vcxo bandwidth

[Stefan Simion]

Hello,

I would some help to understand why the VCXOs "in general" have a relatively
low
modulation bandwidth, 10 to 30 kHz at most? I would like to know if there
is
a fundamental reason for this, or is it that for some practical reason, the
BW is
limited on purpose?

Thanks,
S. Simion

 

[Rich Grise]

Hello,

I would some help to understand why the VCXOs "in general" have a
relatively low
modulation bandwidth, 10 to 30 kHz at most? I would like to know if there
is
a fundamental reason for this, or is it that for some practical reason,
the BW is
limited on purpose?

Thanks,
S. Simion

It's because of the nature of the crystal oscillator, so, I guess you could
say there's a "fundamental" reason for it, nyuk nyuk. The crystal has a
natural resonant frequency. You can change the frequency by adding some
kind of load that will literally slow down its vibration, but not by
much, electrically.

Think of adding a weight out at the end of a tuning fork. Adding capacitance
to change the resonant frequency is like adding a 1g weight to a 100g fork,
so to speak.

Hope This Helps!
Rich

 

[Ken Smith]

Hello,

I would some help to understand why the VCXOs "in general" have a relatively
low
modulation bandwidth, 10 to 30 kHz at most? I would like to know if there
is
a fundamental reason for this, or is it that for some practical reason, the
BW is
limited on purpose?

I think you'll find that this is a case of both being true.

If you attempt to change the frequency of oscillations too quickly the
amplitude of oscillation changes and the frequency doesn't follow the
intended change. This is because the Q of the crystal is very high and
the amplifier's coupling to that high Q system is loose.

If high frequencies were allowed to go into the variactor diode, the
oscillator would tend to lock to harmonics of signals on that input.

 

[Stefan Simion]

OK, thanks. I think I really meant "small signal" BW, i.e. small
frequency changes. It is probably better to say "phase modulation"
of the XO? Is it possible that, in addition, the modulation BW
which can be achieved may also have to do with the proximity of
other (unwanted) modes of the xtal?

 

[Andrew Holme]

I would some help to understand why the VCXOs "in general" have
a relatively low modulation bandwidth, 10 to 30 kHz at most?
I would like to know if there is a fundamental reason for this,

It is fundamental. You simply cannot pull a crystal very far. If you
get your oscillator working way off frequency, it is because the
external LC elements are now controlling the frequency instead of the
crystal. The stability and Q will be poor. The crystal may be
contributing some reactance to the circuit but it will not be
resonating.

 

[Arie de Muynck]

"Stefan Simion" ...

I would some help to understand why the VCXOs "in general" have a relatively
low
modulation bandwidth, 10 to 30 kHz at most? I would like to know if there
is
a fundamental reason for this, or is it that for some practical reason, the
BW is
limited on purpose?

The crystal is a resonator with very high Q.
Bandwidth is Fosc / Q, thus with a very high Q you get low bandwidth...

Example: 30 MHz / 10000 = 3 kHz

Regards,
Arie de Muynck

 

[Ken Smith]

OK, thanks. I think I really meant "small signal" BW, i.e. small
frequency changes. It is probably better to say "phase modulation"
of the XO? Is it possible that, in addition, the modulation BW
which can be achieved may also have to do with the proximity of
other (unwanted) modes of the xtal?

In general the further you pull a crystal the lower the Q gets. At some
point the crystal may as well be left out of the circuit.

If there is a nearby mode, it may hope to it or it may get very noisy.
The noisy case is partly because the phase vs frequency characteristic
gets flattened, partly because noise at the other mode gets through and
partly because the system can have a multivalued responce.

 

[ddwyer]

If there is a nearby mode, it may hope to it or it may get very noisy.
The noisy case is partly because the phase vs frequency characteristic
gets flattened, partly because noise at the other mode gets through and
partly because the system can have a multivalued responce.

A cooperative crystal manufacturer can design with reduced spurious
modes.
Such crystals are necessary for filter design.
Fundamental crystals pull 9 times better than 3rd overtone.
Large diameter fundamental crystals pull further than small.
therefore large diameter with smaller than normal electrodes are
required.
Some suggest that paralleling crystals reduces motional inductance and
increases pulling.
SAW resonators pull further can be run at higher power and can be
divided down.
..