[R-390] Stagger Tuning and IF craziness

[2002tii]

Tisha wrote:

>It turns out that the mechanical filter coils have a bandwidth of right
>around 20 kHz (awful convenient) but it also explains a discrepancy I found
>too where some filters use something other than #41 wire. By tweaking
>things I was able to get the bandwidth of the mechanical filter down to
>12.85 kHz with a different wire gauge and turns ratio (with less parallel
>capacitance to boot).

The response of the driver assembly (coil + magnetostrictive wire) 
has next to nothing to do with the ultimate filter response -- that 
is set by the vibration of the disks.  The driver assembly should be 
flat to substantially greater than the filter BW, to ensure a flat 
mechanical drive impedance to the disk resonator assembly.

>Each stage of the IF deck has a bandwidth and stagger tuning allows you
>to flatten out the bandwidth response to where the wider mechanical filters
>are not clipped off at the edges.

But, as stated in the technical manual and as many of us have 
documented, even when peaked at 455 kHz the IF transformers are 
plenty wide not to interfere with the transition zone of the 16 kHz 
filter (there are several sets of IF response curves on the web, 
taken through a complete IF assembly).  The only possible benefit of 
stagger tuning would be to remove whatever very slight general 
rounding there may be in the middle of the 16 KHz passband, and as I 
stated previously: (1) any such rounding would be a blessing, not a 
problem, in today's band conditions, and (2) the amount of any 
rounding present would be dwarfed by the large passband ripple due to 
the mechanical filter.  (Again, look at the IF response curves on the web.)

>The traditional response is to set everything to 455 kHz and peak away. If
>you do not care about the wideband response at 16 kHz you could either
>stagger tune to narrower values (like was mentioned setting them all the
>same and ending up with maybe 4 kHz of BW, it sort of invalidates the
>reason to have the 8 and 16 kHz mechanical filters) or you could still
>leave them stagger tuned but set them closer.

I believe you misunderstand.  The three tuned IF transformers, all 
peaked at 455 KHz, have an aggregate response that is already wider 
than 16 kHz and does not interfere with the transition zone of the 16 
kHz filter.  Stagger tuning can flatten the gentle peaking the 
transformer stages may exhibit *within* the passband of the 16 kHz 
filter, at the expense of widening their response to much wider than 
16 kHz and thereby sacrificing the transformer stages' ability to add 
very useful close-in stopband attenuation (where the mechanical 
filter exhibits zeroes and a re-entrant response).  Some quick 
figuring suggests that if they are stagger-tuned "by the book" at +/- 
12 kHz, the response of the transformer stages at the -6 dB points 
would increase from about 20 kHz to more than 50 kHz.  As Charles 
said in the post I quoted, think of the transformer stages as a 
"roofing filter."  Not a good idea to widen your roofing filter from 
20 kHz to over 50 kHz.  If someone insists on stagger-tuning, by 
bringing the offsets down to +/- 4 or 5 kHz they will at least 
minimize the damage and end up with a "roofing" response in the 30 
kHz area rather than > 50 kHz.

>I suspect the early idea of "leave these settings alone" was based upon
>what would be the perceived difficulties of field aligning the IF's.

No, it was for the reason stated in the manual -- they are low-Q 
circuits that can generally be relied on not to drift out of spec 
(they fail from time to time, but if not failed they will be good to 
go).  The procedure is given in the manual, and the only thing 
difficult about it is swapping the "test" IF housing in and out.  I'm 
sure Roger can vouch that the military would not hesitate to ask 
radiomen to do that (and much more).  As an aside, the many, many IF 
stages in the Marantz 10B FM tuner -- arguably the best FM tuner ever 
made, and the precision linear-phase IF circuits were largely the 
reason -- were "tuned for life," and I have never yet seen one that 
wasn't spot on if it hadn't been monkeyed with.

>the easy way is to tune to a known signal, put a voltmeter on the
>diode lead and adjust for maximum value. If they are stagger tuned at the
>factory then by letting the field folk do the adjustments might mean that a
>receiver had hot performance but would not hold a lock on a RTTY signal.

Again, I believe you misunderstand.  Putting a voltmeter on the diode 
load and adjusting for maximum value IS EXACTLY how those stages are 
stagger-tuned.  See the manual procedure cited in my earlier 
post.  The only difference is you don't do all of the sections at 455 
kHz, you do 2 at 455, 2 at a + delta, and 2 at a - delta.  As 
mentioned, the "book" delta is 12 kHz, but I've argued that -- if you 
feel you absolutely must stagger-tune the transformer stages -- a 
delta of 4 or 5 kHz would be more appropriate.  If you do this, pay 
close attention to which transformer windings get tuned to each of 
the three frequencies (center, + delta, and - delta).

Best regards,

Don


Copyright (c) 2013.  Not for redistribution