[HBR] Another Receiver Project -- HBR-4, Part 9

[waltah at earthlink.net]

Two problems:

1.  When I installed the brand new 'proper' 80M crystal (18 Mcs), it 
drifted like crazy.   Whoops -- I had been using a 6 Mcs rock on 
the 3rd overtone.  The new crystal was a fundamental crystal 
because that's all ICM does at 18 Mcs.   But it is necessarily 
much thinner and the drive level that was perfectly okay for a 6 Mcs 
HC 49 working on the 3rd overtone was way too high for an 18 Mcs 
HC 49.

The answer to that might have been to find someone to make a 3rd 
overtone 18 Mcs crystal -- other companies do make them.   
However the 42.5 Mcs crystal for 10M (fundamental about 14 Mcs) 
wouldn't be much thicker than the 18 Mcs one.   This is a problem.

Basically the Butler oscillator power level has to be reduced. 

2.  Though I have adequate overall gain, the sensitivity was an 
unsatisfactory 3 uV for 6 db S+N/N.   That calls for greater mixer 
gain, in particular, I needed more drive to the mixer deflection 
plates.   That called for *more* output from the oscillators; since 
the VFO was already producing harmonics, the increased output 
would have to come from the Butler oscillator.

Whoops.

After some thinking, I decided that the answer was to reverse the 
oscillators.   The Butler osc. would do better into the control grid of 
the premixer where less signal is required; redoing the VFO as a 
push-pull oscillator would allow ample drive for the premixer 
deflection plates.   Heck, at 5 Mcs you can probably get five watts 
from a 6J6 with the sections in push-pull; there's not going to be a 
problem getting 10V p-p at a tiny fraction of an mA.

It was easier than it might sound.   I have to do the dial tracking 
from the start (had hoped to avoid that) but that shouldn't be more 
than a few hours of work.   The low warm-up drift of the 6GK5 
oscillator probably can't be acheived but if so, that's the price of 
using a more powerful, lower gain tube and a circuit that gives a 
less favorable oscillator parts layout.

The circuit is the push pull version of a grounded cathode Colpitts.   
Center tapped coil from plate to plate, caps from plates to opposite 
grids and from grids to ground.  (I invariably get the 'opposite' part 
wrong the first time, and this was no exception.)   The tuning cap is 
a two section with a section connected from each plate to ground.  
The coil tap floats; circuit symmetry is established by the plate to 
grid and grid to ground capacitors which are chosen to be most of 
the tank capacitance and should be pretty closely matched.  
Excellent waveform, plus output can be taken push-pull from the 
grids which are a relatively low impedance point.

The lower gain isn't quite the issue it seems because in a push-pull 
configuration there are two tubes.    But the push-pull circuits make 
it difficult to get isolation beween the tube and tank as you can with 
the Vackar or Clapp single ended circuits.  However, my 
impression is that the VFO is sufficiently stable.   

(The 6J6 was certainly the wonder of its day.  1942, perhaps?   It 
was among the very first miniature tubes, probably designed 
specifically for lower power radar applications like the APS-13 tail 
warning radar and APN-1 radar altimeter -- it was used as a pulsed 
self-oscillating transmitter in both sets, delivering I don't know how 
much peak power.   It oscillates pretty easily at 500 Mcs and 
delivers a few watts CW well into the VHF.   

(With a transconductance of 5300 mhos at 100 plate volts it's no 
slouch even by much later standards.   And it's quite beautiful to 
see in operation:  unlike the modern tubes with wimpy little 
filaments all wrapped up in electrodes, the 6J6 spills light all over 
the chassis.   A serious vintage 2M AM station would use a fistful 
of 'em on an open chassis, ending up with maybe a pair of 15Es?  
8025s?  We're talking *decorative* here, not *practical* so an  
832A is out.

(Another project for another day.

(The 6J6 is a highly satisfactory push-pull audio output stage, too, 
delivering the few hundred milliwatts needed for the usual 'station 
filling' volume without a problem.   And I have used 6J6's as push-
push mixers at HF with very good success.)

With the oscillators rearranged I've got way more overall gain than 
is desirable -- the meter pins on any reasonably strong signal.   So 
the IF gain needs to be cut back some ... that's easy enough to do! 
Probably the first step is to adjust the crystal filter loads for a flatter 
passband, since that may lead to a reduction of gain.   As a plus, 
all the connections in the oscillator/premixer area got much shorter 
because the pin layout of the 6JH8 favors this oscillator 
arrangement.  

I finally wired up the calibrator.   *That*, at least, worked the first 
time.    There's space for and I really will add a 25 kcs multivibrator. 
Yeah, another 6J6.   

Sure enough, the 'birdies' are stronger than they were and I'm sure 
there will be more problems but I'm feeling pretty good about the 
project right now.   

Walt
KJ4KV