[HBR] Gain in IF amplifiers

[N2EY at aol.com]

In a message dated 7/24/06 11:49:41 AM Eastern Daylight Time, 
Hopperdhh at aol.com writes:


> Thanks for your thoughtful reply.  Well, its been 40+ years ago so I  can't 
> 
> remember much of the details of what I tried as far as taming the  S-38.  
> Hopefully I've learned something since then, but I do remember  trying 
> everything I could think of at the time.  I must confess that I  haven't tried 
> lately to 
> duplicate my failure.
> 

It's quite possible that the layout would not have permitted stability no 
matter what. I learned that the hard way!

There's another trick I forgot to mention - POLARITY!!

If an IF strip regenerates, it can sometimes be cured by reversing the leads 
to *one* winding of *one* transformer. What this does is introduce a 180 
degree phase shift to the overall strip. Doesn't work in all cases but worth a try.


> In addition to input and output IF transformers, Miller and others also  
> supplied interstage transformers.  I know that the output transformers were  
> 
> closer coupled than the inputs because of the lower load impedance of the 
> diode  
> detector.  I wonder what the difference between the input and interstage  
> transformers was.
> 

I think they may have been coupled somewhat differently in order to preserve 
the bandwidth in an AM receiver. If you're trying for "10 kc" selectivity in a 
BC set, and you have multiple IFTs that are critically coupled, the overall 
BW may be too narrow, or may be shaped in a way you don't like. 


> Your reply certainly makes sense.  However, a voltage gain of 280  times is 
> 
> 48.9 db not 43 db.  

Right you are! That's what I get for doing the math in my noggin!

Assuming that the transformer is exactly  critically 
> 
> coupled, I think that the RL drops to one half of the unload  primary 
> impedance.  
> Usually the coupling is less than critical, but that is  nit picking.  
> Anyway, I 
> think that a gain of nearly 100 db at one frequency  would require a lot of 
> bypassing and shielding.

Yep.

Of course the big unknowns are what the coupling and Q really are, as well as 
how much gain the tubes actually deliver. 

One thing is certain: In a receiver like the S-38 or my old Heath AR-2, with 
no RF stage and only one IF stage, every stage will be designed for max gain.

> 
> The point of my post was only to show that we have control over the gain of  
> 
> an IF stage by choosing the C and L of the transformer (and Q, too).   
> Reducing the gain by increasing the capacitance is a step in  the direction 
> of 
> increased stability.  Any time we lower the plate  load impedance stability 
> goes up, 
> and of course some gain is sacrificed.   

There's another factor: Gain control. With more IF stages, a small variation 
in gain control voltage can have big effects on the gain of the IF strip. This 
permits a "flatter" AGC curve.  

One somewhat extreme example of this 
> 
> technique is in the older R-389 and R-390  Collins receivers which used 1000 
> 
> pF capacitors in the IF transformers along  with resistors (100K on the 
> primary and 470K on the secondary) to control  the Q.  These receivers had 6 
> IF 
> stages at 455 KHz!  The versions I am  talking about are those before the 
> R-390A 
> which used mechanical filters for  selectivity.
> 

They had all those stages because the selectivity came from the IFTs. IIRC, 
the widest setting was 16 kc! 

100K on the primary and 470K on the secondary isn't a lot of loading. Figure 
out the effective RL from an inductor that resonates with 1000 pf and it's 
probably a lot lower. 

A very interseting subject! The Ancient Ones struggled with IF stability, 
having to use triodes. Some of their designs were neutralized! 

Fun fact: Some of the earliest superhets, like the famed Paragon used by 
Godley at Ardrossan in 1921, did not use IFTs at all. They were resistance 
coupled!

73 de Jim, N2EY