CCWN 75:55
DESIGN OF A HIGH QUALITY CW RECEIVER CHAS. WOODSON

          My experience with CW and CCW has led me to conclude
the most critical element in the communication system is the
receiver and that receivers which perform far better than
commercial equipment in my price range can be constructed with
readily available parts and techniques. The purpose of this
article is to describe my current thinking and the direction of
my experimentation toward building my own high quality CW
receiver. I have built the receiver in modular form and continue
to add, modify, and replace modules.            Figure 1 shows a
block diagram of the basic plan of the receiver. Each of the
blocks is a separate metal enclosure, most are Pamona boxes. The
input and output from all modules is low z, about 50 ohms, to
reduce possible radiation and pickup. Modules are connected by
shielded cables.            Separate RF-Mixer modules are
prepared for each band, no bandswitching is done and the elements
can therefore be optimized for a single band. Perhaps in the
future I might make a switching module to select among the
available modules. Considerations of dynamic range are even more
important when working with QRP signals and with QRO signals. One
of the modules uses a crystal filter at 14,049,000 for CCW
experimentation on that frequency.            There are several
alternative HFO units. One will be a oscillator with a range of 5
to 5.5 MHz to tune the 14 to 14.5 MHz range. Another will be a
VFO controlled by a synthesizer for high stability and
resetability. One is a crystal oscillator designed to have low
noise and high stability on a particular frequency; it may
eventually be locked to a standard by a PLL in the manner
suggested by Petit and Woodson (CCWN 75:44). Another is a
5,048,975 to 5,049,025 Hz oscillator. This unit, carefully
calibrated by comparison with the station standard, is used for a
VFO while receiving CCW and CW on 14,049,000 Hz. The one I have
is patterned after the VFO in Hayward (QST, 1973) and is still
undergoing evaluation but apparently is stable within a couple
cycles in normal operation. The emphasis so far in CCW has been
upon divide-by-n logic to obtain frequency precision and
resetability. Eventually we may be able to use other methods but
it will not be easy to get precision and resetability within a
cycle.            The IF strip is similar to the design of Wes
Hayward (QST, 1973). In fact, many of the ideas for the receiver
are adaptations or copies of his work. The IF strip uses two
matched KVG filters for input and output. I have in mind
eventually making separate IF strips for SSB and S! reception.    
       One possibility is to have a product detector made of a
matched quad of HP diodes. The BFO would be at 9,000,500 and use
the same high stability techniques used in the frequency standard
oscillators, again, locked to the standard if necessary.