[Hallicrafters] SX-28 Dial Tracking Issue

[Jim DiMauro]

Hi All:

I recently finished restoring a 1941 vintage SX-28. The rebuild included all
paper caps, a few micas that were out of spec and all out of spec resistors.
I proceeded with the IF alignment, which seemed to respond well to the
procedure in the manual, and then I did the RF alignment. That¹s where
things didn¹t seem quite right.  I set the low frequency and high frequency
alignment points on each band, and the main tuning dial lines up perfectly
at each of the low and high end alignment points. The problem is that the
frequency accuracy of the dial degrades as I move from the end points to the
center of the band, with the error being greatest at the center. I don¹t
expect a radio of this type to track like a Collins PTO would, but the
tracking error is excessive when compared to other radios that I restored,
like my NC-200 and HQ-129-X, both of which are nearly spot on from one end
of each band to the other.


I took detailed measurements that show the magnitude of the error on the
SX-28. Here are sample measurements from each band: Band 1 aligns perfectly
at the alignment end points of 600 kc and 1400 kc, but the approximate band
center shows a 20 kc error (i.e. an 800 kc signal reads 780 kc on the dial);
Band 2 shows a 50 kc error at 2000 kc; Band 3 has an 80 kc error at 4000 kc;
Band 4 is off by 200 kc at 7700 kc; Band 5 is off by 300 kc at 14000 kc; and
Band 6 is off by 500 kc at 28000 kc. Again, in all these cases the dial
reading agrees with the signal frequency at the high and low end alignment
points, and in all cases the dial reading is lower than the actual
frequency.


Interestingly, the error in terms of the physical movement of the main
tuning capacitor seems to be consistent among all of the bands at the band
centers. To clarify, the difference in the main tuning dial logging scale
(the metal disk on the main tuning knob) between the actual and dial
frequency at the center is about 30 points on all bands. If it was due to a
problem with the gear train I would not expect the end points to remain
accurate after repeated tuning up and down the band. I took care to ensure
that I didn¹t align on image frequencies. The band spread dial is installed
correctly and I did the alignment with the band spread capacitor set for
minimum capacitance (100 on the logging scale). There are no obvious
physical issues with the main of bandspread tuning caps, and there is no
evidence that the capacitor plates had ever been ³knifed.² Any ideas would
be appreciated. The radio looks and sounds great, but the magnitude of the
dial error is driving an OCD person like me nuts.

73,
Jim
WA2MER