[Elecraft] Transmitter performance

[Ian White]

To help make this discussion clearer, there is no problem about
referencing the original article [1]. The author, SM5BSZ, has published
it on his website for everyone to see, because this topic of wideband
noise from transmitters needs to be much more widely known.

<http://www.sm5bsz.com/dynrange/dubus313.pdf>

The wideband noise in question is composite noise (meaning both
amplitude noise and phase noise) at wide frequency separations from the
main signal. In simple terms, phase noise comes mostly from the
synthesizer but amplitude noise can come from both the synthesizer and
other sources as well. It's those "other sources" of amplitude noise
that can prove to be important. In many modern transceivers the TX
signal starts out at a very low level, and the low-level TX amplifier
stages need to be designed like a receiver LNA to avoid adding unwanted
noise. Careless designers forget this... and some make even worse
mistakes.

Elecraft owners can relax. The K3 comes top of SM5BSZ's TX noise table
with a measured level of -126dBc (in ARRL's normal reference bandwidth
of 500Hz). This is outstandingly good, 15-25dB better than the chasing
bunch which includes the KX3. The difference between the K3 and the KX3
reflects the completely different hardware architecture of the two
radios, but there is no shame in the KX3 coming 6th because its
transmitter noise performance is still good enough for almost all
practical purposes.  

Dave's concern is that high-level TX noise sidebands can cause problems
to other band users. HF operators never notice this problem because TX
noise is buried beneath other band noise; but it can be a very real
problem on VHF and microwaves where the background noise levels are
extremely low but signals can sometimes be much stronger than on HF.
This places extreme demands on dynamic range - not only in the receiver
(which we hear a hear a lot about) but also in the transmitter.

The KX3 qualifies as 'good enough' because its transmitter noise
sidebands are unlikely to affect anyone else's noise floor unless the
main signal is stronger than S9+40dB... and the K3 is another 20dB
better than that.

The people who really should be worried are the owners of "Brand F", two
models of which measured about 25dB worse than the KX3, and one model
was almost 50dB worse than the K3. (In case that number slid by you,
50dB is also the difference between 10mW and 1kW!) The amplitude noise
sidebands of "model F" at +/- 350kHz are so bad that they can clearly be
seen with an old, noisy spectrum analyser... and that means VERY bad
indeed. SM5BSZ particularly criticizes the ARRL reviewer's statement
that "Overall, the composite noise output as shown in Figure 3 is low"
when that number was way higher (30-45dB) than ARRL's own measurements
of other transceivers. 

Bottom line: Elecraft users can go back to sleep :-)


[1] For full disclosure, I am a volunteer editor for DUBUS magazine
(www.dubus.org). Copyright of articles published in DUBUS remains with
the original authors, so they are free to publish in other media as
well.  


73 from Ian GM3SEK


>-----Original Message-----
>From: elecraft-bounces at mailman.qth.net [mailto:elecraft-
>bounces at mailman.qth.net] On Behalf Of Dave
>Sent: 21 December 2013 00:48
>To: KX3 at yahoogroups.com; Elecraft Reflector
>Subject: [Elecraft] Transmitter performance
>
>I have recently received a copy of a publication known as DUBUS (for
those
>outside of Europe, this is a technical publication dedicated to VHF,
UHF and
>SHF equipment design and operating) and one particular article caught
my
>eye.  The reason being that I'd suffered from some wide bandwidth
signals
>from a semi-local beacon during a band opening on 23cm where the
>transmitted
>signal from the beacon could be detected over a large frequency spread.
>The
>issue, in this case, was apparently down to the power supply used by
the
>beacon.
>
>Now, in the article in DUBUS 3/2013 (Vol 42) a description of potential
>problems for various Amateur radios from the quality of their
transmissions
>was discussed and a table of 'good', 'reasonable' and 'poor' results
from
>tests taken from ARRL and Sherwood Engineering data (which I know
>people
>from Elecraft are often more than happy to quote, so I would guess it
passes
>some engineering muster...) were presented showing not only the
receiver
>performance, but the transmitter performance as well.
>
>Without wanting to quote the whole table (there may be copyright issues
of
>course) but the K3 came top (distant cheering heard from Aptos, CA) but
the
>KX3 came 6th.  The criteria for the ranking being a "TX Wide"
measurement
>in -dBc and the K3 scored 128 at 100kHZz  or 120 at 360kHz - which only
>seemed to
>be included as two radios from a rival company were both ranked at the
>bottom of the table and could muster no better than a measurement in
300
>and
>350kHz bandwidth) but the KX3 scored 107 at 100kHz.
>
>The reason that this is important is that, in an area where people are
>tightly packed together, someone using a transmitter that generates
>significant wide band noise within a particular Amateur band is not the
sort
>of neighbour that you want to have next to you on the VHF/UHF/SHF
bands.
>You might be listening on, say 144.050MHz but if they are transmitting
on,
>say, 144.350MHz then their wideband noise will affect you and, as is
stated
>in the article, they may not even to be transmitting to generate
significant
>signal levels.
>
>So, to my question.  As I'm interested in VHF/UHF/SHF more than HF, why
is
>the KX3 transmitter performance worse than the K3 and, other than
selling
>the KX3 (which I'd prefer not to do) and buying a K3 (which is
attractive,
>but financially a bit of a problem) what, if anything, can I do to
improve
>the 'TX Wide' performance of my KX3?
>
>Dave (G0DJA)