[Boatanchors] "Sweep" techniques

[Richard Knoppow]

----- Original Message ----- 
From: <bonddaleena at aol.com>
To: <boatanchors at mailman.qth.net>
Sent: Friday, January 07, 2011 12:09 PM
Subject: [Boatanchors] "Sweep" techniques


>
>
>
> Hello to all. I've been a ham for a long time and there is 
> one area where not only is my knowledge severely lacking, 
> but I seem to be unable to find any decent info on this 
> technique.
>
> The subject is 'sweep alignment' of IFs and filters, etc.
> I have quite a few Tek 'scopes, and also a nice collection 
> of MODERN Sweep/Function Generators. Krohn-Hite, Exact, 
> etc. I have also purchased and repaired a ton of test 
> equipment including 'scopes and TDRs, in addition to my 
> boatanchor collection.
> The manuals for the Sweepers assume you know how to 
> connect them and use them.
>
> A while back, I noticed that there were several articles 
> in ER Magazine showing the swept response of various 
> boatanchors, (SX-117, etc). Unfortunately, the authors 
> never describe the test methodology. I have purchased more 
> than a few (!!) "How to use your Sweep/Signal Generator" 
> books.
> However, they are pretty old and only deal with old, tube 
> type TVs.
> I even suggested to Ray at ER Mag that this subject would 
> make a great article and he agreed, but I still haven't 
> seen anything.
>
> So, could some of you guys with all the expertise in this 
> area (Carl, Al) at least give me a shove in the right 
> direction?
>
> Yes, I have searched the WWW extensively and no joy......
>
> Thank you!
>
> ron
> N4UE

     In old literature this is called _visual alignment_. 
Its simple but needs some precautions. The earliest sweepers 
were simpy generators that could be frequency modulated over 
a range wide enough to cover the IF bandwidth. They were 
sometimes called wobbulators because the frequency was 
wobbled back and forth using a sine wave.
     To use the sweep technique you need:
1, A sweep generator or wobbulator
2, A display device, an oscilloscope works fine.
3, Some way of calibrating the displayed frequencies. When a 
well calibrated generator and scope are used you probably 
don't need anything else but most sweep generators had some 
sort of marker generator in them. When aligning the IF of an 
ordinary communications receiver with a single crystal 
filter the filter itself is used to mark the center 
frequency. A second, fixed frequency, generator can be used 
to make a beat at some desired frequency, either the center 
or to find the edges.
4, Its helpful, not not necessary to have some sort of 
linear to log voltage converter for the vertical display. 
This is helpful where you want to investigate the skirt 
selectivity far down the slope but is not necessary for 
normal alignment.

     The technique is simple. Just as when using a fixed 
frequency generator the sweeper can be used at either the IF 
frequency with direct injection at some point that will not 
disturb the tuning, or at an RF frquency tunable by the 
receiver. If the latter both receiver and generator must be 
stable enough to remain essentially on frequency during the 
alignment.
     The signal to the scope must be DC so should be picked 
off at the detector load resistor.
     Alignment is done wiith the AVC off and signal level 
into the IF at the lowest level that will overcome noise and 
give you a good display. The reason is that the tuning of IF 
and RF circuits can be affected a little by bias and you 
want the best selectivity of both IF and RF at minimum 
signal levels.
     The signal generator should be set to the center 
frequency without sweeping and the detected signal fed to 
the scope and adjusted for a convenient level. Remember, 
that unless the IF is already well adjuted the display level 
is likely to increase as its adjusted so leave some room for 
it.
     Before begining to sweep you may want to adjust the IF 
for maximum on the fixed signal to get it approximately 
correct.
    Now set the sweeper for a range that will cover the 
expected bandwidth of the IF plus a bit. The sweep rate 
should be slow enough so that the pattern is not distorted. 
Remember that all filters store energy, that's how they 
work. If swept too fast the filter will continue to put 
something out after the signal has gone past it and the 
detected pattern will be distorted. This is why narrow band 
crystal filters "ring". Usually this distortion is easy to 
see in the pattern.
     Next begin the adjustment procedure. If visual 
alignment is specified, as it is for instance for the 
Hammarlund HQ-129-X, follow the book instructions. Otherwise 
begin by setting the center frequency. In most receivers 
with single crystal filters the actual center frequency will 
be determined by the crystal resonance at minimum bandwidth 
(but not in the Collins 51J series where it is always set 
for exactly 500khz regardless of the crystal).
     Once the center is set begin to sweep slowly and note 
the pattern. It is usual to begin with the IF adjustments 
nearest the detector but some receivers specify the 
opposite. If you know the IF is either critically coupled or 
undercoupled there will be a single peak and the procedure 
is t tune each stage for the greatest amplitude at the 
center and a generally symmetrical peak. For overcoupled 
transformers, like some in the HQ-129-X or RCA 88-series, 
the adjustments are done to get a somewhat flat topped 
pattern centered on the center frequency. If sufficiently 
overcoupled the stage may present a double hump which should 
be symmetrical. In most receivers using overcoupled stages 
only some of them are overcoupled, the others being 
critically coupled so that the combination is flat across 
the top. Again, the HQ-129-X is like this and its IF is 
adjusted in steps, the output being mesured at the output of 
the stage rather than at the detector. In general the 
adjustment is for the correct shape at the maximum 
amplitude. IFs with overcoupled stages in general will have 
somewhat steeper skirt selectivity than those with 
critically or under coupled transformers.
    Some receivers such as the Collins 75A-2 can be adjusted 
for either critical or over coupling so that the bandwidth 
can be varied, again, when adjusted for overcoupling the 6db 
down bandwidth will be greater but the skirt bandwidth will 
be sharper.
    In most receivers with simple single crystal filters the 
crystal filter is adjusted last. It may require reducing the 
sweep width and rate to do it right. In late design filters 
which used the Hammarlund patent there are two adjustments: 
one is for peaking the entire filter at the IF frequency, 
the other adjusts the reactive loading on the crystal and is 
adjusted for _maximum_ bandwidth when the crystal is set for 
maximum. The sweeper will also show where the nulls are. 
When the phasing control is set at the center the nulls 
should not be visible. As its move toward the ends the null 
will appear at one end of the band and move toward the 
center as the control is moved toward the end limit. On the 
other side it should start at the other bandwidth limit and 
also move toward the center. At mid positions of the phasing 
control the null should be about symmetrically placed on 
each side of the center frequency. If not make sure that the 
knob on the phasing control is actually set to indicate the 
phasing capacitor position accurately. This can be done 
visually but you really don't have to see the capacitor, you 
can judge the edge positions from the sound. If you the knob 
is correct and you are still not getting a symmetrical 
pattern from the crystal adjut the bandwidth adjustment a 
little and see if it improves. The correct setting is the 
one that gives you the best pattern. In some cases the 
filter will have a balance trimmer inside. In the HQ-129-X 
its a pigtail of wire, in the Super-Pro its a compression 
trimmer. It may need some adjustment of this trimmer to get 
the pattern symmetrical with the phasing control centered.
     Earlier receiers like the older HRO and Hallicrafters 
receivers with crystal filters use the original Lamb version 
which, in general, will not change bandwidth or phasing 
without moving the center frequency. These are much harder 
to adjust because of the interaction but generally one can 
see where they are peaked and set for best response and 
nulling. However, the center frequency and gain will change 
with the bandwidth and phasing controlls no matter how well 
adjusted.
     There is not much else. The method will also allow  you 
to investigate the variation in selectivity patterns with RF 
gain due to detuning. There will be some but in a well 
designed receiver it will be minimum.
     Another note: Some receives with variable IF bandwidth 
(not crystal but regular) use a form of coupling that 
results in the band increasing on one direction only, i.e., 
the effective center frequency changes. Some hallicrafters 
receivers are like that. In any case, where If bandwidth is 
varied by changing mutual coupling in the transformers, as 
it is in the Super-Pro series including the SP-600-JX and 
some others, the alignment should be done at the minimum 
non-crystal bandwidth where the transformers are either 
under or critically coupled and then the pattern checked at 
wider bandwidths where small tweaks may improve its symmetry 
a little.
     Occasionally you will encounter a receiver who's RF 
selectivity is narrow enough to affect the IF pattern if you 
use an RF signal. This can be the case at the low end of the 
broadcast band in receivers where the RF stages are not 
broadened out but will not usually be a problem above the 
broadcast band. Most communications receivers that include 
the broadcast band have some means of broadening out the RF 
on it and also usually restrict the sensitivity since signal 
strength is usually very high as is noise level.
    I hope this is helpful not confusing:-)


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com