[R-390] RF deck alignment

[Roger Ruszkowski]

Tisha,

  
I know that this generally falls under the "leave it alone" category but
has anyone ever gone through the process of aligning the RF deck stages and
bands for linearity?

Generally I know that we pick a few spots on each band and peak through the
RF stages for performance but what happens when a slug has been replaced,
repaired or the spring was accidentally stretched? That one slug will have
a different peak than the other in the string when the band selector switch
is lined up that way.

The only way I could think to do it would be to inject a known RF signal
level and to use a RF microvoltmeter (like a Boonton 92) at test points to
walk through the stages.

Am I wrong-headed to think that the more selective the RF stages are the
better the desired response would be at the IF?


----------------------------------------------------------------------

Tisha,

Mostly this is a moving target. You could sweep the tuned section of interest with a
sweep generator, detector and scope.
As the mechanical tuning sets still we would see a peak on the scope display where
the tuned circuit was currently tuned to and the band pass skirts of the tuning as it sets.

While the sweep generator is running, we could then run the mechanical tuning through
its range and watch the peak move along the scope display. We could watch the peak point
for good level across the tuning range. We could watch the band skirts for symmetry. We could watch that
the peak point matched the receivers dial read out to see that the mechanical track was in line with the electrical track.

A sweep generator is a bit more graphic but you could use a signal generator and a voltmeter at the test points.

A real common practice was to use the signal generator and the diode load voltage. But rather than just
look at the given points in the TM we would look at a bunch of points. On the 16 - 32 octave we would check each
megahertz band as the crystals changed.

You may want to peak the 2 - 4 octave for best output across 3.5 - 4.0 Mhz and the 4 - 8 octave for best output
across 7.0  - 7.3 Mhz. Maybe you would like to peak the 8 - 16 across one of the international software bands.

We had some receivers that were used on fixed frequencies. These got special treatment to peak performance
on those frequencies. You will be amazed at how badly you can pull the band pass out of shape.

Consider that while you may not gain much in peak in band performance, how much more out of band stuff
you can filter out of the mix to reduce the noise floor also counts. The peak signal may not gain but if you
can take out the software AM above 7.2 at night you may hear a whole lot more CW at 7.0 for DX.

There is rhyme and reason for taking on these exercises you suggest. You can take your general coverage
receiver and optimize it for some specific select frequencies with just your spline tool and flat blade non conductive
screw driver (tweeker). Conversely you can get a whole lot better flatter general band pass by testing more than the minimum points
presented in the TM.

Some simple math shows the band points in the Tm are not the best exact corners. they are set on the 100Khz cal tone
points so you can use the cal tones to get close in alignment. We know there are better frequency values to use on each
band to get a better flatter total performance from the receiver.

Once you get the signals up out of the noise to a good level, more work does not get more signals and aggravation begins
to exceed return on investment. But until you get to that point all work is productive.

Roger AI4NI