Elecraft Responds to the ARS Lab's Review of the K-1
special to The ARS Sojourner
design high performance receivers. He voices some strong general opinions
regarding receiver front end designs with NE602 Gilbert mixers, audio derived
AGC etc. We would like to post some comments on the review below.
As usual, Russ did an excellent job of doing the lab tests. His data is well
presented and we really like the comparison charts. Russ accurately states that:
"The K1 is a NorCal 40 / Sierra / SST on steroids." This is exactly what we
intended and our RX specs bear this out, placing the K1 at the top of this
group. What we would like to point out is that there are really two legitimate
categories of RX performance:
1) The 'crunch-less' front end used by higher end, more complicated and
expensive rigs (K2, FT-1000 etc.) Here we agree you should 'go for broke' to get
the best performance from the RX.
2) The 'reasonable' RX performance offered by rigs like the K1, DSW-20, Sierra,
NorCal 40, SST, IC706, FT-100 etc. Rigs in this class are constantly trading
off performance against cost, size, power consumption, ease of construction and
repair by the owner.
Russ appears to have a strong bias against -any- RX based upon a '602' Gilbert
Cell type mixer. (We do too, as you can see in our K2 design - but only for
higher end designs - not those intended for simple construction and portable
use..) We feel Russ misses the point in the case of the K1 and rigs like it. We
DID design it for low current drain portable use and that basically limited us
to a low power mixer like the '602. That's how we achieve its 55 ma RX drain. We
optimized the RX chain so it performs significantly better than other '602
designs and we also made the front end band pass filters extremely narrow to
reject out of band signals. Our K1 users in Europe (with its harsh RF
environment) have been very happy with its performance, even on home antennas.
Its not meant to be a K2, but it performs at the top of its class of simple,
portable, low cost rigs.
The '602' also allows us as designers to significantly reduce the component
count in the rig.
What is really interesting about the review is that the K1 RX performs exactly
where we expected it to: More or less at the top end of the 'Gilbert Cell' mixer
rigs, and surprisingly better in several RX specs than a couple of the 'diode
ring' mixer rigs like the QRP+ and NC20. (Readings within 3-5 dbm on the charts
are essentially equal since the measurements can vary by that much from rig to
rig and lab session to lab session.)
A third order iP in the -5 dBm to 0 dBm range is not bad for this class of rig.
The ARRL numbers for the IC706 are -1.3 to -11 dBm and the TS-50 comes in
between +3 and -7 dBm. Certainly the higher end RX rigs like the K2, FT-1000
etc. have much better numbers, but this comes at additional expense, size,
complexity and current drain.
The negative comments on the variable crystal filter are odd since there is
no supporting detail except for the raw data on the filter. The filter graph is
where we expected it to be and many have noted it works quite well. (Yes, the
wider response setting has a broad roll off, but this width is only intended for
use while tuning the band. We always use the narrower K1 filter settings for the
final QSO.) Any of the many NC40 / Sierra / SST / DSW20 users out there will be
more than happy using the K1.
Russ also made a negative comment about audio derived AGC in general, which is
used in all of the rigs in this class (NC40, Sierra, SST etc., and even by the
the NC20, QRP+ and SG-2020. (The DSW-20, 40 etc. and the SW40 series have no
AGC. They rely on manual RF gain and diode clipping limiting in their AF
stages.) We worked hard optimizing the K1's audio AGC performance to be the best
in this category. IF derived AGC will always have a better attack time, but it
would have added significantly to the size, complexity and cost the the K1.
The bottom line is that we designed the K1 with the following goals in mind:
1. Very low RX current drain for battery / portable operation
2. Very easy to build, understand and debug - even for first time builders.
3. Reasonable cost - well below the K2 and other full feature rigs.
4. Good RX operation in over 95% of operating situations.
5. Excellent TX characteristics (excellent harmonics and a great keying
waveform.)
6. Unmatched features in a single, very small, package: Built in memory keyer,
digital display and options for a noise blanker and internal auto tuner. You can
basically throw the K1 into your briefcase, knapsack or lunch box, along with a
battery and spool of wire, and be on the air where ever you choose to go!
All reviews are subjective. What is excellent about Russ's ARS review
is that while he voices his opinion about RX designs strongly, he follows his
subjective opinions with lab data. While we wish he had made it clearer he
wasn't singling out the K1 RX performance versus other rigs in its category, I
think most readers, after reviewing the data and comparing the K1 to the other
rigs in its class, will not be disappointed at all. So far our K1 customer
feedback has been exceedingly positive.
A final note: Wayne and I are obsessive optimizers. If there is a way to squeeze
more RX performance out of rigs designed for this class (small, portable and
easy to build and understand), you can be sure we will find it!
Eric Swartz, WA6HHQ, and Wayne Burdick, N6KR
www.elecraft.com