"Lucky" ) writes:
Hi Pete

How can that be? You're saying the circuitry was designed that way? They
couldn't do double for all the bands??

Lucky

I'm curious about how it was done, but it was not uncommon to have
receivers in the old days that switched to double conversion only
on the highest band(s).

Back then, the normal IF was 455KHz, this was before crystal filters
in the HF range became available (and once they did, before they became
common). It was pretty easy to have good image rejection at the lower
bands with that IF frequency, and given that one did have front
end tuning tracking the local oscillator, there was little incentive
to do things differently.

But at the higher bands, 10MHz and up (the exact frequency being
more determined by how the receiver breaks the 3 to 30MHz range into
bands), image rejection became more of a problem. On the really
cheap receivers, it pretty much disappeared, as you'd often see
in reviews of such receivers. A receiver like the National HRO series
had two stages of amplification before the mixer, and the extra tuned
circuits apparently helped to keep out images.

So double conversion was added on the highest band(s). A stage of
a combined mixer/oscillator was added one those bands, or more
commonly there was a stage of 455KHz amplification that was
turned into a mixer/oscillator on the highest band(s). The first
IF would then be in the 2MHz or so range, and that extra stage
would drop it down to 455KHz. This counted on having enough selectivity
at the first IF so the 455KHz images were knocked down enough.
The Hammarlund SP-600 did this (and benefitted especially since it
tuned right up to 54MHz), and a Heathkit portable comes to mind too.

Why not make it double conversion through all the tuning ranges?
Maybe because the extra conversion always adds problems, so if
you only have it in place when actually needed, you cut down
on unwanted spurs. Double conversion gives two chances for
images, and a mixer is more prone to overload than an amplifier,
so if you don't need the extra conversion for image rejection,
the extra conversion is a liability.

The other common scheme for double conversion back then was to
have a receiver that tuned a fixed range, and then toss in a
crystal controlled converter ahead of it to add the extra bands.
In some cases, you had single conversion on one band, because
the fixed tuning range would be a useable part of the spectrum.
A classic example would be a receiver that tunes the 80meter
ham band, 3.5 to 4MHz. For that band it would be single conversion.
But to tune the higher bands it would be double conversion, with
the bands converted first to that 3.5 to 4MHz range. Receivers like
this had the advantage that since the tuning was for only one range,
it was worth making a good linear oscillator and a good dial. Which
is where receivers like the Collins R388 came from.

Double conversion became a different thing once crystal filters in
the HF and even VHF range came along. Up conversion was then happening
and that has certain advantages. Once the first IF has a decent
filter, you will get good image rejection for the second IF, and
since you have a relatively narrow bandwidth at that first IF,
overload of the second mixer becomes more difficult, since it's
only seeing a small portion of the spectrum. And then LC oscillators
became synthesized to a crystal reference, and it no longer mattered
where the oscillator tuned.

Michael