With the trend for 9 Mhz IFs and other crystal ladder filters at those
sorts of frequencies, what do people adopt for the reception of
AM and FM, interpolation to a lower frequency, say, 455 kHz?

LC filtering at 9MHz being very broad band

On Fri, 20 Nov 2015, gareth wrote:

With the trend for 9 Mhz IFs and other crystal ladder filters at those
sorts of frequencies, what do people adopt for the reception of
AM and FM, interpolation to a lower frequency, say, 455 kHz?

LC filtering at 9MHz being very broad band

YOu can get good crystal filters at 9MHz, but they get costly. And it's
harder to make crystal filters the wider they get.

There may not even be good (and certainly not cheap) AM and FM bandwidth
filters at 9MHz.

A traditional scheme is 10.7MHz and 455KHz. IN the old days it would be a
few IF transformers at 10.7MHz, nowadays it's a wide ceramic filter
intended for FM broadcast reception. So it does a good job of getting rid
of the image which would be 910KHz (twice the 455KHz IF), but is too wide
for providing any real selectivity for narrow band operation at 10.7MHz.
THey are very cheap of course. For AM, there are lots of cheap ceramic
filters at 455KHz, and even for narrow band FM. There are even crystals
readily available to convert from 10.7MHz to 455KHz, 10.240MHz. That one
is used in some synthesizer ICs, and since there is enough consumer
equipment that uses them, they can be cheap and readily available.

At 455KHz, the deviation of an FM signal is a better percentage of the
signal frequency, so the FM detector is easier. At 10.7MHz, the recovered
audio may be relatively low. The classic CA3089 FM detector strip
required some work to get good recovered audio at 10.7MHz with a low
deviation FM signal (while for FM broadcast, it was fine).

I'm not sure many people building 9MHz crystal filters to put into
homebrew rigs are making them multi-mode. Of course, at 10.7MHz, there has
always been a lack of crystal filters, I think only because if anyone was
doing narrow work, they'd go to 9MHz. But that's not consistent; I found
an SSB CB set at a garage sale a couple of years ago, and it uses a
10.7MHz filter (I assume crystal but don't know for sure), but it's a
relatively wide filter, used for AM too.

If someone is building a multi-mode receiver, they'll probably go with a
10.7MHz first IF, and then have the various selective filters at 455KHz.
Or, they'd be using a roofing filter at some frequency above 30MHz, and
convert down. Early cellphones were a source of FM IF strips like that
(and since they included the required second conversion crystal, once you
had the filter, you had the means to get it down to where selectivty can
work. Some went from that high frequency down to 455KHz directly.

In some cases, there is enough selectivity at 9MHz, but they want to do
tricks, so there is a second conversion. So something like the Kenwood
830 has a decent 9MHz filter, and a lesser filter at 455KHz (which can be
upgraded), so you can get passband tuning. I think it's the 455KHz filter
that is lesser.

Sometimes they want to do tricks. I have a Sony SW1 (the very small one,
that came in a plastic case with ac adapter and active antenna), and it
uses a somewhat narrow 10.7MHz filter, and a good narrow one at 455KHz.
But, they pull the second conversion oscillator (which works since the
10.7MHz filter is wide enough), on every second "channel". This way, the
synthesizer only has to lock down to a 10KHz reference, but you can get
the needed 5KHz steps for shortwave broadcasting. Every second channel,
the second conversion oscillator is pulled by 5KHz. Very neat.

There was one earlier shortwave portable from Japan that had a traditional
analog oscillator feeding the first mixer, which converted down to
10.7MHz, and then down to 455KHz. But the oscillator feeding the second
mixer was somewhat variable, allowing for fine tuning. That wouldn't work
unless the first filter was sufficiently wide, but it has some useful
benefits for the homebrew receiver.

But it's always a tradeoff. Single conversion with a high IF trumps just
about anything, but the various conversion schemes over the years have
tended to have some value to it, either to make do (you don't have a good
filter at a high frequency) or to do things like passband tuning or that
method of fine tuning I mention above.

Now, with fancy DDS synthesizers, and a move to broadband stages (except
when you add tuned circuits), one could build a multi-mode receiver using
whatever IF filters they had. The mixer wouldn't care because it's
broadband except for the filter. The synthesizer can be retuned to get
the right injection frequency, and the readout made to show the signal
frequency, so nobody knows that you shifted from a 9MHz IF for SSB to a
10.7MHz IF for narrow band FM, and then to some other frequency for very
narrow CW.

Michael

"Michael Black" wrote in message
news:alpine.LNX.2.02.1511202332580.15821@darkstar. example.org...



Michael, thank-you for a very full and informative response; the reason for
my query
is that I have two 9MHz filters, one for SSB and the other for CW, which I
purchased for my lifelong project (going back to the mid 1960s :-( ) to
produce my own communications RX, dating back to the days when I was a
teenager with
a passionate interest in amateur radio, but no pocket money to pursue that
interest, and
a belief that the only gear I'd get would be that which I would make for
myself.

It seemed to me nowadays that such an RX would be incomplete without some
means
to resolve AM and FM transmissions.

Unfortunately, with the changes in technology over the years, the goalposts
are
forever changing, and the brushed aluminium 19" cabinet that I got for the
project has remained on the top shelf unsullied for 20 of those years!




On Fri, 20 Nov 2015, gareth wrote:

With the trend for 9 Mhz IFs and other crystal ladder filters at those
sorts of frequencies, what do people adopt for the reception of
AM and FM, interpolation to a lower frequency, say, 455 kHz?

LC filtering at 9MHz being very broad band

YOu can get good crystal filters at 9MHz, but they get costly. And it's
harder to make crystal filters the wider they get.

There may not even be good (and certainly not cheap) AM and FM bandwidth
filters at 9MHz.

A traditional scheme is 10.7MHz and 455KHz. IN the old days it would be a
few IF transformers at 10.7MHz, nowadays it's a wide ceramic filter
intended for FM broadcast reception. So it does a good job of getting rid
of the image which would be 910KHz (twice the 455KHz IF), but is too wide
for providing any real selectivity for narrow band operation at 10.7MHz.
THey are very cheap of course. For AM, there are lots of cheap ceramic
filters at 455KHz, and even for narrow band FM. There are even crystals
readily available to convert from 10.7MHz to 455KHz, 10.240MHz. That one
is used in some synthesizer ICs, and since there is enough consumer
equipment that uses them, they can be cheap and readily available.

At 455KHz, the deviation of an FM signal is a better percentage of the
signal frequency, so the FM detector is easier. At 10.7MHz, the recovered
audio may be relatively low. The classic CA3089 FM detector strip
required some work to get good recovered audio at 10.7MHz with a low
deviation FM signal (while for FM broadcast, it was fine).

I'm not sure many people building 9MHz crystal filters to put into
homebrew rigs are making them multi-mode. Of course, at 10.7MHz, there has
always been a lack of crystal filters, I think only because if anyone was
doing narrow work, they'd go to 9MHz. But that's not consistent; I found
an SSB CB set at a garage sale a couple of years ago, and it uses a
10.7MHz filter (I assume crystal but don't know for sure), but it's a
relatively wide filter, used for AM too.

If someone is building a multi-mode receiver, they'll probably go with a
10.7MHz first IF, and then have the various selective filters at 455KHz.
Or, they'd be using a roofing filter at some frequency above 30MHz, and
convert down. Early cellphones were a source of FM IF strips like that
(and since they included the required second conversion crystal, once you
had the filter, you had the means to get it down to where selectivty can
work. Some went from that high frequency down to 455KHz directly.

In some cases, there is enough selectivity at 9MHz, but they want to do
tricks, so there is a second conversion. So something like the Kenwood
830 has a decent 9MHz filter, and a lesser filter at 455KHz (which can be
upgraded), so you can get passband tuning. I think it's the 455KHz filter
that is lesser.

Sometimes they want to do tricks. I have a Sony SW1 (the very small one,
that came in a plastic case with ac adapter and active antenna), and it
uses a somewhat narrow 10.7MHz filter, and a good narrow one at 455KHz.
But, they pull the second conversion oscillator (which works since the
10.7MHz filter is wide enough), on every second "channel". This way, the
synthesizer only has to lock down to a 10KHz reference, but you can get
the needed 5KHz steps for shortwave broadcasting. Every second channel,
the second conversion oscillator is pulled by 5KHz. Very neat.

There was one earlier shortwave portable from Japan that had a traditional
analog oscillator feeding the first mixer, which converted down to
10.7MHz, and then down to 455KHz. But the oscillator feeding the second
mixer was somewhat variable, allowing for fine tuning. That wouldn't work
unless the first filter was sufficiently wide, but it has some useful
benefits for the homebrew receiver.

But it's always a tradeoff. Single conversion with a high IF trumps just
about anything, but the various conversion schemes over the years have
tended to have some value to it, either to make do (you don't have a good
filter at a high frequency) or to do things like passband tuning or that
method of fine tuning I mention above.

Now, with fancy DDS synthesizers, and a move to broadband stages (except
when you add tuned circuits), one could build a multi-mode receiver using
whatever IF filters they had. The mixer wouldn't care because it's
broadband except for the filter. The synthesizer can be retuned to get
the right injection frequency, and the readout made to show the signal
frequency, so nobody knows that you shifted from a 9MHz IF for SSB to a
10.7MHz IF for narrow band FM, and then to some other frequency for very
narrow CW.

Michael