- For the Output transistor I chose the NTE342
which has more
dissipation than the NTE341, ( replacement for the MRF237). It
has a
Dissipation of 12.5 Watts opposed to 8 watts of the NTE341. It
also seems,
according to NTE specs, to have more gain.
The 342 has the tab connected to the emitter, so I used a Mouser
heat sink
part#532-576802B00 and cut the top two "finns" ( for lack of better
word.... I'm CX , remember), to allow "headroom" for the extra
band
module(s) that I carry attached to the inside of the lid with velcro
stripes. At a later time I "improved" the heat sink with a piece
of anodized
angle behind the Mouser heat sink.
The E and B terminals have to be bent to fit in the holes so the
342 is looking backwards. I preferred to solder it self standing
and with
the heat sink attached instead of using the back cover as a heat
sink,
resulting in shorter leads.
- I enlarged the PCB holes to fit the thicker 342 leads
and had to
scratch a little the mask on the top to make a better electrical
connection, (especially the emitter), since the enlargement of
the original
holes chew the inside sleeves. Filing the 342 leads to make them
fit in the
holes only weaken the leads and compromise its mechanical strenght.
- L10 was replaced with a 4:1 transformer wound in
the same core FT37
using 10 bifilar turns of enamel 26ga. wire. I tried both the twisted
(10
turns per inch) and parallel bifilar winding with no difference
in Ic. FYI
the one now working is the twisted pair for no particular reason.
Solder
the end of one winding with the beginning of the other. One end
to B+, the
other to C46 and the center, ( the junction of the two) to the
342
collector. I didn't cut the PCB trace. Just lifted the collector's
side of
C46 lead and soldered to the new xfrmr "in the air".
- I added another 1N5817 schottky diode in parallel
with D6 since in 80M
band the Ic is about 0.9 Amps. The 1N5817 rating is for 1 Amp,
so I decided
not to challenge the specs.
- The modules were re-touched after the power mod,
but only in the 80M
an 15M L5s I removed one turn. 40, 30, 20 and 17 got away with
squeezing or
separating the turns for best output. In 17M L6 "squeezing" rendered
a big
improvement. The 15 M module is the touchy one. I used the Rat
Shack
plastic tune-up set and had to enlong the yellow one (guess the
best one
for this task) with the red tubbing in such a way that my hand
was abt 8"
away from the rig while tunning. Only a scope was used for the
power mod.
The RX caps peaking was done using my main transceiver as a signal
generator.
- The Buzznot B+ lead was routed under the PCB for
neatness and it seems
to pick up less hum from out of band signals.
- The KC2 VFO connection was made with a straight stiff wire
with no
shielding and I only found one birdie around 18.080
- I did not touch R22 nor anything related to the VFO or PMOs.
The output power is about
6.2W in 80M, Ic of 0.9
5.7W in 40M Ic of 0.8
5.0W in 30 and 20
4.5 in 17
2.9 in 15M with an Ic of 0.63A
The B+ is 12.6 (I use it with batteriess only).