On Mon, 14 Nov 2005 04:07:57 GMT, "Frank Dresser"
wrote:


"matt weber" wrote in message
.. .
What breakthrough has made single
conversion so state of the art?


Absolutle nothing, in fact single conversion sucks unless it is an up
conversion, and even then, mixer noise will wipe out reception above
about 10Mhz absent a good tuned RF amplifier in front.


Why would up conversion mixer noise wipe out reception above 10 MHz? How
would the presumed mixer noise problem be fixed by a further conversion?
The bind is in many low end receiver designs, the mixer is also the
local oscillator, so most SW receivers that are variants of the All
America 5 design (and there were many) had very poor performance above
10Mhz or so.

Of course
providing 3-5 Khz selectivity at 40Mhz tends to be a bit challenging.
Q on the order of 10,000......

Single Conversion with a 455khz IF strip doesn't have problems with
bandwidth, but image rejection in the SW bands sucks big time.

That's true enough with inexpensive receivers which relied on a single
(de)tuned circuit for RF selectivity. But the better receivers would
cascade two or more tuned stages, isolated with RF amplifiers.
Actually most interesting design in a single conversion receiver I
think I ever was was in the mid 1960's Squires-Saunders built one with
a tuned RF stage with a Q muliplier on it, so they had a Q of a couple
thousand on the front end and made a killing on the gain as a result
of gain-bandwidth product. Suffices to say that with that sort of
front end selectivity, image rejection wasn't a problem. Obviously
impedance matching with the antenna was crucial to performance, but it
was undoubtedly the best single conversion HF receiver every
commercially built (and had a price tag to match).

Frank Dresser

matt weber ) writes:

Actually most interesting design in a single conversion receiver I
think I ever was was in the mid 1960's Squires-Saunders built one with
a tuned RF stage with a Q muliplier on it, so they had a Q of a couple
thousand on the front end and made a killing on the gain as a result
of gain-bandwidth product. Suffices to say that with that sort of
front end selectivity, image rejection wasn't a problem. Obviously
impedance matching with the antenna was crucial to performance, but it
was undoubtedly the best single conversion HF receiver every
commercially built (and had a price tag to match).

But the Squires-Saunders had a high IF. According to one quick check,
it was a first IF tuneable from 5 to 5.5MHz, and then 1MHz. That
was part of the wave of receivers with crystal controlled first mixers, in
order to get a nice tuning range and stability. If they'd gone to
a fixed IF, then either the local oscillator would have to be switched
from band to band, or premixed with a crystal oscillator before the signal
went into the signal mixer.

Note it's an example of how down conversion can still work. For so
long, people always thought in terms of the early superhet with the
IF being down in the hundreds of KHz range, but the issue isn't that
it was converted down but that the IF was so low.

That Squires-Saunders arrive as crystal filters were still a new
thing. There's a famous 1963 article in QST by Squires or Saunders (I
forget which), discussing the philosophy and design of the receiver.
Some of the issues were keep rf amplification before the mixer to
a minimun, and use the 7360 beam deflection tube for the mixer for
a well balanced mixer. Few or no receivers used a balanced mixer
before the SS-1R, at least not affordable receivers. So the front
end Q-multiplier was brought in to deal with the simple front end.
For the rest of the decade, the basic concept, a 7360 mixer and a front
end q-multiplier, bounced around in various construction articles. But
in some ways it was just because it had been done, because there
were no front end Q-multipliers after the late sixties or early seventies.
And of course, solid state components came along, making it easier to
build a balanced mixer, be it with schottky diodes or active components,
without a bunch of bulky tubes.

All the receivers I saw that used a front end Q-multiplier used a high
IF, ie at least 2MHz and most often 9MHz.

Michael

) writes:
m II wrote:
Perhaps he meant to say 'any sort of selectivity' ?

I re-read his posting, and I think he meant amplification. In context,
he was referring to the earliest vacuum tube days. The frequency
response of those tubes was limited. If I recall correctly, it was
limited by the physically large size and the spacing between the
filament, the grid, and the plate.


Howard Armstrong received the patent for the superhet, US patent number
1,342,885 in 1920. He wanted to receive what were astoundingly high
frequencies at the time, like in the 2 or 3MHz range.

At the time he cooked it up, even at the time the patent was issued,
there was no commercial radio broadcasting. The spectrum above
what is now the AM broadcast band was deemed useless (which is
why amateurs were relegated to "200 meters and down" after WWI.

I don't recall the schematic in Armstrong's patent, but if you look
in the history books, you find early schematics that use a chain
of RC coupled tubes for the IF strip, no selectivity.

Amplification has always lagged after frequency use. During WWII,
radar development was limited because they had problems getting
receiving tubes to work in the microwave frequencies, so they went
to diode mixers. It's pretty much always been easier to convert
to a lower frequency for amplification.

Michael

"matt weber" wrote in message
...
On Mon, 14 Nov 2005 04:07:57 GMT, "Frank Dresser"
wrote:


"matt weber" wrote in message
.. .
What breakthrough has made single
conversion so state of the art?


Absolutle nothing, in fact single conversion sucks unless it is an up
conversion, and even then, mixer noise will wipe out reception above
about 10Mhz absent a good tuned RF amplifier in front.


Why would up conversion mixer noise wipe out reception above 10 MHz? How
would the presumed mixer noise problem be fixed by a further conversion?


The bind is in many low end receiver designs, the mixer is also the
local oscillator, so most SW receivers that are variants of the All
America 5 design (and there were many) had very poor performance above
10Mhz or so.

OK, but I thought we were talking about up conversion.

I don't think converter tubes lose much gain above 10 MHz, but they are
awfully noisy, and their noise really jumps out at higher frequencies.
Beyond that, many of those old radios had a high impedance antenna input,
and the typical coax run would shunt the high frequencies.

I don't think there was much upconversion above 10 MHz during the converter
tube era, however.



Of course
providing 3-5 Khz selectivity at 40Mhz tends to be a bit challenging.
Q on the order of 10,000......

Single Conversion with a 455khz IF strip doesn't have problems with
bandwidth, but image rejection in the SW bands sucks big time.

That's true enough with inexpensive receivers which relied on a single
(de)tuned circuit for RF selectivity. But the better receivers would
cascade two or more tuned stages, isolated with RF amplifiers.


Actually most interesting design in a single conversion receiver I
think I ever was was in the mid 1960's Squires-Saunders built one with
a tuned RF stage with a Q muliplier on it, so they had a Q of a couple
thousand on the front end and made a killing on the gain as a result
of gain-bandwidth product. Suffices to say that with that sort of
front end selectivity, image rejection wasn't a problem. Obviously
impedance matching with the antenna was crucial to performance, but it
was undoubtedly the best single conversion HF receiver every
commercially built (and had a price tag to match).


Dunno. I never worked with that radio, and I never even worked with a
regenerative preselector. There was a regen preselector projector in one of
my ARRL handbooks, but I've been too lazy and unmotivated to build it. I do
know that regens have a sharp peak in their response curve, but their skirts
have a gentle roll off, typical of a single tuned circuit. I sorta picture
a very strong adjacent signal on the slope breaking through and modulating a
weak signal at the peak Maybe not, but all the other high end single
conversion radios I can think of used multiple RF stages, biased in the
linear (enough) region.

Frank Dresser

"Michael Black" wrote in message
...

) writes:
m II wrote:
Perhaps he meant to say 'any sort of selectivity' ?

I re-read his posting, and I think he meant amplification. In context,
he was referring to the earliest vacuum tube days. The frequency
response of those tubes was limited. If I recall correctly, it was
limited by the physically large size and the spacing between the
filament, the grid, and the plate.


Howard Armstrong received the patent for the superhet, US patent number
1,342,885 in 1920. He wanted to receive what were astoundingly high
frequencies at the time, like in the 2 or 3MHz range.

The story I remember is, during World War One, it was feared the Germans had
developed a way to communicate at 100 meters. Armstrong wanted to intercept
those communications, if they existed.


At the time he cooked it up, even at the time the patent was issued,
there was no commercial radio broadcasting. The spectrum above
what is now the AM broadcast band was deemed useless (which is
why amateurs were relegated to "200 meters and down" after WWI.

I don't recall the schematic in Armstrong's patent, but if you look
in the history books, you find early schematics that use a chain
of RC coupled tubes for the IF strip, no selectivity.

It's worth mentioning that there's a practical limit as to how much gain can
be obtained at any one frequency, and that practical limit was much lower
back in the earliest days. The superhet split it's gain at supersonic and
sonic frequencies, and could have much more gain without breaking out into
uncontrolled oscillation than a simple audio frequency amplifier. The tubes
of that era were just about useless as amplifiers at 3 MHz.

After Armstrong's invention, better triodes combined with better circuits
such as the Neutrodyne, as well as the screen grid tubes, put the TRF back
in the game into the early 30s, or so.





Amplification has always lagged after frequency use. During WWII,
radar development was limited because they had problems getting
receiving tubes to work in the microwave frequencies, so they went
to diode mixers. It's pretty much always been easier to convert
to a lower frequency for amplification.

Michael


Frank Dresser