[Collins] 75A4 Roofing Filter ?

[Dr. Gerald N. Johnson]

On Thu, 2009-01-22 at 23:49 -0800, Adam Farson wrote:
> Hi John,
>  
> The architecture of the 75A-4 is such that a "roofing filter" in the
> modern sense is an invalid concept.
>  
> The 1st mixer down-converts the RF signal to a 1.5 - 2.5 MHz variable
> 1st IF. The 1st LO is crystal-controlled; the bandswitch selects the
> appropriate crystal. The 2nd mixer (excited by a PTO tuning 1955 -
> 2955 kHz) down-converts the variable 1st IF to a fixed 455 kHz IF. The
> 2nd IF filter is a 455 kHz Collins mechanical filter; this is the
> selectivity filter, not a roofing filter.
>  
> http://www.collinsmuseum.com/75a4.html
>  
> I hope my explanation is reasonably accurate, and helpful (I am not a
> Collins owner, but would love to have an HF9500 in my shack!)
>  
> 
> Cheers for now, 73,
> Adam VA7OJ/AB4OJ
> 
> 
That's all correct. What I was going to say when I got around to it.
With different frequencies the explanation also applies to all the 75A,
75S, 51J, 51S, and R390() receivers. The broadband exposed mixers are
the weak points in all these receivers, the multigrid tubes make noisy
mixers and have limited dynamic range. Though the RF bandpass filters
that are often tracked with the LO are not nearly as wide as DC to 30
MHz often seen in the recent receiver designs that upconvert to a
roofing filter at 45 MHz. The tuned RF stage definitely helps keep out
strong signals from other band. It doesn't help keep out the KW station
down the block in the same pile up.

It is possible, and there are numerous articles about, to work over the
mixers in these receivers to improve dynamic range. W0MLY (now SK) used
to sell a solid state mixer for the 75S receiver, though he and I
disagreed on which mixer was most critical. One published mod for the
'A4 replaced the first mixer with a 12AT7 long tail mixer circuit that
was lower noise, but also lower gain.
-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer