced signals and 5 kHz for
detail in The ARRL Handbook, Chapter          receiver front-end dynamic-range         narrow-spaced signals. W8JI has used
26. Shortwave DX hunters and contest          measurements performed by the ARRL       10 kHz for wide-spaced signals and 2
participants have requested that              laboratory at various signal spacings.   kHz for narrow-spaced signals.
testing of receiver front ends be made           We can see a big difference between
at conditions representing real on-the-       5- and 50-kHz test results; that is,     BDR and Two-Tone Third-Order
air situations. That is, we should test       blocking dynamic range (BDR) and         Dynamic-Range Tests
receivers when extremely weak DX              intermodulation-dynamic-range (IMD           BDR is the difference, in decibels,
signals from the other end of the world       DR) measurements for widely spaced       between the minimum discernable
are present at the same time as several       signals produce much better results      signal (MDS) and an off-channel signal
strong local signals that are close to that   than for 5-kHz spacing. That explains    that causes 1dB of gain compression in
tiny DX signal. In contrary to standards      why manufacturers are opting for         the receiver. Two-tone third-order
set by Amateur Radio community,               wide-spaced measurements.                dynamic range (IMD DR) is the
equipment manufacturers prefer that              Closely spaced tests can inform us    difference between MDS and the levels
                                              much more realistically about a          of two interfering signals causing IMD
Skrytka Pocztowa 738                          receiver’s usefulness for DXing and      products just equal to the MDS.
25-324 Kielce 25                              contesting on our crowded HF Bands.          IMD DR depicts the strong-signal
POLAND                                        The ARRL laboratory, G3SJX and           capabilities of a receiver; that is, how it                                   W8JI have published measurement          behaves under real-world conditions,

                                                                                                           Sept/Oct 2002 1
when strong signals are delivered from          Table 3 demonstrates 10 and            not as good as we want them to be.
the antenna to the receiver input.          2-kHz-spacing test results of BDR and         Considering the decrease in BDR
Receiver IMD immunity is determined         IMD DR for some HF transceivers            and the lowering of IMD DR between
by the limits of its linear signal-         tested by W8JI. Table 3 includes the       widely and closely spaced tests, I
handling capabilities. Those, in turn,      same additional columns as in Table        consider the best receivers for split-
are determined by the limiting effects      2. In column four, the decrease in BDR     frequency operation with DX stations
of receiver active circuitry such as the    is calculated between the 10 and           to be those of the following HF
preamplifier, mixer and first IF            2-kHz tests. Consequently, in column       transceivers:
amplifier. Passive components may also      6, the decrease in IMD DR is cal-          • Elecraft Model K2
exhibit such limiting effects. For          culated between 10 and 2-kHz tests.        • Ten-Tec Model Omni-VI+
instance, fast RF silicon diodes used for   Italic numerals distinguish 2-kHz test     • Heavily modified Drake R-4C
receiver input-filter selection and for     results.                                      The three best results in Table 2 and
receive-transmit switching or pream-            The 2 and 5-kHz closely spaced         one result in Table 3 are distinguished
plifier/attenuator activation often cause   receiver tests represent real-world, on-   by boldface lettering.
additional IMD in some present-day HF       the-air DX hunting (split operation),
transceiver models. Moreover, overload      when many strong signals are very          K2 and OMNI-VI+: Design
of varactor diodes in automatically         close to a very weak DX station signal     Concepts and Features
tuned preselectors, as well as submin-      barely copied in the noise. Less              Elecraft and Ten-Tec manufacture
iature, inexpensive inductors and           degradation of BDR and IMD DR              the K2 and OMNI-VI+, respectively
monolithic two-pole first-IF filters        values means better receiver               [The Omni-VI+ has been discontinued
placed immediately after the first up-      performance for strong closely spaced      as of 2001 in favor of a superior design–
conversion mixer, can have a role in        signals. You can see that some             Ed]. Drake discontinued the manu-
IMD generation and receiver perfor-         receivers perform better and some are      facture of the R-4C about 20 years ago.
mance degradation.
    Table 2 demonstrates 20 and 5-kHz-
spacing test results of BDR and IMD
DR for some HF receivers tested in the
                                            Table 1—IC-765 Receiver Front-End Dynamic-Range Measurements
ARRL laboratory. Two columns are
added for convenience in analysis. In       Signal Spacing Blocking DR (dB)                        IMD DR (dB)
the fourth column, the decrease in BDR      (kHz)         IF Shift Off IF Shift On            IF Shift OffIF Shift On
is calculated between the 20 and 5-kHz      5                120           91                      85          73
tests. In the sixth column, the decrease    10               130.5        105                      90          88
in IMD DR is calculated between 20 and
                                            20               151.5        139.5                    97          95
5-kHz tests. Italic numerals distinguish
5-kHz spacing test results.                 50               152          152                      99          99

Table 2—20 and 5-kHz-Spacing BDR and IMD DR for some HF receivers tested by the ARRL
Manufacturer Model                  BDR (dB)      BDR Decrease              IMD DR (dB)      IMD DR Decrease
Elecraft          K2              133 and 126       only 7 dB                 97 and 88          only 9 dB
ICOM        IC-706MkIIG           120nl and 86       34 dB!                   86 and 74           12 dB
ICOM            IC-746             113 and 88        25 dB!                   92 and 78           14 dB
ICOM         IC-756PRO            120 and 104        16 dB                    88 and 80         only 8 dB
ICOM         IC-775DSP            132 and 104        28 dB!                  103 and 77           26 dB!
Kenwood     TS-570S(G)             119 and 87        32 dB!                  97nl and 72          25 dB!
Kenwood       TS-570D                   “               “                         “                  “
Kenwood        TS-2000            121nl and 99       22 dB!                   92 and 67           25 dB!
Ten-Tec       OMNI-VI            128nl and 119      only 9 dB                100 and 86           14 dB
Ten-Tec       OMNI-VI+                  “               “                         “                  “
Yaesu          FT-847             109nl and 82       27 dB!                   89 and 73           16 dB
Yaesu Mark-V FT-1000MP            126 and 106        20 dB!                   98 and 78           20 dB!

Table 3—10 and 2-kHz-Spacing BDR and IMD DR for some HF transceivers tested by W8JI
Manufacturer Model                BDR (dB)      BDR Decrease             IMD DR (dB) IMD DR Decrease
ICOM          IC-751A            98 and 83.5        14.5 dB               91 and 79         12 dB
Drake      R-4C (stock 1)*       109 and 57          52 dB!               82 and 48         34 dB!
Drake      R-4C (stock 2)†       116 and 80         36 dB!                86 and 68         18 dB
Drake R-4C (heavy mod)†† 131 and 127               only 4 dB             119 and 118       only 1 dB
*Stock 1 has MOSFET second mixer.
†Stock 2 has vacuum-tube second mixer.
††Heavy mod is rebuilt with solid-state doubly balanced high-level mixers and Sherwood 600-Hz roofing filter.

2 Sept/Oct 2002
For CW-oriented DX hunters, the R-4C           of the ham bands and prevent                  The K2 itself is devoted to CW QRP
is not an impressive receiver when             receiver front-end overload and IMD       enthusiasts, but could be tailored for
compared to recent models. But after            In designing its K2, the main goal       other preferences by adding following
radical modifications, an upgraded           of Elecraft was to construct an HF          options:
R-4C is a good receiver for weak DX          transceiver devoted only to the ham             • The SSB option offers an
signal CW reception on crowded               bands, useful for DX hunting—mainly         adjustable speech compressor and
amateur HF bands, thanks to the low          CW—with SSB as an option. As Table          optimized seven-pole, 2.2-kHz-wide IF
phase noise of the R-4C PTO (perme-          2 indicates, this has been done             crystal filter,
ability tuned oscillator). As shown in the   successfully.                               • 100 W PA Module (offered since the
ON4UN questionnaire results in the              The K2 HF transceiver implements            Dayton 2002 convention)
second edition of The Antennas and           a single-conversion superhet receiver:      • 160-meter band with second receive
Techniques for Low-Band DXing, a             • A doubly balanced diode mixer offers         antenna
significant number of responders have          excellent dynamic range. Narrow           • An automatic antenna tuner
reported using the R-4C for DXing on           and ham-band-only double-tuned            • A noise blanker
80 and 160 meters.                             preselector filters are switched by       • An auxiliary I/O RS-232 interface
    The K2 and OMNI-VI+ BDR for                relays, so the receiver front end         • An audio filter, eliminating residual
5-kHz spacing between strong signals           offers much better IMD response              noises outside the desired passband
is (126 – 106) 20 dB and (119 – 106) 13        than when diode switching is used             The K2 is sold in kit form with
dB, respectively, greater than that for      • A switchable HF preamplifier and          assembly instructions that are well
the third-ranked FT-1000MP Mark-V              switchable attenuator increase the        written. Anyone can complete the kit
(106 dB). Accordingly, the two-tone            range of receiver sensitivity adjust-     and buy what one really prefers. The
third-order dynamic range (IMD DR) of          ments, which allow the operator to        K2 Product Review, written by Larry
the K2 and OMNI-VI+ for 5-kHz                  adjust the receiver to particular         Wolfgang, WR1B, appears in QST
spacing from two strong signals is,            propagation conditions and the            (March 2000, pp 69-74). “Impressions
respectively, (88 – 80) 8 dB and (86 –         receiving antenna actually in use         of the Elecraft K2 Transceiver” by Rich
80) 6 dB better than for the third-          • AGC is derived from the IF signal.        Arland, K7SZ, appears in QST (April
ranked IC-756PRO (80 dB). This                 AGC offers fast attack time and           2001, p 99).
advantage is especially useful for DX-         smooth operation (without any                 In designing the OMNI VI+, Ten-
oriented operators.                            popping effect on strong signals) for     Tec has also departed from the
    Such good receiver front-end               fast and slow settings. It is even        prevailing general-coverage receiver
parameters prove the design concepts           possibile to switch the AGC off,          concept and returned to ideas used 20
implemented by Elecraft and Ten-Tec            which is sometimes the last chance        years ago. Ten-Tec have abandoned:
in the K2 and OMNI-VI+ models. Both            to copy extremely weak DX sur-            • Wide semi-octave, noisy first local
makers have abandoned ideas                    rounded by strong signals—                   oscillators generated by synthesizers
commonly exploited during last 20              experienced DXers know it.                • First-IF up-conversion into the 50 to
years by most other makers of HF             • A sharp IF crystal filter is close to        90-MHz region
transceivers and returned to proven            the mixer and because of the              • Wide first-IF roofing filters
designs used previously but with               relatively low IF (4.915MHz), the             The OMNI VI+ HF transceiver is
modern implementations.                        crystal filter greatly attenuates out-    designed for ham bands only, from 160
    The K2 and OMNI-VI+ use the                of-IF signals. That helps to prevent      to 10 meters. There are only two mixers
following crucial design ideas in the          receiver overloading by strong            in receiver chain: first IF = 9 MHz,
receiver front end:                            signals from outside the IF-filter        second IF = 6.3 MHz. All ham bands
• HF ham-band coverage only, no                pass-band. The IF crystal filter offers   are covered in 12 segments of 500 kHz,
   general coverage capability                 an adjustable passband for CW from        each having 30-kHz margins at lower
• Only single (K2) or double (OMNI-            wide (2000 Hz) to narrow (200 Hz).        and upper band edges. This model is a
   VI+) conversion is used instead of a      • A low-phase-noise PLL local oscillator    successful comeback of already proven
   chain of several mixers commonly             Implemented microprocessor               concepts but with an implementation
   used by other makers                      control offers:                             using present-day components:
• Both models have excluded the first        • Split operation with two VFOs             • The first local-oscillator signal is
   up-conversion IF into the 50 to           • Dual-range RIT and XIT                       produced with band-dependent
   90-MHz range with the associated          • Memory operation for mode (CW or             crystal oscillators mixed with a low-
   wide bandwidth first-IF roofing filter      SSB), dual VFO A/B split operation,          noise 4.97 to 5.53 MHz PLL.
   (with its passband set wide enough          receive IF crystal-filter passband           Therefore, all synthesis noise
   for narrow FM transmission and              selection, receive CW sideband               problems causing reciprocal mixing
   adequate for noise-blanker operation)       selection (allows canceling of one-          have been avoided.
• Both models use a relatively low first       side interference from strong nearby      • The first IF is low enough to
   IF that allows installation of narrow       station by switching to opposite             implement a narrow IF crystal filter
   SSB/CW crystal filters with good            received sideband—a rudimentary              with a good shape factor (having a
   shape factors to greatly attenuate          IF-shift function),                          passband adequate for SSB and
   out-of-band IF signals just at the        • Direct keypad entry of frequencies           CW) offering great attenuation of
   front of the IF amplifier                   and memory channels                          out-of-passband signals
• The main IF selectivity of the crystal     • Three tuning rates: 1, 10 and 100             The first IF at 9 MHz and can be
   filters is very close to the receiver       kHz per main-knob revolution              fitted with the following passband IF
   front end, which helps substantially      • 10-Hz tuning resolution                   crystal filters:
   to obtain high BDR and good IMD DR        • Adjustable receive CW offset with a       • SSB: 1.8 kHz or 2.4 kHz
   even for closely spaced strong signals      tracking sidetone                         • CW: 250 Hz or 500 Hz
• Both models implement ham-band-            • Auxiliary I/O RS-232 interface for        • A special 500-Hz, 6-pole IF crystal
   only preselector filters that substan-      computer logging and remote-                 filter centered for digital modes
   tially suppress strong signals outside      control purposes                              The second IF at 6.3 MHz can be

                                                                                                            Sept/Oct 2002 3
equipped with following pass-band              performance of the OMNI VI+ and K2.           We can find in pile-up situations that
crystal filters:                               The figure demonstrates a typical             many combinations of 2F1 – F2 and
• SSB: 1.8 kHz                                 situation where a barely heard DX             2F 2 – F 1 are present. Those will
• CW: 250 Hz or 500 Hz                         station—only a few decibels above the         produce intermodulation products on
   Such a mixing concept allows                receiver noise floor (dotted line)—is         the weak DX station’s frequency and
installation of narrow crystal filters in      operating SSB on 14.195 kHz. That DX          these IMD products will interfere with
both IF chains right at the beginning          station is operating split and listening      or distort the weak DX signal. They
of first and second-IF receiver ampli-         upward a few kilohertz. A pile-up of          can even completely bury the DX
fiers. Therefore the receiver main             strong stations is calling where he is        signal in noise and hiss. As the tables
selectivity filters are close to the mixers,   listening. For simplicity, only four          show, some receivers are more and
where they should be according to              signals are shown on the graph. Also          some are less prone to IMD.
DXers—and where they are not in most           for illustration, let us say that a QSO          Most present-day HF transceivers
ham radio HF transceivers made in the          is in progress just 3 kHz higher on the       implement synthesizers to produce LO
last 20 years.                                 neighboring frequency of 14.198 kHz.          signals for mixing. Analyzing BDR and
   Depending on chosen crystal-filter              The ability to copy such a weak DX        IMD DR results, you can judge for
combinations, the following good               station in presence of many nearby            yourself which makers do better and
shape factors should be achieved:              strong signals will depend on several         which ones are not as good—look for
• 1.3 for 2.4-kHz first and second-IF          receiver qualities: selectivity, BDR,         noise-limited remarks in test results.
  crystal filters for SSB reception            IMD DR and the amount of phase                Some synthesizer designs produce more
• 1.4 for 1.8-Hz first and second-IF           noise on the LO signal.                       phase noise than one can obtain using
  crystal filters for SSB reception                We can presume that almost any            methods implemented by Elecraft in
• 2.6 for 500-Hz first and second-IF           modern HF receiver has enough                 the K2 and by Ten-Tec in the OMNI VI+
  crystal filters for CW reception             sensitivity and selectivity to copy the       models. Therefore, K2 and OMNI VI+
• 2.9 for 250-Hz first and second-IF           weak DX station with no other signals         models are better predisposed to deal
  crystal filters for CW reception             present. Nevertheless, for real, on-the-      with pile-ups on crowded ham bands.
   Other combinations of first and             air situations when plenty of strong             The dotted line on Fig 1 indicates
second IF crystal filters are possible.        signals are present near DX-station           the receiver noise floor. The noise-floor
All installed IF crystal filters can be        frequencies, some receivers will do           levels of the OMNI VI+ and K2 do not
selected independently of the mode.            better than the others. That will             change in the presence of strong
Superior receiver selectivity signifi-         depend on how great is their BDR,             nearby signals, because the OMNI VI+
cantly decreases interference even             how great is their IMD DR and how             and K2 have much less phase noise
from very close signals.                       much phase noise accompanies the LO           than most HF receivers using fre-
   DSP noise reduction (5 to 15 dB),           for any particular HF transceiver.            quency synthesis. The dot-dash line
DSP auto-notch elimination of inter-               If the receiver has only average          illustrates the general situation for
fering carriers and DSP low-pass (five         BDR, even a single adjacent signal—           synthesized LOs. The presence of
choices) help to customize receiver            for instance, on 14.198 kHz, if it is         many strong signals near a weak DX-
selectivity in addition to the selectivity     strong enough—will desensitize that           station frequency leads to the
already offered by IF crystal filters.         receiver and the weak DX station will         appearance of reciprocal mixing
                                               not be heard in the presence of strong        signals on the DX frequency that will
Influence of Phase Noise                       interference.                                 interfere with that signal. When LO
   The main limiting factor of modern              When many strong stations are             phase noise and calling stations’
receiver performance is local-oscillator       calling, spread out 3-20 kHz up from          signals are high enough, then
phase noise. Phase noise contributes           weak DX signal, the IMD DR plays a            reciprocal-mixing products can bury a
to poor receiver BDR in the form of            big role in performance of the receiver.      DX station signal completely in noise.
desensitization by nearby strong
signals resulting from reciprocal
   In the OMNI VI+, phase noise is –
122 dBc for 1-kHz spacing, –123 dBc
for 10-kHz spacing and –138 dBc for
20-kHz spacing. In the K2, phase noise
is –120 dBc for 4-kHz spacing and –
126 dBc for 10-kHz spacing.
   Therefore, both OMNI VI+ and K2
have superb ham-band performance
with an extremely high close-in
dynamic selectivity. That enables
reception of very weak signals from DX
stations when strong signals are only a
few kilohertz away. Several on-the-air
A/B reception comparisons (using the
same switchable receive antenna) of HF
transceivers made by other makers              Fig 1—A representation of a typical DX pileup situation in the frequency domain. Vertical
against OMNI VI+ and K2 have been              lines represent the strengths of incoming signals. There is a weak DX station (shaded) at
made recently. Generally, these                14.195 MHz. The dotted line is the receiver noise floor for low-phase-noise receivers. It
comparisons favored the OMNI VI+ and           does not change in the presence of nearby strong signals and allows the tiny DX station
                                               (shaded) to be heard. The dot-dash line indicates the noise floor for a noisy synthesized
K2, especially in the case of CW               local oscillator, which has increased in the presence of nearby strong signals. The
reception on 160-meter band.                   increased noise floor hides the DX station at 14.195 MHz. The dashed line is the S9 signal
   Fig 1 explains the superior                 level.

4 Sept/Oct 2002
That case is illustrated by the shaded      600-Hz Sherwood roofing crystal filter        strong adjacent signals. In my opinion
bar around 14.195 kHz.                      in the first IF. You can replace the poor     Elecraft in the K2 and Ten-Tec in the
                                            second mixer with a high-level-input          OMNI VI+ have properly designed
Summary                                     doubly balanced low-noise mixer and           receiver front ends for DX-oriented
   The K2 by Elecraft and the OMNI          add more gain after the narrow IF             hams. This article was written late in
VI+ by Ten-Tec are relatively new.          filters following the second mixer (using     the autumn of 2001. Since then, Ten-
American makers have manufactured           a solid-state IF amplifier instead of a       Tec has announced their ORION
both. As far as I know (as of October       tube version). An R-4C upgraded that          Model new HF Transceiver. I believe
2001), there is no response to the call     way, with gain properly distributed in        this is a big step in the right direction.
for superior dynamic range from other       the receive chain, could offer better
makers of HF transceivers yet. DX           performance for extreme DX situations         References
hunters can optimistically expect that      than most modern HF transceivers.             I’ve drawn material from many sources to
                                                Perhaps I am an old-fashioned man.           produce this article. These include:
good times have come at last for them                                                     Many Product Review articles in QST. My
and other makers will offer their new       But my motto is: If equipment is                 articles published in SP HF Magazines
models designed appropriately for DX        designed properly to achieve best             J. Devoldere, ON4UN, Low Band DXing,
hunting and contesting. Nevertheless,       performance in some specific and                 second edition (Newington, Connecticut:
this is still a market economy and the      narrow area—in this case solely for              ARRL). The third edition is available from
next steps of other makers will depend      reception of weak CW and SSB DX                  ARRL as Order No. 7040, ISBN: 0-87259-
on how much popularity and admir-           signals only on crowded HF ham-                  704-0; $28.
ation the K2 and OMNI VI+ achieve           bands—you can expect better perfor-           Web sites: W8JI’s (, receiver
among the DX community.                     mance from it than from general-                 measurements as of 8 Aug 2001) and sev-
                                            coverage multiband machines.                     eral others: html;
   I’ve analyzed equipment-review                                                  ;
articles published in QST and some              Therefore, if an HF Transceiver is ;;
articles devoted to receiver front ends     used mainly for CW and SSB DXing                 Elecraft mailing list Elecraft@mailman.
published in QEX for some time now.         only inside the ham bands, a general-   (throughout the summer and
At the same time, I was gathering           coverage receiver with its associated            autumn of 2001).
components to build my own homemade         up-conversion and its first-IF wide              SP7HT has been involved in DX
dream receiver to perform better in         roofing filter is not the best way to reach   hunting for last 45 years. During the
extreme DX-hunting situations than          the main goal. Adversely to the concept       first 25 years of his activities all of his
equipment offered commercially on the       used in general-coverage receivers, the       HF equipment was homemade (inclu-
market—European QRM on low HF               main bandwidth selection should take          ding SSB crystal-filter production). In
bands is much, much stronger than in        place at a point as close to the front end    recent years, a homemade rig has been
other parts of the world. I’ve planned      as possible. That will enable us to           the only way to be on-the-air from this
to begin construction upon retirement.      achieve the greatest immunity against         part of the world. The last 20 years he
Recently, I’ve noticed that there are       strong adjacent signals.                      has used several ICOM and Kenwood
models on the market performing                 Unfortunately, that crucial demand        HF Transceivers, but he’s had no
almost as well as I need. Additionally,     is not acted upon in most of HF               experience with other makers.
Elecraft offers the K2 as a kit. Its many   transceivers offered in the ham-radio            Tadeus was the very first DXer from
options can be purchased and tailored       market at the present time. Being             Poland to reach the DXCC Honor Roll
according preferences, without the          myself a devoted DX hunter, I                 (1981) and DXCC Honor Roll #1 (1984).
unnecessary bells and whistles found        recognize the concepts implemented            For last 28 years his occupation has been
in general coverage multipurpose            by Elecraft in the K2 and Ten-Tec in          associated with microwave satellite
machines.                                   the OMNI VI+ as a step in the right           telecommunication. Actually he is a
   According to W8JI, there is also a       direction. To meet demands of DX              specialist at the Polish Telecom Satellite
challenge for ambitious constructors to     hunters, first of all, we need very good      Services Center “TP SAT” in Psary,
upgrade old R-4Cs having the narrow         receiver performance and immunity to          Poland. He will retire in January 2003.

                                                                                                               Sept/Oct 2002 5

Shared By: