On Thu, 03 Mar 2011 10:15:42 -0800, Joel Koltner wrote:

"Tim Wescott" wrote in message
...
Does Mouser still sell them?

Somewhat surprisingly, yes, they do. (Although of the many hundreds
they have listed in their system, it's only something like a couple
dozen they actually have stock of.)

The point of the IF transformer was twofold: to provide selectivity,
and to give a good impedance match between stages. I think there must
have been a standard receiver design, as there just seemed to be one
choice for each coil -- this in spite of the fact that as soon as you
start juggling feedback and/or standing currents, you change the
impedances, and therefor the required transformer.

Agreed -- and I additionally suspect that many people who made those
receivers didn't necessarily understand the design itself that well.
I.e., if it basically worked and wasn't clearly deaf, it became a
product -- few if any AM/FM receivers found on the shelves of, e.g.,
Sears listed things like their sensitivity, adjacent channel rejection,
etc. (From the flip side, though, I suppose that's one of the things
the FCC did: Coordinated frequencies and power levels such that just
about any radio would work "reasonably" well;

In fact, they did: if you pay attention to the analog TV channels and the
radio channels in any one market, you'll find that they're spaced apart
by at least one 'dead' channel, while those 'dead' channels are 'live' in
adjoining broadcast areas.

For instance, in Portland, Oregon the TV channels were 2, 6, 8, 10 and
12, while Salem got the odd-numbers.

it's a very different
market from, say, amateur radio where there's little or no coordination
of these parameters and the customer may very well want to try to
listen to some QRP station on 14.15MHz while there's some big gun
blasting away on 14.14MHz...)

A different problem, too: broadcast implies a few transmitters and lots
of receivers, so you hold down system costs by skewing the requirements
very heavily toward cheap receivers, even at the cost of efficient
bandwidth usage.

Contrast that with pre-broadcast satellite TV, where you save some bucks
in the satellites, at the cost of needing big dishes and sensitive
receivers. That's OK, because the dishes are few (before they started
getting into the hands of the consumers). Then look at the direct-to-
consumer satellite systems, where the dish is just two feet across, and
you have to assume that the satellite is way more expensive.

I think if I were going to design a broadcast-band receiver, I wouldn't
just re-do the old schematic from 1960 -- I'd start from a clean sheet
of paper, and see where I could go from there.

I agree insofar as the actual design goes, but I like to study these
older technologies because I think it's all too easy to not realize just
how good the performance of some of the old designs were (for your new
design you'd like to start with specs that are hopefully some
improvement or at least as good as the old ones...), and also because a
lot of the same *techniques* can be applied to modern designs just as
well as they could to old ones (e.g., varactor tuning is just an
evolution of mechanical tuning, neutralization applies just as much to
BJTs as it does to tubes, etc.).

True, but you need to know where the old techniques aren't going to serve
you well (e.g. the passives involved in neutralizing a transistor stage
to get a few more dB gain cost more than another stage).

(Actually, I think the first
thing that'd go onto that clean sheet of paper would be an ADC -- I'm a
luddite in a lot of ways, but not in how I'd like to see a receiver
laid out).

That works, but consider that if you digitize the entire broadcast FM
band at once (all 20MHz of it), compared to a $20 superhet receiver:

-- Your weak signal sensitivity may be worse, since your dynamic range
is spread across the entire band rather than just what'll fit through an
IF filter.
-- You'll likely suck rather more power from a battery. -- It'll
probably cost more.
-- For the channels you can receive well, you'll have infinitely more
options on being able to reduce noise, change your audio bandwidth,
recovery stereo, decode RDS, an so on. :-)

Good point. But getting things through an IF filter, _then_ going into
an ADC isn't a bad thing, unless you're trying for the absolute minimum
of power and circuit cost.

--
http://www.wescottdesign.com