[R-390] Stagger tuning 390A IFs (again)

[2002tii]

Some time ago, Charles wrote:

>In the 390A, the mechanical filters provide the actual IF bandwidth
>response (the interstage LC filters do clean up the stop band, where
>the mechanical filters have zeroes and ripple, so they give the IF a
>lower ultimate stop band attenuation than it would have had with the
>mechanical filters alone).  Think of the interstage LC filters in the
>390A as a roofing filter, although in this case they follow rather
>than precede the mechanical filters that actually set the IF bandwidth.
>
>In some cases, the LC "roofing" filter in a 390A may have a narrower
>or more peaked response than the 16 kHz mechanical filter.  If that
>is a worry, the interstage LC filters can be stagger tuned to broaden
>the overall LC response so that even the 16 kHz mechanical filter
>effectively sets the IF response.
>
>I'm not aware of there being a version of the 390A IF that was
>designed to be stagger tuned, but that's not to say there wasn't
>one.  *   *   *  The only real benefit of stagger tuning a 390 IF is
>to broaden the 16 kHz response [if it would otherwise be narrower
>than the 16 kHz mechanical filter]


Some 390As were apparently shipped with stagger-tuned IFs, and in 
1985 the Space & Naval Warfare System Command technical writers 
specified that those IFs should be stagger-tuned if they needed 
adjustment (i.e., when an IF transformer was replaced).  However, it 
is not entirely clear why these radios were stagger-tuned, and the 
information in the tech manuals raises some further questions.

The Fowler manual (SPAWAR 0913-LP-009-1400, 15 May 1985) says the following:

"3.2.10.2  Adjustment of transformers T501, T502, and T503 is 
normally not included in the IF amplifier alignment procedure.  They 
are initially tuned during receiver assembly, and should require no 
subsequent adjustment.  The bandwidth of these transformers is 
sufficiently wide to have negligible effect within the bandpass of 
even the 16 kHz mechanical filter.  Their most important function is 
that of providing attenuation of IF signals more than 8 kHz removed 
from 455 kHz."

So, as a general matter, the 390A IF transformers are wide enough 
that they do not narrow the IF response below the width of the 16 kHz 
mechanical filter, but they do help with stopband attenuation.  They 
should be left alone, all peaked at 455 kHz.

However, on some particular later chassis, they were stagger-tuned 
from the factory:

3.2.10.2 (continued):  "In the IF subchassis with MOD numbers 1 and 
higher on Order No. 363-Phila-54, serial numbers 600 and higher on 
Order No. 08719-Phila-55, and all IF subchassis on Order No. 
14-Phila-56, transformers T501, T502, and T503 are stagger-tuned to 
increase bandwidth.  When one of these transformers is replaced in 
any subchassis, stagger-tuning procedures should be followed 
(paragraph 6.2.7.1)."  [Paragraph 6.2.7.1 details a procedure for 
peaking the T501 secondary and T502 primary at 467 kHz, the T501 
primary and T502 secondary at 443 kHz, and the T503 primary and 
secondary at 455 kHz.]

These two sections of Paragraph 3.2.10.2 leave us with a mystery -- 
why were those particular IFs stagger-tuned, and why should "any" IF 
chassis with at least one replacement IF transformer be 
stagger-tuned, but NOT "any" IF chassis just being aligned if no IF 
transformers have been replaced?

One possibility is that the IF transformers used in those chassis, 
and all replacement IF transformers, had higher Q (and thus narrower 
bandwidth) than the transformers used in other chassis.  But if that 
were the case, one would expect the manual to advise replacing T501, 
T502, and T503 as a set in any older (straight-tuned) chassis that 
required at least one replacement, and stagger-tuning them.  Note 
also that the stagger-tuned IFs are tuned to 455 kHz and to 455 +/- 
12 kHz.  Even if the transformers were higher-Q, with this wide 
staggering the aggregate response would be considerably wider than 
the original non-stagger-tuned response, and some of the reduction in 
close-in stopband signals would be lost.  One would expect the 
staggering with higher-Q transformers to be more like +/- 4 kHz.

So, why else might the 1985 tech writers think that any IF chassis 
with at least one replacement IF transformer should be 
stagger-tuned?  It could be that all IF transformers are nominally 
the same with respect to Q/shape, and they considered stagger-tuning 
to provide a better overall response.  But if so, why didn't they 
instruct techs to stagger-tune ALL radios when they were being aligned?

Here's my guess (but note that it is just that -- a guess):  All 390A 
IF transformers are nominally the same.  The 1985 tech writers did 
think that stagger-tuning provided a better overall response, 
probably because it flattened the 16 KHz passband slightly (not the 
~1-2 dB ripple from the mechanical filters, but a gently rounded 
overall shape).  But the tuning procedure is difficult (you need to 
change out the shield cans with test cans that have holes for tuning 
tools, then change back to the originals), so they didn't consider it 
justified unless you had to tune one of them anyway because you replaced it.

If this is correct, it might lead one to believe that stagger-tuning 
the 390A IF is "better" and should be done by any fanatical 390A 
owner as a matter of course as part of an IF alignment.

I disagree.  IF this supposition is correct and all IF transformers 
are nominally the same with respect to Q/shape, then the issue 
becomes which is more important -- flattening the rounded top of the 
16 kHz response, or having the best close-in stopband 
performance?  Since the passband has serious ripple in any case due 
to the mechanical filter, I submit that the gentle rounding from the 
LC peak is the least of our worries.  Indeed, for almost any 
practical use today, the small rolloff away from the IF center 
frequency will actually be an advantage, because a 16 kHz passband is 
just too wide for any application in the current RF environment.  To 
the extent that stagger-tuning may slightly improve the phase/group 
delay response of the IF, any problems in that area are swamped by 
the horrible phase response of the mechanical filters, which is at 
least 1000x worse, so this is no justification for 
stagger-tuning.  Finally, improving the close-in stopband performance 
is actually quite important, so the benefits gained by sacrificing it 
if you stagger-tune the IF transformers would have to be substantial 
to justify doing it.  And as explained just above, not only are there 
no substantial benefits to stagger-tuning in this instance, there are 
actually additional detriments.

So, I believe the best course of action is:  (1) Leave the IF 
transformers alone unless you have a damn good reason to think they 
need tuning, and (2) If you do see the need for retuning, peak all of 
them at 455 kHz.  And (3) if you do feel compelled by the ghost of 
Art Collins to stagger-tune them, use something like +/- 4 or 5 kHz 
staggering (455/451/459 or 455/450/460) rather than 12 kHz.

This, I believe, is certainly true for radios that were originally 
NOT stagger-tuned, and is very likely true for radios that WERE 
originally stagger-tuned, because I believe it is very likely that 
the IF transformers are the same in both cases.  However, I think 
rule (1) is paramount -- if you have a radio that was stagger-tuned 
from the factory, and no damn good reason to think its IF 
transformers need tuning, I would not change it.

It would be interesting to know if the serial number "breaks" in the 
three contracts, mentioned in Paragraph 3.2.10.2, represent one date 
on which the change was made.  Also, this seems like an especially 
noteworthy change and one might expect there to have been an ECO or 
other documentation -- but I'm not aware of any.  I'll leave these 
questions to the 390A historians among us.

Best regards,

Don



Copyright (c) 2013.  Not for redistribution