Hi Ron
If used for the BITX design, do you think the squarewave
output signal can remain unchanged and directly be fed into
Q7 of the BITX ? Or will a low-pass filter be required ?
I think it should be fine. If you look at the waveform at the mixer port of
the original BITX it isn't very sinusoidal anyway. When I get a chance, I
will put my homebrew spectrum analyser on the original VFO and see how the
harmonic content looks.
About the timebase. Compared to the original Huff&Puff (timebase
of several tens of MHz, e.g. 60 MHz) the 32... KHz frequency is
very low. What will happen when, for example, a 1 MHz X-tal will
be used instead? Will the design continue to work? I assume the
lock frequencies will be much closer to each other but since
nothing is for free there will also be a disadvantage...
The design should continue to work. The distance between lock points simply
increases. In the extreme with distances of several kHz between lock points,
you would need to increase the range of the correction loop - this means
decrease the series resistor between the integrator and the varicap (1M in
my original). The maximum division ratio of the 74HC4060 is 2^14 therefore
with a 1MHz crystal the smallest lock step you could attain would be 61Hz,
which is still quite acceptable. Of course, if you are prepared to add a
third IC for more timebase division, you can have whatever division ratio
you like! Note if you decide to use MHz crystals: for higher frequencies,
you should decrease the resistances in the oscillator circuit: I suggest 2K2
and 1M in place of 330K and 10M.
The advantages of the 32.768kHz crystal are:
1. They're cheap and everywhere, (clocks and watches)
2. They're ordinarily very small physical size
3. Small stages of division required to get good step size
4. Low power consumption
On the latter point of low power consumption, this is the original reason
that I started considering the low frequency watch crystals. Originally, I
wanted to build a better version of my 2-chip frequency counter
. The Mk2 version of that
frequency counter, with the low current LED's, consumed between 4 and 5mA
current, depending on how many LED's were lit. However, fully 3mA of that
was spent on the crystal oscillator! It seemed highly unreasonable to spend
so much current consumption on the timebase when normally one would expect
that the most important place to spend power would be the LED display.
The reason for the relatively high power consumption in the oscillator was
that the CMOS logic must work hard to whip the capacitance back and forth,
4,096,000 times a second. The only solution: lower frequency crystal. The
32.768kHz circuit consumes a matter of some tens of uA.
After much correspondence with Arv K7HKL he inspired and encouraged me make
a combined Huff Puff and frequency counter, which eventually resulted in the
current minimalist designs.
Have fun and let us know how the results are
73 Hans G0UPL
