[HBR] Another Receiver Project -- HBR-4, Part 24

[waltah at earthlink.net]

'Harder than I thought'

Well, injecting a maximum signal at the antenna jack gives 0.6V p-
p across the mixer grid tank circuit.   I don't have a signal generator 
capable of doing that directly; it would have to be a purpose-built 
power oscillator and I'd be right back to signal purity issues.   

Running a few numbers, though, I estimate the circulating current 
in that tank at that signal level (which is modestly above where 
intermodulation distortion sets in) at a sizable fraction of an amp.   
Ten turns on the small tunable coil, so we're looking at a few 
ampere-turns, in other words, not that far from the filament 
transformer range ... could that cause saturation of the core? 

Remember that we're not talking about visible distortion of the 
current waveform but only enough saturation to cause intermod 
levels nearly *80 db below* the input signal ... in other words, not a 
very large amount of distortion at all.   It certainly seems possible 
to me.

My hypothesis is that the limiting factor on IFDR is saturation of 
the powdered iron core of the coil tuning the RF stage output.   

After some squinting at the chassis and a look at the parts 
archives, I think I see a way to change the front end tuned circuits 
to fixed coils using a tuning cap without doing major surgery. I can 
use large enough coils -- a T68 toroid for the small coil and a T-106 
for the large one used in series on 40/80 -- to (I think) handle at 
least an order of magnitude more flux.   

Am I remembering something indicating that one of the FTDX-9000 
options is air-core coils in the front end?   Hummm ...

One of the things that's noteworthy about Ted Crosby's HBR-series 
is that their strong-signal performance was better than most 
contemporary commercial receivers.   I don't think that's unrelated 
to his use of high-Q air core coils at a time when the industry 
standard was a 3/8" paper form with a powdered iron slug.

In another department altogether, I think I have figured out why 
everyone went with antenna input on the G1 of the beam deflection 
tube mixer tubes.

It is true that a mixer is a multiplier and that output-wise it therefore 
does not matter whether a 1 V LO signal is put on a deflection 
plate and a 1 uV antenna signal on a G1, or the reverse.   However, 
what about a 10 V LO signal?   That's easy to achieve and would 
give 20db more conversion gain (than 1 V) if used on the deflection 
plates -- but would simply drive G1 into hard limiting if injected 
there.   

I think that the possiblity of higher gain by using higher LO voltages 
started us down the 'G1 for antenna input' path and that hobbyist 
designers may not fully have considered the trade offs when this 
choice was repeatedly made thereafter.   Namely, that going the 
other way with the superior-performing 6JH8 and using a highly 
linear moderate gain RF stage could give the best of all worlds -- a 
truly linear mixer receiver with potentially remarkable signal 
handling ability.

Of course some of the following designs were intended to be 
simple ones; for those, the G1 antenna input made sense all 
around.  But those 75A4 conversions should have been tried the 
other way.  

More to come ...

Walt
KJ4KV