[HBR] Relay switched coils?

[Walt Hutchens]

> Would using small relays instead of a bandswitch could provide
> sufficiently good connections and low strays to preserve the good
> performance of an HBR-type front end without plug-in coils?

How to handle the front end tuning is one of the fundamental -- and
often most difficult -- of the receiver design questions.

For narrow band front ends -- those that tune to the signal frequency,
rather than being broadbanded -- high Q tuned circuits greatly improve
performance. High Q coils must either be large (as are those in the
HBR designs) or they must be wound as toroids on modern high Q cores.

Typical small slug tuned coils have a hard time getting a Q over 50;
Q's for good air wound coils (a la HBR) or toroids, 200 or so. That's
a 4:1 difference in bandwidth that's not easy to get in any other way
and translates to a worthwhile difference in receiver performance,
particularly in intermodulation distortion (IMD).

In a multiband receiver (say 80-10M) front end lead lengths must be
minimized in order to avoid feedback problems that will seriously
compromise performance on the higher bands. Even if the RF stage
doesn't oscillate, feedback will force some combination of low gain
and poor noise figure on the bands where this matters -- say 20M and
higher.

There are several ways to tackle the lead-length issue.

1. Toroids can be small enough and have so little external field that
they can be packed closely enough to permit bandswitching in a home
design, although the need for very small trimmer caps may seriously
increase costs and precise coil and capacitor values are needed.

2. If more than one RF stage is needed, lower Q requirements make
small slug tuned coils practical. Military sets in which the need to
reduce LO radiation dictated two RF stages are an example. But ham
sets don't have that requirement and designing two stages for low IMD
is more than twice as hard as doing it for a single stage. Coil layout
to avoid stray resonance problems (the 80M coil might self-resonate on
10 or 15M, killing gain on that band) can be a challenge.

3. Complex 3-D designs such as turrets and placing the coils around
the bandswitch sections (perhaps on a plate supporting each section)
rather than on a conventional chassis can work, but the mechanical and
constructional difficulties are daunting.  (As in "it can be designed
but you can't wire it, and even if you could, you couldn't adjust the
coils.")

4. Use of plug in coils reduces the lead requirements to those of a
single coil per tuned circuit and eliminates the bandswitch. If large
coils are used, then relatively large trimmer capacitors can be
mounted in or on the coils, simplifying the capacitor requirements.

The major benefit of relay switching is that it lets you lay out a
complex switch without regard for mechanical issues. For example if
you are switching filters in an IF string that runs parallel to the
front panel, then you'll need a right angle gear drive to get the
shaft for a rotary switch to and through the front panel. With relays
you can place the knob anywhere you like without regard to where the
filter switching is done.

A secondary issue is that with relays you don't have to run a shaft
parallel to a string of high gain stages or between the input and
output of a filter: When very high isolation is required, signal
propagation along even a nominally grounded shaft can be an issue.

In other words, relay switching solves problems that one doesn't have
to have in the front end of a high performance multi-band receiver for
home construction. Plug in coils give the easiest design and
construction; a conventional bandswitch will work fine if you are
willing to use toroids (and rather fussy part values); and a modest
reduction in performance will let you use small slug tuned coils.

The genius of the HBR-series is that they could be built by any ham
with modest construction skills who could follow the directions AND
would deliver performance superior to almost any multiband receiver of
the time.  They're still very hard to beat.

Walt
KJ4KV