Jeff Goldsmith wrote:

BC 312/342 IF (at least mine) is 470KC.

The 3-6 MC ARC 5 receiver IF is 1415, the 6-9.1 range ARC 5 radio IF
is 2830.

Nice thing about these ARC 5 sets is often the transformers have been
fully enclosed for 60 years and are still shuny and new looking. A beat
up 6-9.1 range one wouldn't be too hard to turn up (I've got one here
somewhere if anyone seriously wants to give this notion a shot).
High IF's do sound good. The Philco 42-350 (IIRC) uses the same ~5MC
IF for all the bands. Quite a bit less gain per stage than 455KC will
deliver, but closer to simple receiver quality audio.

5MHz is high, and many old TVs have 4.5 MHz IFTs within that could be altered.

The final Q of a normal 455 kHz 2 stage IF system is usually 65.
If the final Q of the 5 MHz IF system was also 65, the pass band would be 77 kHz,

allowing stations only 45 kHz apart on the BCB to be all too plainly heard,
although the audio response could be out to 38 kHz.

That's why I proposed an IF just above the BCB at 2 MHz,
where an overall Q of 65 would give a pass band of 31 kHz,
with some hope of attenuating stations only 45 kHz away from the wanted station.
Audio BW could be up to 15.5 kHz.
So a slightly higher overall Q would be permissable.
There is still need for 6 consecutive tuned circuits to give a steep
skirt slope just away from resonance peaks.
It may be possible to have all single tuned LC circuits, each with a Q
of say 40, or a pass band of 50 kHz, so that after 6 circuits this has been
reduced to
20 kHz or therabouts.

But then each single tuned circuit has to be coupled in pairs, which
involves an insertion loss, and pairer critically coupled IFTs and active but low
gain
stages are required.

Patrick Turner.







Jeff Goldsmith

Henry Kolesnik wrote:

TRFs are nice receivers, I had several back in the 1950s becasue that's all
I could find for free or next to nothing. The problem with a tuneable TRF
is that the bandwidth increases as you tune to higher frequencies. BW=f/Q
with Q being relatively constant over the BCB. Since you seem to have a
passion for a fix tuned TRF I'd suggest using old IF transformers for your
coils. It should be relatively easy to remove turns to get the freq up.
Some receivers, particularly military ones were made with IFs in the BCB.
The BC-454 that covered 6 to 9 Mcs had an IF of 1415 Kc. The BC-348 has an
IF of 915 Kc and the BC-312 and BC-342 also skipped the BCB but I can't
recall their IFs but I'll bet Google will find em. Scour the hamfests and
you'll find plenty but prices are going up. If you can find a BC-946 you'll
have a gem for a BCB Rx.
--
73
Hank WD5JFR

"Jon Noring" wrote in message
...
In Patrick Turner's latest message replying to the thread I started on
building a high-performance AM tube tuner, he stated his skepticism
that a pure TRF circuit will, in a practical sense, meet the specs I'm
looking for (to meet the specs will require an impractical number of
RF amplifier stages, such as six or more.)

Within the design constraints he selected, I cannot disagree with his
conclusion.

However, one of the constraints he made, and that most make, is that
tuning is to be done by an infinitely variable multigang air
capacitor.

I wonder if the same conclusion would be reached if we approach this
from a single channel perspective? That is, what if we fix the
frequency we want to receive (e.g., 830khz, or 1420khz), and then for
each stage optimize the parallel RLC circuit (or use some other tuned
circuit with the right resonance and bandwidth response -- crystals?)

With the much greater design freedom this should allow, could we now
possibly meet the specs with only two or three RF amps, for any
frequency in the 500 to 1800khz range? For a simplified diagram of a
two RF stage TRF receiver, see page 7 of:


http://www.technology.niagarac.on.ca...531unit6rx.ppt

Now, if this single channel approach indeed makes TRF more attractive,
then the next step is obvious -- consider building the TRF AM tuner
with selectable channels. Obviously, the downsides to this a the
need for a switch, the complexity and cost of having lots of channels
if we want to cover the whole BCB range (e.g., 120+, every 10khz in
North America, and 9khz in Europe and elsewhere), and the issue of
oscillation due to interference between the RF stages (shielding and
proper layout design will likely be very critical.) I envision the
tuned circuits for each channel being put on a mini-board which will
plug into a slot, but then there may be other ways to handle the large
number of tuned circuits necessary to cover the whole BCB (360+ of
them -- this does sound quite expensive.)

Thoughts?

*****

Of course, the other alternative is to modernize a proven design of a
super-het tube radio from yesteryear (the final audio amplifier will
be stripped out, replaced with a line-out.) Which radio circuits
would you nominate (with improvements), which provide excellent audio
fidelity (probably with variable bandwidth control), very good
sensitivity and selectivity, and other desirable specs? In addition,
the circuit design should be "forgiving", so those who build it from
kits (with pre-made PCB boards and proper guidance on layout and
shielding) will likely be successful -- the parts, too, should be
purchasable as NS from somewhere (NO SCROUNGING for old used parts as
I've noted before). We probably need to make special arrangements for
producing the various coils, since not everyone is going to want to
wind their own coils.


Jon Noring