38S: OUTPUT MOD for stock 38S

---------

From: Paul Harden (pharden@aoc.nrao.edu)
Date: Sun Jan 26 1997 - 16:49:56 EST


NorCal 38-Special - stock version
OUTPUT FILTER MOD

This is the result of some careful bench testing of the 38S, mostly for
the IRF510 mod. The "stock" output filter for the 38S is a very good
filter. The virtual square-wave from the HC240 is causing some
spurious power, most of it BELOW 10MHz. Most of it is more than 30dB
below the fundamental (-30dBc), but can cause a slightly elevated output
power reading on most power meters. This simple mod knocks down these
low frequency spurs considerably and boosts the output power of the
fundamental a bit (about 1.5dB or nearly 100mW). This mod is not
MANDATORY; the stock 38S is well within FCC compliance. This mod will
boost the output power a bit, attenuates the low-frequency spurs
(which produces the raspy sound when mistuned), and seems to make adjusting
TC2 a little more distinct.

For clarification, below is the output filter schematic. I am designating
the optional capacitor for the filter input as C501, as that is how it
is labeled on the 38S PCB (right next to L3). It is shown as C505 on
the IRF510 mod. C501 and C505 is the same capacitor. The space for it
is already on the board, although no component should currently be there.
   

OUTPUT FILTER SCHEMATIC

                     -----> to receiver
                     |
                     |
    From C27 | L3 L4
    HC240 ---- .1 ---*--- T37-2 ---*--- T37-2 ---*----> ANTENNA
    PA | | |
                     | | |
                   C501 C28 C29
                   270pF 820pF 560pF
                     | | |
                     | | |
                    gnd gnd gnd

MODIFICATION:
1. Add a 270pF capacitor at C501 (next to L3 on the 38S board).
   Any value from 200-270pF does a good job; 270pF is best.
   Above 300pF, output power is reduced while the overall drain
   current increases on transmit and the low frequency spurs are
   smoothed into a big power peak around 4MHz.
2. Change C26 from 220pF to 330pF. C26 is on the pin 10-11 end
   of U4, the HC240. It is probably easier to leave C26 alone, and
   add a 100pF cap across C26 on the solder side for a total of
   320pF. The Xc of 300pF is 50 ohms; this gives the reflected
   power from the output filter and HC240 drivers a 50 ohm termination,
   which cleans up the output purity. C26 is the capacitor from the
   single HC240 driver, pin 8, to ground.
3. Readjust TC2 for a compromise between good output power and a
   pure tone. This mod makes the area of the pure tone a bit smaller
   (20-30 degrees of TC2 rotation), but also much more distinct and
   easier to set. (With the TicK keyer mod, you are not able to
   hear the direct sidetone on keydown; you'll have to adjust TC2
   for best sounding output one another receiver).

SOME DISCUSSION
The output filter was injected with a wide band noise source and
observed on a spectrum analyzer to "paint a picture" of the overall
response. Peak power occured around 10MHz as one would expect, with
secondary power peaks occuring at 4, 8 and 25 MHz.

The output filter was then reconnected to the HC240 drivers at C27
and the input and output impedance was measured on an H-P 4815A RF
Vector Impedance Analyzer, which tells you the complex impedance values,
which was Zin=110 ohms at +80 degrees (+ phase shift means it appears
inductive). The Impedance Analyzer is then scanned in frequency to
find the points where there is zero phase shift ... this identifies
points of self resonance. This method also showed self resonance at
4,8 and 25 MHz. The above two mods were performed and the above tests
repeated, which showed the 4,8 and 25 MHz resonance points were
attenuated by 12-15dB, and the input/output impedance was lowered to
nearly 50 ohms with a -85 degree phase shift, meaning the output filter
now appears capacitive. With a 50 ohm dummy load added, Zout went to
48 ohms at -5 degrees, or very well nulled for 50 ohms at 10 MHz.

With or without this mod, the 2nd harmonic remains at -40dBc, or 10dB
better than the FCC requires.

72, Paul NA5N


Search QRP-L Archives

[ QRP-L Archive | ]
[ 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 ]

---------

This archive was generated by hypermail 2b29 on Fri Jun 02 2000 - 11:32:50 EDT

kd4ab@kd4ab.org