[ARC5] Selectivity improvement in ARC-5 receivers.

[Brian Clarke]

Hello Ken,

All is not lost, except access to the synaptic connections to your long 
unused memory cells.

The matter of which you speak is covered well at a quasi-conversational 
level in the RSGB (1987) 'Radio Communication Handbook' 5th edition, section 
on 'Interstage coupling' (pp 6.20 to 6.25). Fred Terman (1955) 'Electronic 
and radio engineering' Ch 3, 'Properties of circuits with lumped constants' 
gives a more rigorous treatment with more formulae in 38 pp. The selectivity 
curve, ie, the slope of the skirts, shown in the RSGB handbook appears as 
Fig 3-11 in Terman. Fritz Langford-Smith (1960) 'Radio design handbook' 
supplies a 49 pp chapter - Ch 26 'Intermediate frequency amplifiers'; it is 
similar to Terman in terms of level and design detail; there is an appendix 
devoted to calculating the coupling coefficient, and there is a major 
section on selectivity and the calculation thereof.

The slope of the skirts, in dB per decade, depends on the frequency, the Q 
of the reactive elements and the coupling coefficient between the coils. Q 
is not only that of the reactive components, but also includes generator and 
load impedances. The coupling coefficient is related to coil spacing and 
other matters of geometry, such as the polarisation of the windings and 
their co-axiality.

Terman's and Langford-Smith's books are for designers; the RSGB handbook is 
for talented amateur technicians. I suspect the ARRL Handbook would be 
similar in level to the RSGB handbook - they are not really design reference 
manuals in the way Terman and Langford-Smith are.

In short, it ain't just frequency. And ultimate stop-band attenuation will 
depend on shielding as well.

Cheers, Brian, VK2GCE.


On Monday, June 02, 2014 7:03 AM, Ken posited:


> This month's Electric Radio Magazine has Jeep Platt's article on the 
> subject
> above in it. Jeep's article is very well written and very useful.
>
> However, I have had a question concerning ultimate stop-band selectivity 
> for
> some time now after a discussion with another of our list-members during 
> our
> discussion on Jeep's final method.
>
> It is my understanding that the ultimate stop-band selectivity is 
> determined
> NOT by any physical characteristic of the IF transformers, but, rather, by 
> the
> IF frequency itself.
>
> I.e., at the 60 db down point, the true selectivity is determined more by 
> the IF
> frequency than the separation of the two IF coils.
>
> I vaguely remember from my EE course in filters that it was "so many db 
> per
> octave", or something similar.
>
> Can anyone here tell us what I am even talking about here, and fill in the
> details?
>
> I'm at a loss, but still very confused.
>
> Kenneth G. Gordon W7EKB