jax wrote:As 3 mtr
jim stated the pll transmitters stephen sold evolved from a vfo exciter
that operated at quarter of the frequency with a pll that was
programmed with wire jumpers, this then evolved into the 50meg version
with virtually the same pll circuit with a few improvements such as dip
switch frequency selection way before paul hollings got involved.
A
bit later on, I helped with the further evolution of the PLL PRO III.
It was my suggestion to use the 74HC4059, because we wanted to get away
from using a "programming chart" and the BCD code inputs (and the lack
of a reset period) would allow a move to BCD rotary switches (when he'd
used up all the DIP switches!). These would have directly shown the
selected frequency. I also designed a frequency display for the rig -
which would have been the first NRG product to contain a PIC! I also
added the OLPD to the board layout, so that the output was switched off
(by raising the emitter voltage on the driver stage).
I
redesigned his little PRO III stereo coder so that it knew its left from
its right, and the input sockets on the board could be correctly
labelled. I also added the facility to switch to mono, and the two LEDs
(to match the PRO III exciter) which indicated mono and stereo. The
modifications radically improved his old basic product and made it
sufficiently different from the "Veronica" thing (that didn't know its
left from its right and put big spurs out 4.684MHz either side of the
carrier because of the poor board layout).
The
PRO IV Stereo coder was my design. It was something I'd been playing
with for a while, and Stephen wanted a new product that would made his
range stand out more from the charlatan down the street. I took a
Veroboard prototype up there, and Stephen and I spend a day laying out
the PCB. We etched and drilled a couple of prototypes, and populated
the boards with the ICs I'd brought along for the purpose. One minor
improvement to the design involved a cut track and an earth strap (to
reduce a tiny hum), and we were astonished by how good the thing
sounded. The measured results were amazing - the noise, distortion,
separation, phase accuracy, and frequency responses (and lack of spurs)
were all superb. They even got sold to a lot of professional,
commercial broadcasters, and they are still in widespread use in many
countries in their 1U 19" rack version with balanced audio inputs.
Stephen's
oscillator was a dual Hartley type, derived from the "Kallitron"
oscillator developed in the 1930s (by Cheirix and Mesney), and had a
number of things to recommend it. Even without the PLL, it was quite
stable, and if the capacitors were carefully chosen, and the coils made
from sufficiently stiff wire, it would largely stay put. The big bonus
was that it was "self-doubling" and developed the Band II signal when
the two oscillators were cross-coupled. There's a sort of "current
robbing" phenomenon going on, and the two oscillators inherently
synchronise themselves with the right phase relationship to double their
basic frequency. It also has the benefit that the highest frequency
the PLL has to sample will be 54 MHz. This can easily be counted by
cheap 74LS TTL ICs. The output of the oscillator was fed through a
highpass filter to eliminate any small ½f content that could otherwise
leak through due to minor component tolerance imbalances. Steps were
taken in the component selection and in the board layout to ensure the
closest balance possible.
Another innovation of mine was to use
FETs to do the sampling from the oscillator, and we also experimented
with diode switching of capacitors into the oscillator (selected by the
4059 programming lines) to select the part of the band the oscillator
was going to. This eliminated the trimmers altogether, but never went
into production.
Incidentally,
if you short the two bigger square pads beneath the 2SC1947 with a
piece of flat copper braid (taking care not to short anything else out)
you'll increase the gain of the final stage and the board will give
>5 Watts ight across the band.
Incidentally,
there's a subtlety in the driver stage - it doesn't give the same power
out across the band. The final needs a few extra milliwatts at 108 and
fewer at 87.5, so a gain slope was introduced to the driver to flatten
out the output power.
The Hollings rubbish never managed the
performance of Stephen's gear. Hollings lost in court (I was an expert
witness), and it cost him a lot of money. We demonstrated that
Stephen's designs actually were his, and that the Hollings products were
just poor, cut price copies with very inferior performance.
Unfortunately, Stephen also had to pay out a lot in legal fees as there were no costs awarded.