[HBR] HBR-11/2000. Comments? -- And GC-HBR progress

[N2EY at aol.com]

In a message dated 7/14/06 9:52:40 PM Eastern Daylight Time, 
waltah at earthlink.net writes:


> The most interesting thing I found 
> looking at the May '65 Goodman QST article ("Some thoughts on 
> home receiver design") is his note that he too dealt with the 
> problem of BFO pickup in the IF amp input.  That's a problem 
> because you can't get proper S-meter action and if the problem's 
> bad enough you lose dynamic range too.  

Even a little leakage can be a dynamic-range problem. 
> 
> The way I've been dealing with that in recent designs is by very 
> careful layout, sheilding, and filtering, AND using a bridge circuit 
> copied from the Tempo-ONE transceiver that feeds some voltage 
> from the BFO into the detector with reversed phase to cancel out 
> the unwanted stuff.  Goodman's approach is cleaner in some ways  
> -- he does another conversion immediately ahead of a no-gain 
> detector so the BFO isn't at the IF.   The price is one more stage 
> and one more oscillator.

Yep - and there are more benefits:

1) If you're trying to get a lot of gain between the mixer and the detector, 
as is needed in a no-RF-stage receiver like W1DX's "Miser's Dream", simply 
adding more IF stages can be impractical. A mixer at the end of the IF chain can 
have gain without adding to the stability problem.

2) Some detector circuits can pull a tunable BFO, particularly because the 
BFO freq is so close to the signal freq. An xtal BFO can resist pulling, but it 
will require multiple crystals of precise frequency for LSB, USB, CW, FSK, 
etc. If the BFO is xtal controlled and the last conversion oscillator is 
variable, you get the effect of a variable BFO. 

3) If you have filter(s) for one IF, but no BFO xtals for those filters, a 
last conversion may let you use xtals you do have.   

4) Swapping the conversion oscillator from high side to low side can invert 
the sideband.

5) Putting a second xtal filter at the end of the IF chain can clean up 
things nicely, and improve shape factor. A last conversion lets you use filters of 
different frequency.

> If all the gain is at one frequency you have to do the filtering, 
> sheilding and all that anyway, or you will never get the IF passband 
> right.  The tiniest bit of feedback causes it to be terminally screwed 
> up.   
> 
> Hmmmm ...
> 

Yep.

73 de Jim, N2EY