I looked at this chip some time ago for use as the LO in an HF receiver,
and I abandoned it concluding that the noise performance of the VCO is about
20 dB shy of the minimum you need for good HF receiver overload performance
('reciprocal mixing").

As to the VCO tuning range, less is usually better from a noise standpoint,
as most varactors don't have very good Q specs and the tighter the varactor
is coupled into the overall resonant circuit, the poorer the resonant Q will
be. Also, high tuning sensitivity in the VCO makes the oscillator
susceptable to modulation from 'junk' on the power supply or the tuning
voltage.

Shorting out turns in the resonant inductor as a means of bandswitching
isn't so great either, as the shorted turns inductively couple into the
resonant circuit as resistors (the resistance of the pin diode or other
switching element), lowering the circuit Q and raising the noise of the
oscillator.

In general, it's best to band-switch discrete elements such as separate
inductors or capacitors and to keep the tuning sensitivity as low as
possible in order to provide continuous coverage of a band. If high tuning
range is necessary, every attempt should be made to use high Q components
and to maintain a high loaded Q in the circuit they're used in. Also,
parallelling of varactors has been shown to improve noise performance in
wide-range oscillators.

Joe
W3JDR


wrote in message
oups.com...

john jardine wrote:
Done a bog standard, single transistor Colpitts, that covers 110-220Mhz
using a cheap SMV1255-000 varicap. It's part of a BFO arrangement and
intentionally limited to a 110MHz swing (1.5 to 4.5V) but there's a bit
of
steam left in the Varicap and 120-240 seems not unreasonable.

That's a Kv of about 36 MHz / V. Naive question ... doesn't a higher
Kv
have the potential for greater noise? What is the effect of a higher
Kv
for the LO VCO on the overall performance of a receiver?