From: Ken Scharf on Sun 26 Jun 2005 22:41

wrote:
From: Ken Scharf on Jun 25, 6:26 pm
wrote:

I am looking for a nice schematic, and perhaps additional info, on a 12V
(fil. & plate) tube superhet Rx; just something fairly simple for CW/SSB
reception without bells and whistles, and preferably for 75M/40M reception.

Setchell-Carlson managed to do it with loctal-base tubes running
off of a 24/28 VDC aircraft bus back in WW2 times. No dynamotor,
just that cute little box referred to as a "range receiver" or
BC-1206.

Converting a "Q5-er" from tubes to FETs would be rather easy.
No real need for the 2N2222s. Would have gobs of space left over
inside an already compact box.

This is the idea of a cascade fet-bipolar circuit. The drain of the fet
drives the emitter of the bipolar and the bipolar's base is biased so
that the base sits about half way up the power supply (two resistors).
The fet is self biased with a source resistor for the desired current.
The gain of this circuit is higher than a fet by itself, and the output
impedance will be higher.
____ __
-|_ \_/
| |

If the 135 KHz IF has an equivalent total Q of 100, the BW would
be about 1.3 KHz, not all that swift for CW and too shart for SSB.

The AnArc-5 Q5'ers had 6 tuned circuits in their If but the '1206
only has 3 or 4. So it's BW might be a tad wider. The 1206 also has
no bfo.

The ARC-5 LF version has IF TRANSFORMERS but the insides are
easy to see for detail. The 6 to 9 MHz version IFs (2830 KHz)
had only one tuned circuit in each IF can.

Can't remember what the BC-1206 used in their IF cans, but that
is easy to change inside the present can (it will look just the
same on the outside).

I can't see any advantage of going cascode (JFET-bipolar) when an
insulated gate FET would closely approximate a vacuum tube for
both input impedance (all capacitive) and drain/output impedance
(many hundreds of KOhms). Depends on what is available for the
vacuum-to-solid conversion.

Note: Both input and output impedances will affect the impedance
and Q of the tuned circuits...and the GAINS...for a conversion
with minimum circuit changes. All that has to be taken into
account.

However, the BC-1206 tuning range of 195 to 500 KHz results in
an image at the converter input of 390 to 1000 KHz away from the
desired band. That's worse than the 910 KHz image of an old
455 KHz IF.

The 1206's dial only goes up to 400khz, but it tunes past that. Maybe
it goes to 420 or 450khz, I don't think it goes as high as 500khz.
(so maybe I don't have to expand the range for my needs).

The variable tuning capacitor max:min ratio is roughly 11:1 (no
external parallel capacity). Since the resonant tuning range is
the square root of that, you could have it tune 190 to 570 KHz
with a 9:1 max:min change in resonance capacity. The only real
problem is getting the variable LO tuning to track the front end
since it would tune 325 to 705 KHz (for the high-side) and that
would be a 2.169:1 frequency ratio or 4.706 capacity change ratio.


The linear-in-capacity rotation of the 3-gang variable is good
for an RX Noise Bridge having an expanded parallel-C range...
which is what I used them for. :-)

True, but the rig is just to cute to canabilze.

Heh heh heh. I agree with you there. It probably is most dense
of any other 1940s era design using 8-pin "medium" size tubes,
including the very compact ARC-5 receivers. Not quite as dense
as the SCR-300 Walkie-Talkie (BC-1000) which stuffed 18 7-pin
"miniature" tubes, a half-dozen coil/transformer cans, and a
five-gang (!) variable capacitor into an approximate 5" x 10"
chassis. [Galvin/Motorola did real good in packaging there]

KEEPING the tubes as-is, with the possible exception of the AF
out (28D7 ?) would yield a straight-AM receiver. ADDING a BFO
via under-chassis bipolar transistor would result in minimum
cannibalization (one extra switch on front panel)...RF and IF
could remain the same. I've forgotten the audio output circuit
of it but recall that AF output power for old aircraft was
rather high (ambient noise in old warbirds was terrible). In a
quiet land environment, the audio output needed is rarely 200
milliWatts into a speaker, much less for headphones. The SCR-300
walkie-talkie (headphones only) had a maximum of 3 mW audio out!

Hint on supply conversion: A 12.6 V "filament" transformer can
do the job for all-12 V filaments in parallel. A voltage
doubling rectifier (even a tripler if need be) will yield the
quite-low B+ for plate and screen. I once did that (long time
ago) with a single transformer and voltage tripler for B+, all
working into a set of power resistor load boxes, all in a half
hour of tack-soldering. Got 30 VDC for B+ at 115 VAC line,
less than 5% ripple. Didn't have the BC-1206 then. A critic
looking on said "that isn't a good enough test...gotta have the
'real' load on it to 'prove' it." :-) [some folks just aren't
convinced until someone else has made a finished production
product!]

wrote:
From: Ken Scharf on Sun 26 Jun 2005 22:41


wrote:

From: Ken Scharf on Jun 25, 6:26 pm

wrote:

I am looking for a nice schematic, and perhaps additional info, on a 12V
(fil. & plate) tube superhet Rx; just something fairly simple for CW/SSB
reception without bells and whistles, and preferably for 75M/40M reception.

Setchell-Carlson managed to do it with loctal-base tubes running
off of a 24/28 VDC aircraft bus back in WW2 times. No dynamotor,
just that cute little box referred to as a "range receiver" or
BC-1206.


Converting a "Q5-er" from tubes to FETs would be rather easy.
No real need for the 2N2222s. Would have gobs of space left over
inside an already compact box.

This is the idea of a cascade fet-bipolar circuit. The drain of the fet
drives the emitter of the bipolar and the bipolar's base is biased so
that the base sits about half way up the power supply (two resistors).
The fet is self biased with a source resistor for the desired current.
The gain of this circuit is higher than a fet by itself, and the output
impedance will be higher.
____ __
-|_ \_/
| |


If the 135 KHz IF has an equivalent total Q of 100, the BW would
be about 1.3 KHz, not all that swift for CW and too shart for SSB.


The AnArc-5 Q5'ers had 6 tuned circuits in their If but the '1206
only has 3 or 4. So it's BW might be a tad wider. The 1206 also has
no bfo.


The ARC-5 LF version has IF TRANSFORMERS but the insides are
easy to see for detail. The 6 to 9 MHz version IFs (2830 KHz)
had only one tuned circuit in each IF can.

Can't remember what the BC-1206 used in their IF cans, but that
is easy to change inside the present can (it will look just the
same on the outside).
I seem to remember two trimmer caps so I guess there were a total
of four tuned circuits. I think the ham radio article I saw on
converting the rig did add a bfo using a tv horizontal coil cut
down as the bfo coil. I could also remove some wire from an old
BC453 bfo coil, of which I have a few in the junk box. (Not to
metion about a dozen 85khz if cans).

Which reminds me.....
I was thinking of building a solid state version of the old
Drake 2B receiver with it's 80meter first (tunable) IF, 455khz
second if, and 85khz third if (using Arc5 IF transformers) instead
of Drakes 50khz (close enough). I'd use a BC454 tuning cap padded
down to cover 3.5-4.1mhz just like Drake. The first oscillator
would be a PLL to cover everything (beats grinding on ft243 rocks).


I can't see any advantage of going cascode (JFET-bipolar) when an
insulated gate FET would closely approximate a vacuum tube for
both input impedance (all capacitive) and drain/output impedance
(many hundreds of KOhms). Depends on what is available for the
vacuum-to-solid conversion.
Junction fets are rather low gain, though compared to tubes with
28 volts on their plates, probably a good match. The cascade circuit
can be built with either two jfets, or one jfet and one bipolar. The
jfet/bipolar version combines the high gain of a bipolar transistor
with the high input impedance of the fet. That's the reason I'd give
it a try (plus I enough of BOTH kinds in the junk box). I'd replace
the converter tube with three fets, two in differcial for the mixer,
and one for the oscillator.


Note: Both input and output impedances will affect the impedance
and Q of the tuned circuits...and the GAINS...for a conversion
with minimum circuit changes. All that has to be taken into
account.


However, the BC-1206 tuning range of 195 to 500 KHz results in
an image at the converter input of 390 to 1000 KHz away from the
desired band. That's worse than the 910 KHz image of an old
455 KHz IF.


The 1206's dial only goes up to 400khz, but it tunes past that. Maybe
it goes to 420 or 450khz, I don't think it goes as high as 500khz.
(so maybe I don't have to expand the range for my needs).


The variable tuning capacitor max:min ratio is roughly 11:1 (no
external parallel capacity). Since the resonant tuning range is
the square root of that, you could have it tune 190 to 570 KHz
with a 9:1 max:min change in resonance capacity. The only real
problem is getting the variable LO tuning to track the front end
since it would tune 325 to 705 KHz (for the high-side) and that
would be a 2.169:1 frequency ratio or 4.706 capacity change ratio.

Well that doesn't take into account the distributed capacitance of the
coil windings, which would be considerable at this frequency what with
the required inductance. (ever wonder why a grid dipper covers a much
higher min-max frequency ratio as the frequency band goes up?) Also
IIRC there are trimmer caps in parallel with the tuning caps for
alignment as the coils are fixed. To make the oscillator track, a
series padder cap is used to reduce the max. capacitance value of the
oscillator section. -OR- a parallel padder could be used to swamp out
the range of the section.



The linear-in-capacity rotation of the 3-gang variable is good
for an RX Noise Bridge having an expanded parallel-C range...
which is what I used them for. :-)


True, but the rig is just to cute to canabilze.


Heh heh heh. I agree with you there. It probably is most dense
of any other 1940s era design using 8-pin "medium" size tubes,
including the very compact ARC-5 receivers. Not quite as dense
as the SCR-300 Walkie-Talkie (BC-1000) which stuffed 18 7-pin
"miniature" tubes, a half-dozen coil/transformer cans, and a
five-gang (!) variable capacitor into an approximate 5" x 10"
chassis. [Galvin/Motorola did real good in packaging there]

KEEPING the tubes as-is, with the possible exception of the AF
out (28D7 ?) would yield a straight-AM receiver. ADDING a BFO
via under-chassis bipolar transistor would result in minimum
cannibalization (one extra switch on front panel)...RF and IF
could remain the same. I've forgotten the audio output circuit
of it but recall that AF output power for old aircraft was
rather high (ambient noise in old warbirds was terrible). In a
quiet land environment, the audio output needed is rarely 200
milliWatts into a speaker, much less for headphones. The SCR-300
walkie-talkie (headphones only) had a maximum of 3 mW audio out!

You can't get much power out of a tube at 28v on the plate. The 28D7
was up to the task being a double power pentode with the sections in
parallel (why not push pull?). I wouldn't call solid state conversion
canabilzing.

Hint on supply conversion: A 12.6 V "filament" transformer can
do the job for all-12 V filaments in parallel. A voltage
doubling rectifier (even a tripler if need be) will yield the
quite-low B+ for plate and screen. I once did that (long time
ago) with a single transformer and voltage tripler for B+, all
working into a set of power resistor load boxes, all in a half
hour of tack-soldering. Got 30 VDC for B+ at 115 VAC line,
less than 5% ripple. Didn't have the BC-1206 then. A critic
looking on said "that isn't a good enough test...gotta have the
'real' load on it to 'prove' it." :-) [some folks just aren't
convinced until someone else has made a finished production
product!]

From: Ken Scharf on Jul 7, 10:19 pm

wrote:
From: Ken Scharf on Sun 26 Jun 2005 22:41
wrote:
From: Ken Scharf on Jun 25, 6:26 pm
wrote:


The 1206's dial only goes up to 400khz, but it tunes past that. Maybe
it goes to 420 or 450khz, I don't think it goes as high as 500khz.
(so maybe I don't have to expand the range for my needs).

The variable tuning capacitor max:min ratio is roughly 11:1 (no
external parallel capacity). Since the resonant tuning range is
the square root of that, you could have it tune 190 to 570 KHz
with a 9:1 max:min change in resonance capacity. The only real
problem is getting the variable LO tuning to track the front end
since it would tune 325 to 705 KHz (for the high-side) and that
would be a 2.169:1 frequency ratio or 4.706 capacity change ratio.

Well that doesn't take into account the distributed capacitance of the
coil windings, which would be considerable at this frequency what with
the required inductance. (ever wonder why a grid dipper covers a much
higher min-max frequency ratio as the frequency band goes up?) Also
IIRC there are trimmer caps in parallel with the tuning caps for
alignment as the coils are fixed. To make the oscillator track, a
series padder cap is used to reduce the max. capacitance value of the
oscillator section. -OR- a parallel padder could be used to swamp out
the range of the section.

As an aid to setting a tuning range, check out the February 1977
issue of Ham Radio magazine (under my byline). It covers adding
parallel caps, series caps, series-parallel and parallel-series
combinations. Easy formulas, all algebraic.

Worst case distributed capacity with typical inductances used at
this frequency might be 12 pFd.

Given a 32 to 352 pFd variable min:max, the C ratio = 11:1

Add 12 pFd to each section and C ratio = 8.0889:1 and F ratio =
2.8441:1. Add a total of 18 pFd in parallel and C ratio - 7.3800:1
with F ratio = 2.7166:1 and that is good from 190 to 516 KHz.

There's a near-infinite number of solutions for series-parallel
combinations for the LO but only a few will track adequately
with the antenna and RF amp stage. I suggest doing a simple
computer program for that kind of thing. It can check resonance
of the LO tank against the Antenna-RF tanks and find the
difference in resonance at as many frequencies within range as
you want.