[Yaesu] Mechanical vs. Crystal filters

[Walt Hutchens]

I'm not a guru either but I'll take a stab anyway, and maybe one of 
the gurus will correct any errors.

Crystals have extremely high Q's.   All you've got is a quartz 
crystal displaying piezo-electric properties -- a crystal lattice and 
an electrostatic field.  Where are the losses?   Some in the plated 
on electrodes and the mounting, some, doubtless in whatever gas 
surrounds the crystal ... TINY.  

The elements of a mechanical filter are good but not THAT good.  
You've got those coils ... UGH -- coils have resistance, meaning 
losses.  And the filter elements themselves aren't a single crystal --
 there's bound to be some loss in a vibrating polycrystaline solid.  

So for a given number of elements, the crystal filter should have 
both lower losses and a better aspect ratio.   Plus the many possible 
modes of vibration that can be excited in a crystal allows their use 
in a filter at any practical frequency.   Mechanical filters are much 
more limited, frequency-wise.   

So why mechanical filters at all?   Well, at the time Collins made 
them popular, most radios achieved selectivity in an IF at under 1 
Mcs -- typically 455 kcs.  I believe that it was cheaper to build a 
five or six element mechanical filter for these frequencies than make 
the five or six crystals necessary for the crystal filter.   Plus, at 
least some parts of the mechanical filter could be patented, where 
crystal filters using individual crystals were in the public domain. 

Today, the situation is probably somewhat different.   Crystal 
filters can be built on a single crystal and the process is highly 
automated.   Mechanical filters are still a hand assembled item.  
However, 'Collins mechanical filter' has something of a cachet.

What I fail to understand is why any modern radio has a 455 kcs IF.  
Could it be just the fact that every frequency determining part you 
can imagine is made for that IF, so parts are cheap?

I guess there are more tidy frequency conversion schemes available if 
the final IF (that is, the largest signals in the receiver) is well 
below the lowest frequency at the antenna.  But all those mixers ... 
UGH.  

Walt Hutchens
KJ4KV