Switching to a Mechanical Filter from 1967 for my HRO-ish Receiver (with video)

From RSGB Handbook 1982

Having overcome the difficulties with the National NPW Dial and Gearbox, I turned my attention to the 455 kHz filter.  I had been using this old Toyo CM - 455 kc filter (Date stamped August 1969).  CM stands for "Crystal-Mechanical."  These filters are hybrid with some of the features of a crystal filter and some of the features of a mechanical filter. For more details go here: 
https://soldersmoke.blogspot.com/2016/04/the-meaning-of-cm-in-toyo-cm-455-filter.html

I was disappointed by the CM filter.  It seemed very lossey, and it just didn't seem to be of sufficiently high Q -- it seemed very broad. I could hear the other side of zero beat.  It was barely a "single signal" receiver,  and being "single signal" is the whole point of a superhet. 

I remembered that Pete Juliano had sent me a Japanese-made 455 kc mechanical filter. Maybe this would do better.  Last night I did a quick comparison test and -- wow -- Pete's filter was much better.  The Fifth Edition of the RSGB Handbook seems to agree with my assessment, noting that mechanical resonator filters were superior to the Crystal Mechanical hybrids (see pages 4.17 and 4.18) 

Pete's filter is from the Kokusai Electric Company.   Part# MF 455 ZL.  (Date stamped May 1967). "ZL"indicates lower sideband.   I checked and indeed the passband goes from just above 452 kc up to about 454.5 kc.  This is a 40 meter receiver and SSB on 40 is LSB, so this filter would work perfectly right?  Not so fast!  Sideband inversion had to be considered.  

I was running my VFO from about 7455 to 7755 kHz.  This means that the modulated incoming signal would be SUBTRACTED FROM the VFO signal to get to  the 455 kHZ IF.  And when that knd of subtraction happens, we have sideband inversion.  The LSB signal will look like a USB signal when it reaches the filter. 

My BFO was running right at 455 kHz, using a ceramic resonator at that frequency. I briefly considered just shifting it down to 452 kHz, but this proved to be difficult.  Then I got a better idea. 

I could just shift the VFO down to 6545 to 6845 kHz.  This would mean that the VFO frequency would be subtracted from the incoming modulated frequency.  There would be no sideband inversion.  I had been thinking about doing this frequency shift anyway, thinking that VFO stability gets better as you go lower in frequency. 

REMEMBER THE RULE:  If you are subtracting the modulated (signal) frequency from the frequency of the local oscillator or VFO, only then will you have sideband inversion.  See: 
https://soldersmoke.blogspot.com/2015/05/sideband-inversion.html

Moving the VFO was easy. I am using a variable capacitor with several variable caps on the same rotor.  I just moved from the smallest variable cap to the middle variable cap -- this added capacitance to the system and lowered the frequency.   I also added three additional turns on the coil.  This put me very close to where I needed the frequency to be.  I added one additional 9 pf cap and this put the VFO freq right where I wanted it. 

I was really  glad to include Pete's filter in this receiver.   The mechanical resonator technology fits very well with the very mechanical old-tech theme of this project (it already had a gearbox -- a mechanical filter seemed to fit right in).   It is a fascinating device -- it is almost like having a set of tuning forks all tuned to 455 kc (see above for the RSGB description of how it works).    And having it from from Pete adds a TREMENDOUS amount of mojo, juju, and soul to the new machine. 

Icing on the cake: As I type this, I am listening to Fred K3ZO converse in Spanish with hams all through South America.  Fred preceded me by three decades at the U.S. Embassy in Santo Domingo, but when I got there the local hams were still talking about him  -- he was much loved and admired by the Dominican hams.  TRGHS. See Fred's story here (scroll down a bit): http://www.gadgeteer.us/DRDISP.HTM