Product Review and Short Takes

Product Review and Short Takes from QST Magazine
April 2005
Product Reviews:
SDR-1000 Software-Defined HF/VHF Transceiver
Elecraft XG1 Receiver Test Oscillator
Short Takes:
KU4AB 2-Meter and 70-cm Antennas
Copyright © 2005 by the American Radio Relay League Inc. All rights reserved.
SDR-1000 Software-Defined HF/VHF
Reviewed by Steve Ford, WB8IMY
QST Editor
The debut of the FlexRadio SDR-1000
opens a new chapter in the history of
Amateur Radio. I’m not indulging in
hyperbole by making such a statement—
it is a fact. For the first time in ham history, you can purchase “off the shelf” an
HF and 6 meter transceiver that uses software to define its functionality—a software-defined radio.
What does this mean? Is the SDR-1000
just a software product?
Not necessarily. The SDR-1000 most
definitely has hardware. If you purchase
what I like to call the “full Monty” version of the SDR-1000 with the 100 W PEP
RF amplifier and RF expansion board, you
are presented with nondescript 10×81/ 2×4
inch black box. On the front of the box
there is an ON/OFF rocker switch, a fourpin microphone connector and a cooling
fan. On the back, you’ll find ports for
computer connections, dc power and, of
course, an antenna.
Inside the box there are a few circuit
boards, as shown in Figure 1, but they are
not populated in the manner you may be
accustomed to seeing in a “traditional”
transceiver. The receiver is an advanced
direct-conversion design using direct digital synthesis. It converts RF directly to
audio. Separate in-phase and quadrature
signals are then fed to the computer sound
card for digital signal processing using an
innovative in-phase and quadrature (I and
Q) image-reject approach. The connectors
are provided on the rear panel as shown
in Figure 2.
For transmitting, the SDR-1000 hardware is designed to take processed audio
from the sound card and convert it to RF.
In the case of the SDR-1000, used in this
review, the transmit conversion includes
boosting the RF to 100 W on 160 through
10 meters.
It’s quickly apparent that the true heart
of the SDR-1000 transceiver is not within
the black box. Data is the lifeblood of this
radio; the hardware is just a portal between
the analog and digital worlds. To invoke a
different metaphor, the SDR-1000 hardware is like unformed clay on the potter’s
wheel, waiting for the hands of the artist
to shape it into something meaningful.
The artist—and the artistry—is in the software that runs on your computer. The SDR1000 software—known as PowerSDR—
determines how a received signal will be
demodulated. It also creates the transmitted signal according to the mode you wish
to operate. Therefore, the SDR-1000 is a
software-defined radio in the most literal
sense of the term.
Don’t confuse the SDR-1000 with microprocessor controlled radios that offer
updateable firmware. The changes imple-
om Line
The SDR-1000 may mark the
beginning of a new generation of
Amateur Radio equipment, but the
pioneers who take it up may need
a bit of frontier spirit!
Figure 1—A look inside the SDR-1000 transceiver.
Joel R. Hallas, W1ZR
Assistant Technical Editor
From April 2005 QST © ARRL
mented by a firmware update are limited
in scope because the inflexible hardware
defines (and constrains) what can be done
with the radio. In the SDR-1000, the software is the radio to the greatest extent possible. When you modify the software in a
software-defined radio, you can make very
large changes indeed.
How Large?
How about adding a new operating
mode? Or totally revamping the digital
signal processing (DSP) functions? Or
designing a completely new “front panel”
monitor display? Depending on the nature
of the change, you could install new software and suddenly have a very different
And unlike firmware-based transceivers, the SDR-1000 software architecture is
completely open. This means that anyone
with enough computer savvy can modify
the software (and, hence, the radio) to suit
their individual needs. It also means that
hams throughout the world can pool their
collective genius and create new software
for the SDR-1000. So rather than a static
box full of hardware, the SDR-1000 will
evolve through the years as clever hams
take up the “clay” and create new works of
engineering art.
With the evolving nature of software
in mind, I should point out that this is a
review of the SDR-1000 as it existed in
late January 2005. By the time you read
this, it is certain that the software will have
changed with the addition of even more
features and improvements.
Also, it is important to note that our
review was conducted with Windowsbased PCs because, at the time of the review, the SDR-1000 software was only
available for Windows 2000 and XP systems. However, a Linux version is in the
works for later this year.
The Critical Sound Card and PC
The sound card is the engine that enables the SDR-1000. In particular, the
dynamic range and distortion performance
of the SDR-1000 is directly related to the
quality of the sound card. At the time of
this writing, FlexRadio Systems officially
supported only the SoundBlaster Audigy2
ZS, Audigy2, Extigy, MP3+ and the Turtle
Beach Santa Cruz sound cards. This is not
to say that other sound cards cannot be
used, but the radio may not perform as
specified. And if you run into trouble with
a non-supported card, FlexRadio may not
be able to help you.
In addition, not all sound cards have
the separate line input, line output and microphone jacks necessary to work with the
SDR-1000. They may also lack mixer controls with independent level adjustments
for the line input and microphone input.
From April 2005 QST © ARRL
Table 1
FlexRadio SDR-1000, amplifier serial number 0445-058
Manufacturer’s Specifications
Measured in ARRL Lab
Frequency coverage: Receive, 0.01-65 MHz;
Receive and transmit, as specified.
transmit, 1.8-2, 3.5-4, 5.33-5.4, 7-7.3,
10.1-10.15, 14-14.35, 18.068-18.168,
21-21.45, 24.89-24.99, 28-29.7, 50-54 MHz.1
Power requirement: Receive, 1.0 A max;
transmit, 25 A (max).
Receive, 0.9 A; transmit, 15 A.2
Tested at 13.8 V.
Modes of operation: SSB, CW, AM, FM.
As specified.
Receiver Dynamic Testing
CW sensitivity, 500 Hz bandwidth, 26 dB INA
setting: –141 dBm.3
Noise floor (MDS), 500 Hz filter:
Preamp off
Preamp on
3.5 MHz
–127 dBm
–134 dBm
14 MHz
–127 dBm
–134 dBm
AM sensitivity, 10 dB S/N, 30% modulation:
Not specified.
10 dB (S+N)/N, 1 kHz tone:
Preamp off
3.8 MHz
6.8 µV
Preamp on
0.68 µV
FM sensitivity, 12 dB SINAD: Not specified.
For 12 dB SINAD:
Preamp off
29 MHz
2.5 µV
Preamp on
0.66 µV
Blocking dynamic range: Not specified.
Blocking dynamic range, 500 Hz filter:
20 kHz
5 kHz
Preamp off/on Preamp off/on
3.5 MHz
93/90 dB
93/90 dB
14 MHz
93/90 dB
93/90 dB
Two-tone, third-order IMD dynamic range,
500 Hz filter, 90 dB:
3.5 MHz
14 MHz
Third-order intercept: Not specified.
3.5 MHz
14 MHz
20 kHz
Preamp off/on
87/82 dB
87/85 dB
5 kHz
Preamp off/on
86/82 dB
86/84 dB
20 kHz
Preamp off/on
–4/–5 dBm
0/–3 dBm
5 kHz
Preamp off/on
–4/–5 dBm
–1/–5 dBm
Second-order intercept: Not specified.
Preamp off/on, +69/+62 dBm.
FM adjacent channel rejection:
Not specified.
20 kHz channel spacing,
preamp on: 29 MHz, 37 dB.
FM two-tone, third-order IMD dynamic range:
Not specified.
20 kHz channel spacing,
preamp on: 29 MHz, 37 dB.*
S-meter sensitivity: Not specified.
S9 signal at 14.2 MHz: See Note 3.
These sound-card feature requirements
spell trouble for laptop users. Most laptops
offer only a microphone input and a headphone output. The good news is that it is
possible to use an external USB sound
device such as the SoundBlaster Extigy or
In addition to a quality sound card,
you’ll need a quality (read: “fast”) PC.
The minimum requirement for the SDR1000 is an 800 MHz Pentium computer.
In our tests, an 800 MHz system was just
adequate. Stepping up to a machine with
a clock speed greater than 1 GHz makes a
Figure 2—The rear panel of the SDR-1000.
Manufacturer’s Specifications
Measured in ARRL Lab
Squelch sensitivity: Not specified.
At threshold, preamp on:
SSB, 14 MHz, 5.0 µV; FM, 29 MHz, N/A.
Receiver audio output: Not specified.
See Note 4.
IF/audio response: Not specified.
Range at –6 dB points, (bandwidth):
CW (500 Hz filter): 460-935 Hz (475 Hz);
USB: 211-2816 Hz (2605 Hz);
LSB: 213-2778 Hz (2565 Hz);
AM: 2-5980 Hz (5978 Hz).
IF and image rejection: Not specified.
First IF rejection, 14 MHz, 116 dB;
image rejection, 14 MHz, 56 dB.
Transmitter Dynamic Testing
Power output: SSB, CW, FM, 100 W high,
low, not specified; AM, not specified.
CW, SSB, FM, typ 100 W high, <1 W low;
AM (carrier), typ 25 W high, <1 W low.
Spurious and harmonic suppression:
Not specified.
47 dB; meets FCC requirements.
SSB carrier suppression: Not specified.
70 dB.
Undesired sideband suppression: Not specified.
70 dB.
Third-order intermodulation distortion (IMD)
products: Not specified.
See Figure 5.
CW keyer speed range: Not specified.
0 to 54 WPM.5
CW keying characteristics: Not specified.
See Figure 6.5
Transmit-receive turn-around time
(PTT release to 50% audio output): <20 ms.
S9 signal, 25 ms.
Receive-transmit turnaround time (tx delay):
Not specified.
SSB, 150 ms; FM, 150 ms.
Unit is not suitable for digital modes.5
Composite transmitted noise: Not specified.
See Figure 7.
Size (height, width, depth): 4"×10"×8.5"; weight, 6 pounds.
*Measurement was noise-limited at the value indicated.
100 W operation on HF bands only. Pre-selector required below 1.8 MHz.
Current consumption higher on 160 meters (see text).
Dependent on software calibration. Once calibrated (see XG-1 review for possible calibrator),
the S-meter is relatively accurate.
Dependent on PC sound card (SDR speaker output requires amplified speakers).
Dependent on PC (see text).
substantial difference (it also allows you
to tweak the SDR-1000’s software CW
keyer to function above 15 WPM). For my
on-air tests, I used a 2.4 GHz system.
The SDR-1000 requires a connection to
the computer’s parallel (printer) port for all
the control lines going to and from the transceiver hardware. For most desktop PCs this
isn’t an issue, but you’ll need a 25 pin
straight-through male-to-male computer
cable, which isn’t a common animal these
days. Fortunately, FlexRadio sells the cable
if you can’t find it at your favorite dealer.
If your computer has two parallel
ports, you’re in luck. You can connect
your printer to one port and the SDR1000 to the other. My computer is limited to a single parallel port, so I found
that I had to unplug my printer and plug
in the SDR-1000 whenever I wanted to
get on the air. For a permanent installation, I’d need to install an A/B switch.
The SDR-1000 provides a port on the
rear panel for receive audio, but it isn’t
intended to drive a speaker directly. Instead, you must feed the audio to an amplifier, or a so-called “amplified speaker”
(a speaker with an amplifier built in). For
this review, I used an amplifier to drive
two large station speakers.
The SDR-1000 manual isn’t printed;
you must download it as a PDF file from
the FlexRadio Web site and print it yourself. At well over 100 pages, this can consume a considerable amount of paper and
printer ink (or toner). My solution was to
view the manual on my monitor and print
only the pages that I knew I’d need to refer to repeatedly during the setup.
Fortunately, the hardware installation
is straightforward. You connect the parallel cable between the radio and the
computer. Three audio cables connect between the rear of the SDR-1000 and the
LINE IN, LINE OUT and MIC IN ports on
your sound card. The leads from the
13.8 V dc power supply (not included)
connect to terminals on the SDR-1000 rear
panel. Finally, your antenna system
coaxial cable attaches to the SDR-1000’s
BNC output connector.
The True Challenge Begins when
you Install the Software
My first step was to download the latest version of PowerSDR from the
FlexRadio Web site. The file is less than
1 MB in size, so that step went quickly.
When I ran the setup program to install the
software, however, it came to an immediate halt and informed me that I didn’t have
the Microsoft .Net Framework installed on
my PC. Oops!
I jumped to the Microsoft Web site and
downloaded .Net. That’s a 23 MB file. With
my broadband connection I had the file
within a few minutes. A dial-up user may
need an hour or longer.
Okay, so .Net is installed and we’re ready
for action. The PowerSDR setup runs and
uses a “wizard” to lead you through each
step. My sound card is a Turtle Beach Santa
Cruz and it appeared on a drop-down menu
during the setup process.
Unfortunately, the setup wizard discovered that I didn’t have an ASIO (Audio
Stream Input Output) driver installed. An
ASIO driver offers lower latency and
higher bit depths. That’s critical to getting maximum performance from the
SDR-1000. However, many sound cards—
including my Santa Cruz—do not come
with native ASIO drivers. That revelation
entailed another trip to the Web to grab
ASIO4ALL, a piece of free software that
allows your sound card to achieve “nearASIO performance” (according to the
SDR-1000 manual). With the ASIO driver
installed, I assumed I was good to go.
I fired up the PowerSDR console software and clicked my mouse on the POWER
button. Relays clicked and the SDR-1000
came to life…sort of. I saw quivering lines
on the spectrum display, but heard nothing
coherent in the speaker. Now what?
I rechecked all the cables and the
PowerSDR console settings. Thirty minutes later I was about to unfurl the surrender flag when I spied the ASIO4ALL driver
icon on my Windows desktop. I double
clicked on the ASIO4ALL icon and found
a menu that asked me to choose between
my Turtle Beach sound card and my ATI
video capture card. Ah-hah! The ATI capture card had been automatically selected
by ASIO4ALL, not my Santa Cruz. I
changed the selection to the Santa Cruz
and instantly heard 40 meter SSB audio
in my speaker.
Now I had signals. Boy, did I have signals! The S-meter was pegged and the
quivering lines were bouncing off the top
of the spectrum window. Clearly, someFrom April 2005 QST © ARRL
Figure 4—One of nine PowerSDR setup screens. This one can
be used to tweak the performance of several DSP options, as
well as the AGC.
Figure 3—A histogram view of PSK31 activity on 20 meters.
thing was still not right. It was time to
calibrate the receiver.
There are three receiver calibration
steps: frequency, signal level and image
null. Frequency calibration was relatively
easy. I tuned in a strong WWV signal on
15 MHz and clicked the START button.
Seconds later the SDR-1000 was frequency calibrated.
According to the manual, the signal-level
calibration procedure requires an external
signal source such as the Elecraft XG1 for
best results. Not having an Elecraft XG1 at
hand, I had to use the alternative procedure:
Remove the antenna and set the calibration
level at –110 dB. I clicked START and within
20 seconds the PowerSDR software measured
the no-antenna noise level and calibrated
itself accordingly.
The final step would be the image null
calibration to compensate for phase and
amplitude imbalances on the I and Q signal lines, but that requires a high quality
signal generator or a separate transmitter
with a dummy load. Having neither, I was
forced to skip that step.
When I reconnected the antenna, I was
rewarded with signals at realistic levels.
The spectrum display looked great (see
Figure 3) and the S meter was responding
The SDR-1000 offers several modes,
with more to come as the software evolves.
The software version used for this review
(1.0.5) included SSB, CW, AM, FM and
DRM (Digital Radio Mondiale). The receive frequency range extends well below
160 meters, but FlexRadio recommends an
outboard preselector if you choose to
plumb those depths. There are also optional filter slots on the RF Expansion
board that will likely help in this regard.
Our SDR-1000 included the RF Expansion
board, but not the optional filters.
From April 2005 QST © ARRL
POWER and SQUELCH controls are all
mouse selectable. You can enter specific
values or click on the up/down arrows. The
same is true for the receiver and transmitter incremental tuning (RIT and XIT).
Tuning the SDR-1000 is accomplished
in one of several ways. I quickly discovered that my favorite method was to click
my mouse on the frequency display and use
the mouse wheel to move up or down the
bands. The more direct method is to simply type the frequency into the display and
press the ENTER key. You can also tune
within the spectrum analyzer display.
When you right click the mouse, a crosshair
appears. Move the crosshair cursor to the
signal of interest and left click to instantly
tune the SDR-1000 to the signal.
If the idea of tuning with a keyboard
or mouse makes you uncomfortable,
FlexRadio offers a Griffin Powermate
USB tuning knob. The USB knob works
quite well, giving more of a “traditional
radio” feel to the SDR-1000.
As you cruise the airwaves, you have
your choice of various DSP-enhancing
tools such as noise reduction, automatic
notch filtering (to remove those annoying
“tune up” signals) and a noise blanker. The
characteristics of all three can be varied in
the setup screen. For example, I found the
default setting of the noise blanker to be a
bit “weak,” especially when my neighbor
was using his power tools. It took just a
few mouse clicks to increase the blanker
performance to the point where the staccato pops ceased to be a problem. A typical setup screen in shown in Figure 4.
AGC is variable within several steps
and you have the luxury of choosing a
number of DSP filters, depending on the
application. For instance, I found that the
2.6 kHz filter offered the best fidelity
when listening to SSB, but when the going became rough, I’d switch to 2.1 kHz.
When it really got rough, I clicked on one
of the VAR (variable) filter buttons and
“designed” my own filter on the fly. You
simply drag the high and low frequency
cutoff sliders until you achieve the filtering you desire. Under difficult SSB conditions I set up a 1.3 kHz filter—drastic
and a bit odd sounding, but at least the
QRM was kept at bay.
With the selectable and adjustable filters, you can narrow the window to a mere
25 Hz for extremely selective CW reception. This was a remarkable thing to behold—and even more remarkable to hear.
Speaking of CW, don’t forget that two
receive audio channels (I and Q) are available at the sound card. This means that you
have the ability to select either one (referred to as U or L on the mode buttons), or
listen to both in the binaural mode. Binaural CW reception is a strange experience,
especially with headphones. The signals
seem to float in space, shifting to right,
center or left as you tune the radio. This
odd effect comes in handy when you’re trying to pick a signal out of a crowd (such as
a DX pileup or a contest).
AM reception was pleasant with excellent fidelity. The PowerSDR software also
provides a synchronous AM (SAM) detector, a feature that is only found on highend receivers. In the SAM mode, the
SDR-1000 replaces the varying carrier
from the transmitting station with its own
internally generated carrier. The result is
astonishing if you’ve never heard it before.
SAM is such a rare treat for me, I spent
most of an evening just tuning through the
shortwave broadcast bands. SAM even
makes music listenable.
Listening to DRM requires additional
software, which I did not have available.
Proprietary DRM software must be
downloaded from the DRM Software
Project following purchase of a license at
And what software radio would be
complete without a signal database?
PowerSDR offers a big database capable
of holding all your favorite frequencies
and modes, along with call signs and other
information. This is ideal for shortwave
listening, but it can be used for other applications as well. You can even choose
to scan through the database memories.
As enjoyable as it is to explore the
bands with the SDR-1000, there comes a
time when you feel the need to transmit.
The SDR-1000 manual states that you
must do an RF output calibration procedure using a dummy load. I didn’t have a
dummy load available at the time, so I
used the alternative approach—to accept
the PowerSDR default settings.
My first on-air experience with the
SDR-1000 was on 40 meter SSB. I used
a microphone/headset combo and a
footswitch to activate the push to talk
(PTT). The power metering doesn’t respond quickly to voice peaks and at first I
had the impression that I was only generating about 10 W output. My response was
to fiddle with the power amplifier settings
(adjustable for each band) in an effort to
extract more “juice.”
Suddenly, I noticed the telltale odor of
overheated components. If Scotty from
Star Trek were standing in the room, he
would have been screaming, “She can’t
take the strain, Captain. She’s gonna
blow!” For a moment, I experienced the
sinking feeling that comes with the realization that you may have just fried a set
of transceiver finals. Fortunately, the
SDR-1000 opted to forgive my blunder—
at least this time.
Connecting an external peak-reading
wattmeter, I discovered that the SDR-1000
was indeed peaking at 100 W even when
the meter seemed to indicate that I was
barely generating power at all.
With the output at a safe level, I proceeded to enjoy a number of contacts.
Everyone reported good audio, although
some commented that it seemed to lack
“low end.” I adjusted the PowerSDR equalizer and that helped considerably. I also
tested the speech compression feature, but
didn’t want to push my luck and overstress
the finals.
During the review period, I was lucky
enough to run into a DX operator who was
using split, which gave me the ideal opportunity to exercise the SDR-1000’s dual
VFO feature. Setting one VFO to the receive frequency and the other to the transmit frequency was as easy you’d expect
on a traditional transceiver—and it worked
just as well.
My CW experience with the SDR-1000
was disappointing. The problem appeared
when I tried to use my CW paddles. There
were timing issues that resulted in maddening delays in the sidetone audio. In
addition, when attempting to work semi
break-in, the first character seemed to be
clipped. In fact, the “dit” lengths seemed
to vary. These two issues combined to
make it nearly impossible to carry on a
CW conversation. The good news is that
FlexRadio is presently working on an external keyer module that will solve the
problem. Until it is available, CW is best
sent using an external keyer or the keyboard and memories (an anathema to purists, but a solution for contests).
One more surprise on the transmit side:
The SDR-1000 requires more current from
the 12 V supply on 160 meters than on the
other bands. It will put out the specified
100 W, but at a load that exceeds the 25 A
specification (and fuse size). The maximum
power output of our unit with a 25 A power
supply current was 82 W. The manufacturer
is investigating this condition.
FlexRadio is anticipating PowerSDR
updates in the near future that will add digital modes such as PSK31 and RTTY to the
console. Not content to wait, I crafted my
own solution using a laptop computer connected to the SDR-1000 microphone jack
through a West Mountain RIGblaster interface. Using this approach, I was able to
make a number of PSK31 contacts and
briefly participated in the January BARTG
RTTY contest. (It was a joy to set up a
250 Hz RTTY filter to corral the QRM on
20 meters. I had never experienced that
luxury before!) It was fascinating to switch
between the spectrum, waterfall and histogram displays and observe the signal characteristics of different digital modes.
If you have the capability to use two
sound cards simultaneously in your
PC, that would be another solution for
digital operating with the SDR-1000.
Either way, take care to limit the RF output to 25 W as the manual recommends.
After my experience on SSB, I was very
careful in that regard.
VHF on 6 and 2 Meters
Although reception seemed good on
6 meters, I didn’t gather much experience
on the transmit side. The output on
6 meters is limited to only 500 mW. I did
manage to make a local contact, but for
serious work you must invest in a power
amplifier. You’ll also need a method to
switch the SDR-1000 output cabling
between your 6 meter amplifier/antenna
system and your HF antenna(s).
For 2 meter operating, the SDR-1000
has a provision to add a Down East Microwave DEMI144-28FRS transverter within
the enclosure. This transverter covers 144
to 146 MHz, converting down to 28 to
30 MHz. With a transverter RF output of
about 100 mW, you will also need an amplifier to achieve the power necessary for
long-distance work. We did not test the
DEMI144-28FRS for this review.
The SDR-1000 is a work in progress,
and it always will be. This review is a snapshot of the transceiver captured at a particular point in time.
As I’ve mentioned before, the performance of the SDR-1000 is dependent on
the quality of the computer sound card,
particularly as it concerns dynamic range.
If you use a mediocre sound card, you’ll
get mediocre results. If you spring for an
expensive, high-end card such as those
made by M-Audio (,
you will see impressive results. The commercial cliché “your mileage may vary”
applies here.
In our measurements, as shown in
Table 1 and associated figures, the receive
performance of the SDR-1000 was at least
comparable to a traditional transceiver in
its $1300 price class. Of course, the crucial difference is that the SDR-1000 will
undergo continuous updating and improvement for years after the initial purchase. A hardware rig remains essentially
the same forever.
As it stands today, the SDR-1000 is a
decent SSB, FM, AM and digital transceiver. As a CW rig, it needs work, but
that is reported to be in system test now.
My wish list for other SDR-1000 improvements includes:
• PSK31, RTTY and other digital
modes completely integrated into the
PowerSDR console.
• A peak-reading power output meter
that responds in real time to rapidly fluctuating signals (such as SSB). FlexRadio
reports that their latest release includes
this feature.
• A better manual. The current manual
describes the controls and functions, but
it lacks descriptions of how to actually use
them to operate the radio in various
modes. It should follow the example of
traditional transceiver manuals in this regard. For instance, there should be separate sections that describe how to set up
the SDR-1000 to operate SSB, CW or any
other mode the radio supports.
• Calibration routines that don’t require external test equipment.
• A built-in speaker, or an external
speaker port that doesn’t require an amplifier.
Is the SDR-1000 a radio for all hams?
At this point, probably not. The current
incarnation of the SDR-1000 is best suited
to the amateur who knows his or her way
around a computer. It takes a ham with
From April 2005 QST © ARRL
Reference Level: 0 dB PEP
Reference Level: - 60 dBc/Hz
Vertical Scale: dBc/Hz
Frequency Offset (kHz)
Figure 5—Worst-case spectral display of
the SDR-1000 transmitter during two-tone
intermodulation distortion (IMD) testing
on HF. The worst-case HF third-order
product is approximately 29 dB below
PEP output, and the worst-case fifthorder is approximately 38 dB down. The
transmitter was being operated at 100 W
output at 28.35 MHz.
intermediate or advanced computer skills
to get the most out of an SDR-1000 with
the least amount of frustration.
But this is just the first step into a new
era. As the SDR-1000 evolves, new versions are likely to emerge that will be
“friendlier” and well within the understanding of any amateur. The manufacturer reports that they plan to offer a
turnkey sound-card and radio solution in-
Frequency Sweep: 2 to 22 kHz from Carrier
Figure 6—CW keying waveform for
the SDR-1000 showing the first two dits
in full break-in (QSK) mode using
external keying. Equivalent keying
speed is 15 WPM (limited by PC buffer
size). The upper trace is the actual key
closure (first closure starting at left edge
of figure); the lower trace is the RF
envelope. Horizontal divisions are 40 ms.
The transceiver was being operated at
100 W output on 14.02 MHz. A PC faster
than our 866 MHz unit will support faster
Figure 7—Worst-case tested HF spectral
display of the SDR-1000 transmitter
output during composite-noise testing
at 14.02 MHz. Power output is 100 W.
The carrier, off the left edge of the plot,
is not shown. This plot shows composite
transmitted noise 2 to 22 kHz from the
carrier. Note the spurs that are above the
level of the noise output—but within
regulations. Their position is frequency
cluding a CD with drivers and calibration
routines by the time you read this. If so,
that will be a real plus. With the collective intelligence of the global Amateur
Radio community at work, the potential
of the SDR-1000 is almost limitless.
Manufacturer: FlexRadio Systems,
8900 Marybank Dr, Austin, TX 78750;
tel 512-250-8595; e-mail; Model
SDR-ASM/TRA (fully assembled transceiver with 100 W amplifier and RF expansion board): $1325; SDR-ASM/TR
(fully assembled transceiver with 1 W
output and RF expansion board): $875;
SDR-PKG/TRA (partially assembled
transceiver with 100 W amplifier and RF
expansion board [enclosure is a kit]):
$799; SDR-1000/TR (1 W board set
only—no enclosure): $648.
Elecraft XG1 Receiver Test Oscillator
Michael Tracy, KC1SX
ARRL Test Engineer
S-meters—the majority of radios have
them, but like snowflakes, no two are exactly alike. Most amateurs realize this on
some level, but there isn’t much awareness of exactly how much variation there
is between the meters of two different radios, or even on the same radio when used
on different bands. Worse, there seems to
be a lot of complacency in the practice of
using these meters to provide the strength
indication of RST reports (and that is aside
from the infamous, “You are 59 here OM,
could you please repeat your name and
(–73 dBm for 50 Ω inputs) is equal to S9,
with every lower S unit decreasing the
level by 6 dB.
Virtually none of the S-meters found
on commercial amateur equipment follow
this “quasi-standard.” In fact, a lot of
knowledgeable folk call these devices
“guess-meters” instead. As explained in
great detail by the late Doug DeMaw,
W1FB, in 1977,1 S-meters are typically
controlled by a receiver’s AGC voltage,
which can vary from band to band and
with changes in preamp or attenuator settings. For a sampling of S9 levels from
QST Product Reviews, see Table 2. In addition to the variation at S9, there is quite
So What Should They Read?
While there has never been an official
standard for S-meters, Collins adopted an
internal standard that has become widely
accepted among the amateur community
(it was also adopted as an IARU Region 1
recommendation for HF use in 1990). According to this standard, a level of 50 µV
From April 2005 QST © ARRL
om Line
Priced considerably less than a
typical wattmeter, this is one station accessory that really adds
value to your shack.
a bit of variation in the level change per
S-unit. Some rigs are around 4 dB per increment, some are less (as low as 2 dB
per) and still others are more. Two samples
are shown in Table 3. What’s more, the
level change is typically not consistent on
the very same radio; the increase required
to go from S1 to S2 can be very different
from S4 to S5 and different again at S8 to
S9, as shown. Quite a mess!
What’s a Ham To Do—Elecraft to
the Rescue
So what’s an operator to do to make the
S-meter useful? Part of the problem has
been a lack of accurate level signal sources
at reasonable cost. Certainly a lab-grade
signal generator would do nicely, but its
cost can easily exceed that of the transceiver being measured! Enter the Elecraft
XG1. The XG1 is a crystal oscillator with
accurate output levels at 50 µV and 1 µV
at 7.040 MHz. The output on the
oscillator’s harmonics is not as precise, nor
as high, as would be expected, but none-
theless provides some useful functionality
there as well, as we’ll discuss.
When coupled with a step attenuator
with a 1 dB step size, the XG1 can provide a full characterization of the S-meter
levels below S9. For those who don’t have
an attenuator, several companies offer
commercial units, at least one company
offers a kit, and they are fairly easy to design2 and build.3,4,5
The XG1 is meant to be simple and
inexpensive. Everything is mounted on a
1.5×3.5 inch double-sided circuit board
and no enclosure is provided nor needed.
Stick-on mounting feet are included to
keep it up off the workbench, but if you
have a cluttered bench, you may want to
consider keeping it in a plastic container
when you aren’t using it to prevent accidentally shorting out the battery. Speaking of the battery, it is a fairly large, but
very common, button cell type. The XG1
has a low current drain, so it will last quite
a long time—80 operating hours according to the manual. I once accidentally left
it on for a whole day, and the output remained unchanged when I finally discovered my faux pas and tested it. The RF
connector is a BNC type, so an adapter is
required to connect it to the SO-239 UHF
connector found on most transceivers. As
small and light as it is, you can dispense
with coax jumpers and attach it directly
to the rig if there’s room.
The unit I tested was built by Joel
Hallas, W1ZR, QST’s Product Review
Editor. Joel reported that the assembly
instructions were clear, all parts were
there, the assembly was straightforward
Table 2
S-meter Signal Levels for S9 Reading, 20 Meters
Radio Model
ICOM IC-7800
Kenwood TS-480SAT
Ten-Tec Orion
Yaesu FT-857
Preamp Off
58 µV
87 µV
135 µV
17 µV
Preamp On
7.2 µV
18 µV
33 µV
6.6 µV
Table 3
Difference Between S-meter Readings on Two
Amateur Transceivers
S Value
Radio 1
–89 dBm
2 dB
–87 dBm
2 dB
–85 dBm
3 dB
–82 dBm
2 dB
–80 dBm
2 dB
–78 dBm
3 dB
–75 dBm
3 dB
–72 dBm
3 dB
–69 dBm
Radio 2
not taken
–114 dBm
11 dB
–103 dBm
9 dB
–94 dBm
5 dB
–89 dBm
5 dB
–84 dBm
5 dB
–79 dBm
4 dB
–75 dBm
4 dB
–71 dBm
Table 4
Elecraft XG1
Manufacturer’s Specifications
RF output level: 1 µV and 50 µV
Frequency: 7.040 MHz, ±100 Hz
Current consumption:
Size (W×D): 1.5" × 3.5."
Measured in the ARRL Lab
As specified (±1 dB).
As specified.
2.5 mA typical.
and it was completed in about an hour. It
even worked the first time!
A green LED lets you know the XG1
is on, a yellow LED warns of a low battery and a red LED warns of accidental
application of transmit power of 100 mW
or more. As explained in the manual, damage will occur at significantly higher levels, but don’t leave it connected to your
transceiver when you’re done testing, and
wonder why your SWR is so high the next
In addition to providing a “sanity
check” for your S-meter on 40 meters, the
XG1’s 1 µV output level can be used to
evaluate receiver sensitivity, so you can
make sure your rig is still up to snuff as it
gets older. The 1 µV level can also help
determine the rest of the S-meter scale,
since it should read 2 dB above S3, call it
S3.3, by the Collins standard. You can also
use it to check for problems after the inevitable visit by infamous Murphy or for
those times when you wonder whether the
bands are really that dead or you had a
lightning strike while you were sleeping.
It also makes an ideal carry-along for
checking out rigs at hamfests (provided
the rig has power, of course), or for helping check out rigs for friends. The unit is
remarkably accurate considering its price
and the person who assembled it! See
Table 4 for the details.
The output on the harmonic frequencies of 14.08, 21.12 and 28.16 MHz ranges
from 7 to 24 dB down from the fundamental, and these levels will vary somewhat—
the manual says ±3 dB, but when the
50 µV setting is used, the level of the harmonics is still useful for checking approximate S-meter response on these other
bands. Additionally, the fundamental
crystal can be changed to another HF frequency if desired (note that using a crystal socket is not recommended). Some
folks may find it helpful to have several
XG1s around the shack with crystals for
their preferred bands of operation. Output level accuracy is only guaranteed on
the original 40 meter frequency, however.
Manufacturer: Elecraft, PO Box 69,
Aptos, CA 95001-0069; tel 831-662-8345;
fax 831-662-0830;
Price: $39.
D. DeMaw, W1FB, “What Does My S-Meter
Tell Me?” QST , Jun 1977, pp 33-36.
H. Silver, NØAX, “Hands-On Radio, Experiment #13—Attenuators,” QST , Feb 2004,
pp 69-70.
P. Pagel, N1FB, B. Shriner, WAØUZO, “A
Step Attenuator You Can Build,” QST , Sep
1982, pp 11-13,
D. Bramwell, K7OWJ, “An RF Step Attenuator,” QST , Jun 1995, pp 33-35.
P. Ostapchuk, N9SFX, “A Rugged, Compact
Attenuator,” QST, May 1998, pp 41-44.
From April 2005 QST © ARRL
KU4AB 2-Meter and 70-cm Antennas
The square loop has been a popular VHF/UHF omnidirectional antenna for decades and it is easy to understand why.
The simple design packs solid horizontally polarized performance in a very small package. While it can’t compare to a
directional antenna, an omni loop is useful for antennarestricted environments, weak-signal mobile operating and even
amateur satellites.
Phil Brazzell, KU4AB, designs and builds a line of
“SQ loops” for 6 meters through 70 cm. Phil’s increasingly
popular loops are fashioned from solid aluminum rods for lightweight durability. Because the antennas are often installed outdoors, the rest of the hardware is crafted from stainless steel to
resist corrosion.
For this review, I purchased 2-meter and 70-cm loops for
my cramped attic antenna farm. The 2-meter SQ loop is only
about 12 inches across and the 70-cm loop is just under
41/2 inches wide (Figure 1).
SQ loops arrive completely preassembled. You simply lift
them from their shipping containers and they’re ready to install. The antennas also arrive pre-tuned for the “weak signal”
portions of the bands, which
is where most hams will use
them. However, you can adjust the tuning by increasing
or decreasing the distance
between the ends of the loop
Using a piece of PVC tubing as a mast, I placed the
2-meter loop about a foot above
the attic floor and the 70-cm
loop near the attic roof (see
Figure 2). With a diplexer in the
attic, I “split” a single feed line
to feed both antennas. The
Figure 1—The 70-cm SQ loop
antenna is less than 41/2 inches
loops sport standard SO-239
coaxial connectors.
The results of my SWR
measurements on 2 meters
and 70 cm are shown in Figures 3 and 4, respectively. As
you can see, the low-SWR
points fall right where
KU4AB intended. You’ll also
note that the SWR curve for
the 2-meter loop is a bit
sharper than the 70-cm
So How Well Do They
My first test was with the
AMSAT-OSCAR 51 satellite.
I was pleasantly surprised at
the strength and consistency
of the 70-cm downlink signal.
Figure 2—My attic SQ loop
Steve Ford, WB8IMY
From April 2005 QST © ARRL
Figure 3—The SWR curve with the 2-meter SQ loop.
Figure 4—The SWR curve with the 70-cm SQ loop.
At the highest point of a high-elevation pass, I enjoyed S7 signals, which resulted in full-quieting FM audio from the bird.
Another interesting test occurred during the January ARRL
VHF Sweepstakes. I couldn’t hear most of what the guys with
the big beams were working, but with my 50 W of RF, I was making contacts out to about 90 miles (and I am not on a hilltop).
And even with the polarity mismatch (and despite the high
SWR when operating in the repeater subbands), the SQ loops
were also usable for local FM work.
KU4AB’s SQ loop antennas perform as advertised, if not better. They are a snap to install and extremely rugged. The SQ loops
are also very reasonably priced. With those considerations in
mind, I suspect SQ loops will find homes at many stations.
Manufacturer: KU4AB Antennas, 339 Venice Cove,
Collierville, TN 38017; Model
SQ-144m1: $34.95; SQ-432: $28.95.
QST Editor
Download PDF