[Hammarlund] Re: Phasing control null on crystal filter for SP-600

[David Wise]

Andy Moorer writes:

>L37 is always a problem to align. It is a very broad peak at the best of 
>times. The ONLY way to align it is exactly the procedure in the manual - 
>including the AC voltmeter on the audio output. You have to put the 
>bandwidth select to 1.4 (the most narrow non-xtal setting) and then you can 
>sorta get a peak.

Small correction: 1.4 is the widest xtal setting, not the
narrowest non-xtal setting.  Selecting 1.4 is in the procedure.

I was tweaking L37 the other night, and I noticed that
with a 400Hz modulated carrier, the broad peak had pure
400Hz audio at the center, with 800Hz harmonic on either
side.  (I ignored the S-meter; it didn't correlate with
any of this.)  Maybe that pure spot is the target.
At least it's sharper than the peak.

Tektronix and HP used a similar crystal filter arrangement
in some of their older spectrum analyzers, using a crystal
as the upper leg of a voltage divider.  At series resonance,
a crystal's impedance is minimum.  It rises sharply on either
side.  You can affect the bandwidth of the overall circuit
by adjusting the load placed on the crystal.  For reasons
that escape me, Hammarlund and Tek used a parallel-resonant
LC circuit for the load.  (HP used a resistor.)  A low Q tank
is a heavy load, and the crystal's own resonance dominates,
giving a narrow bandwidth.  A high Q tank is a light load -
even when it's slightly off resonance, because it's broader
than the crystal.

Back to the SP-600.  As you go from 1.4 to .5 to .2,
the LC tank is spoiled to a lower and lower Q.
If it's tuned exactly to the crystal series peak,
you get a symmetrical response.  If it's mistuned,
well, you don't, and I can't figure it out any further,
nor explain how the phasing control does its thing.

Hammarlund changed the spoiler resistors in Issue 2.
They didn't change the crystal part number or anything
else around the filter, so the bandwidths must have
been wrong in Issue 1.

Dave Wise
SWL in Portland Oregon
R-390A '56 Motorola
SP-600 R-274A (Issue 2)
HQ-100
SX-96
FRG-7
Tektronix 535/535A/545/547 o-scopes
HP 141T spec-an
GR 1001A
HP 8640B
Etc