RadioBanter - View Single Post

August 20th 04, 04:29 PM

(Bob Monaghan) wrote in message ...
I'll try to find the cite to the 73 article, it is in my articles files
IIRC ;-) Can't be that complicated*, but whether it will work well with
all older radio receiver designs? Should be available via interlibrary
loan, or I can mail you a photocopy if you can't find or get it locally?

[stuff deleted]
grins bobm

*Since my Drake R4B has IF shift, in a tube/transistor hybrid, it can't be
that complicated to have IF shift in your radios ;-) ;-)

But there's a world of difference between something built from the ground
up, and trying to modify an existing radio.

The basic concept is tuning the receiver so the unwanted signal is out
of the IF passband. Technically you could do it manually, and indeed the
concept existed before receivers had it as a feature. For an AM
signal, it is simply a matter of mistuning the receiver; people doing
this without thinking of the process. But once you "mistune" the
receiver so the unwanted signal is knocked out, an SSB or CW signal will
not be tuned properly. You have to adjust the BFO so it beats the wanted
signal down to it's proper frequency. Imagine tuning the receiver up
1KHz, what was a 1KHz beatnote before now becomes a 2KHz beatnote, so
you've got to set the BFO so it's again a 1KHz beatnote.

Go back enough decades, and the BFO was tuneable, so one could play around
with this. But once BFOs became crystal controlled, it wasn't possible and
even with a tuneable BFO the technique can be cumbersome.

I seem to recall that one of the early Collins SSB receivers had it,
and did it simply by having a control that tuned both the main tuning a
bit and the BFO frequency. But that requires a tuneable BFO and an analog
main tuning.

One, or some, of the Drake receivers actually tuned the filter. This
was possible because those receivers used an LC-tuned filter at 50KHz,
and so unlike a crystal filter, one could add a front-panel control
to tune that filter.

But most applications of IF shift require some kind of conversion stage,
either in the signal chain or outside it, in order to use the BFO as
both the BFO and a control to shift the IF passband.

The Drake R4-C (I couldn't easily dig up a block diagram for the "B")
has three IF frequencies, one at 5.645MHz (a somewhat wide crystal
fitler), one at 6.695MHz (a mode-specific crystal filter there)
and the third at 50KHz (an LC filter).

A B
-----5.645MHz---Mixer---5.695MHz---Mixer---50KHz---Product
filter | filter | detector
| | |
| fixed |
| osc |
| ---------------------------------|
variable
50KHz
oscillator

So as the oscillator signal into Mixer A is varied, it tunes the
rest of the receiver through the passband of that 5.645MHz filter.
But, since that oscillator also feeds the product detector, the
beat note remains constant.

One can get away without so many conversions in the signal chain,
but it still requires mixers. The Kenwood TS-820, for instance,
is single conversion. But the BFO signal is mixed with the local
oscillator signal (and filtered carefully) before going to the
mixer between the antenna and the IF stage. So again, as
the BFO is varied, the filter is moved but not the beat note.

There is no simple way to add IF shift to a receiver. At the very least
it requires bringing the IF out, and passing it through some stages
of mixers and a filter, and then feeding it back, into the receiver.
I've seen such articles, and it risks degrading the receiver, if you
don't use the right conversion frequencies or good mixers.

Or you modify the whole local oscillator chain, so it works like
the Kenwood above.

Neither is an easy step.

Like I said, my recollection of that 73 article was that it was
nonsense. I'm sure it simply added a varicap to an IF transformer
in the existing receiver, as if tuning one LC circuit would allow
for tuning the whole IF strip. Even if I'm misremembering, or
missed the explanation (as I said, the article did not give a good
indication that the author understood IF shift), merely making an
oscillator variable does not make for IF shift.

Michael