--- In BITX20@..., Ron Brink <pa2rf@y...> wrote:
Thanks for the good info dear Allison,
Yes, in the meantime I discovered MOSFET IRF's are not most
favourable candidates for a 6 m PA. The 2SC Japanes transistors are
more common and most likely much easier to use.
A line up of 2n4124, 2sc1970 and a pair of 2sc1971 will do
9-11W CW power at 12V nominal.
About the synthesizer circuitry, in the past I have used a divider
4059 in combination with a 4046.
If after the VFO a hi-speed divider (e.g. 74F74 or U256) is used you
can go up to the GHz range and have a rockstable vfo.
That is certainly one way to go. Another is to use the 4046 phse
comparator in a 1:1 tracking loop with the reference being a VFO in
the 1-4mhz range.
Myself I build PLLs using chips like the MM55104, MC145106 and
MSM5807. Most of these are designed for 10khz challening used in US
CB radio but are easy to apply, found in junk radios and CMOS.
My preferd chip is the MC145106. Using those in a loop has a
twist. The divide by N counter in most maxes at 255 (255*10khz is
only 2.55mhz). This would imply a limited ability to do PLL in
the 40mhz range however there are easy ways around it. The most
common is to mix the VCO with a local crystal osc (better yet a VXO)
and have the result end up on the 1-2mhz range for counting. Then
the /n counter set the channel step and if you use the VSO you
can get interpolation between 10khz channels. One the base design
is worked out (copied from CB rig) scaling to any LO is easy as the
loop is the same and only the VCO and VXO is moved to suit. If built
carfully with good shielding and ground plane the 10khz spurs should
be at least 70db down. Since noise is related to the /n value this
scheme tends to be cleaner than those using a VHF counter. The side
effect is tuning range is limited to /n in the range of 10 to 255
or about 2.4mhz. I've built several using those chips and I package
them in a box of PCB material around 1.75" square and about .75" tall.
usual power consumed is under 60ma @5v. I have one I've built already
for upcomming Bitx-6.
Oh for tuning the pll without a micro. Take a 27c256 or similar
Eprom and program is so the address lines take 2 or 3 BCD coded
switches and translate that to a binary code at the output to suit.
The processis static and requires no clock (no cpu noise!) and is low
power.
My bitx-6 is progressing well. One of the design goals is to fit
it in a box 5Wx2Hx7D (inches) without headset, mic and DC power.
power ouput goal is in the 3-5W (cw power) range using 2SC1971.
Allison
KB1GMX
