[HBR] HBR2K Chapter 6 -- The story resumes

[hbr@mailman.qth.net]

Synopsis:  In earlier chapters (January - August '01) I set out the 
goal of updating the Crosby receiver idea from a 1950 concept to 
about the 1970's, while keeping it a set that the moderately skilled 
home constructor could hope to duplicate.  

I began trying to homebrew a 'hollow state' receiver based on front 
end and other parts from an FT-101E.   This approach seemed to 
offer a chance to build a receiver better than anything available on the 
ham market in the 60's. The FT-101 not only has much usable sheet 
metal (everything important but the chassis plate and front panel) and 
a nice cabinet, but nearly all of the miscellaneous parts -- switches, 
an S-Meter, a box in which to build the power supply, and most of 
the hardware to put the thing together.  Many of the parts -- 
particularly the permiability tuned front end, the VFO mechanicals, 
the tuning cap and filters -- are equal or even superior to anything 
used in tube ham gear, yeah, even Collins, in my opinion.   
(Powdered iron cup-and-core front end coils -- who else did that?)   
However the FT-101 and other gear of the 70's suffered from the 
various drawbacks of transistors, particularly in strong signal 
handling. And because 60's manufacturers were already struggling, 
none of the ham tube equipment was built with the best of the 60's 
tubes and circuits.  

It ought to be possible to have the best of all worlds, with FT-101 
parts sets in the $100+/- range it wouldn't be too expensive, and 
gosh -- it shouldn't even be that hard.

Should it?

The truth began to emerge in Chapters 1-5, posted as I got serious 
about the project.  I started as usual, at the audio stages, and got 
back as far as the second mixer before other life-things got in the 
way.   By that time I was dealing with a couple of serious problems.  

One was that I had somehow overlooked the fact that the FT-101 
crystal filters need to see about 1200 ohms on both ends.   I tried 
various simple measures but what would you expect, from simple 
measures?   Yesterday I did it right, added a cathode follower to drive
the filter at the proper impedence, and changed the 1st IF tube grid
resistor to 1200 ohms.   End of problem.

This may not cost an extra envelope because I used the triode half of 
a 6KE8 (6U8 with a better pentode) for the C.F.   The pentode half 
can be used for the 100 kcs calibrator leaving an empty socket in the 
planned location.   Total 13 tubes.    

One resolve during my 16-month break was to replace my highly 
modified IF transformers with units as close as possible to junkbox 
stock command set units.   That turned out to be simpler than 
expected; the 1415 kcs units from the 3-6 Mcs receivers require only 
changing the internal fixed caps to work fine at 3180 kcs.   (18 mmf 
primary and 22 mmf secondary for the first two, 18 mmf for the 
secondary on the third, replacing 180 mmf.)   Added bonus -- the last 
IFT has to be tapped because of the detector circuit and the 
command set output unit is already so.  I wound up using tapped 
transformers for all stages (higher Q and lower gain) but that was 
probably unnecessary and doing it again, I would not.  

The 2830 kcs units from 6-9.1 Mcs receivers seem a more natural fit, 
but
the more plentiful SCR-274N equipment used single tuned air core 
IFTs in
this receiver.   

I had had oscillation in the IF stages.   Changing the filter 
connections and improving grounding made things much better, even 
with the higher gain of the command set IFTs.  But ... the S-meter 
didn't act right.   It jumped around when I tuned the GDO across 
3180 kcs but didn't show any real increase.   Backtracking the 
problem, the AGC voltage was behaving correctly, but the 2nd IF 
stage voltages (the S-meter is connected between the screen and 
cathode) were not.  

????!?!   

For no particular reason I realigned the IF's and got a clue:  when the
set was turned on, a nice clean whistle went through the filter 
passband
and disappeared.   I went looking for a steady state oscillation but 
found
nothing -- after a minute of warmup, it was gone. Humm ... could the
bizzare S-Meter behavior be the result of a forced oscillation that
shifted the tube operating point to largely counter the effect of the 
AGC?
  Weird ...  

It could only be in the 2nd IF because the 1st IF has a 1200 ohm grid
resistor.   On a hunch I reversed the 3rd IFT coil connections, placing
the IF tube on the small end of the tapped winding rather than the big 
one
-- 1/3 of the total impedence rather than 2/3.   Bingo!   

The change also reversed the phase of the signal to the detector 
relative to the earlier stages.   If the detector was in the feedback path
(it didn't seem to be, but ...), that could have been the true fix.  

The 6EH7 (12,000 umhos transconductance) may be overkill, gain-
wise, but it has such amazing crossmodulation characteristics that 
it's the clear choice for serious receiver design.   

The tube was designed for TV IF service.   At -6.5 volts on the grid it
requires 100 mv of signal to produce 1% crossmod.   At -20 volts, 
(near
cutoff) it will handle half a volt of signal at 1% crossmod.   The
nine-pin socket makes it a difficult retrofit in most ham sets but I've
done it a few times (usually as an RF stage, sometimes also as the 
first
IF) and it makes a huge difference in some sets in which signals are 
clean
on an empty band and 'muddy' sounding on 80m in the evening.  Of 
course on
some sets the main problem is the mixer(s).

Anyhow the HBR2K is wired back as far as the output of the second 
mixer  and (seemingly) all working -- I'm listening to an HF broadcast
station on a clip lead connected to the second mixer grid.  Yeah, gain
isn't going to be a problem.   

Next thing is a couple of clean-up operations; I foolishly wired the five
position function switch of the FT-101 (LSB-USB-TUNE-AM-CW) for
LSB-USB-CW-AM, leaving an unused position.   I need to rewire it for
LSB-USB-CW-AM 2kcs-AM 4kcs.   At an early stage I made some 
dumb changes
in one of the chassis shield partitions; I have to replace that before I
can do any of earlier stages.

Then wiring the tunable first IF.   This is a design change from the FT-
101, which used a bandpass first IF. 

Then the RF stages and first mixer.   Because of the close quarters 
that's going to be a challenge.   I'm transplanting the FT-101 
bandswitch, coils, crystal socket and trimmer boards as nearly intact 
as
possible but there are still a whole lot of connections.   And again I
have to convert an oscillator from solid state to hollow -- the first
oscillator, which selects the band.   

Mixers, at least, are rarely unstable.   I hate that stuff.

Walt Hutchens
KJ4KV