twin crystal oscillator

[colin]

while im waiting for my OCXOs to arrive and thinking about what makes one
OCXO better than others with regard to 1hz phase noise, i thought what if 2
crystal/amp stages were used for the oscillator, the bandwidth must be
reduced quite considerably and hence the close in phase noise too.

not much came up on a search, maybe i will build one and see. i have read
that very old crystals are better becuase their ageing has settled down to a
low rate and are the larger size gives better Q.

Colin =^.^=

 

[Tim Wescott]

colin wrote:

while im waiting for my OCXOs to arrive and thinking about what makes one
OCXO better than others with regard to 1hz phase noise, i thought what if 2
crystal/amp stages were used for the oscillator, the bandwidth must be
reduced quite considerably and hence the close in phase noise too.

not much came up on a search, maybe i will build one and see. i have read
that very old crystals are better becuase their ageing has settled down to a
low rate and are the larger size gives better Q.

Colin =^.^=


I wouldn't try two crystals _in_ an oscillator, but I believe that you

can get good results by following an oscillator with a crystal filter.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Michael Black]

Tim Wescott (tim@wescottnospamdesign.com) writes:

colin wrote:
while im waiting for my OCXOs to arrive and thinking about what makes one
OCXO better than others with regard to 1hz phase noise, i thought what if 2
crystal/amp stages were used for the oscillator, the bandwidth must be
reduced quite considerably and hence the close in phase noise too.

not much came up on a search, maybe i will build one and see. i have read
that very old crystals are better becuase their ageing has settled down to a
low rate and are the larger size gives better Q.

Colin =^.^=


I wouldn't try two crystals _in_ an oscillator, but I believe that you
can get good results by following an oscillator with a crystal filter.

Or there's that circuit, I first saw it in an article by Rohde and

I thought it was his idea, where you extract the signal from the side
of the crystal away from the active device. I can't remember the
circuit, or even if it was special, but he moved the side of the
crystal that normally was grounded off ground, I think with a
resistor, and took the signal from there. The crystal was
then acting as a filter between the oscillator and the rest of
the world.

Michael

 

[colin]

"Tim Wescott" <tim@wescottnospamdesign.com> wrote in message
news:116aj537n0mkr19@corp.supernews.com...

Michael Black wrote:

Tim Wescott (tim@wescottnospamdesign.com) writes:

colin wrote:

while im waiting for my OCXOs to arrive and thinking about what makes
one
OCXO better than others with regard to 1hz phase noise, i thought what
if 2
crystal/amp stages were used for the oscillator, the bandwidth must be
reduced quite considerably and hence the close in phase noise too.

not much came up on a search, maybe i will build one and see. i have
read
that very old crystals are better becuase their ageing has settled down
to a
low rate and are the larger size gives better Q.

Colin =^.^=



I wouldn't try two crystals _in_ an oscillator, but I believe that you
can get good results by following an oscillator with a crystal filter.


Or there's that circuit, I first saw it in an article by Rohde and
I thought it was his idea, where you extract the signal from the side
of the crystal away from the active device. I can't remember the
circuit, or even if it was special, but he moved the side of the
crystal that normally was grounded off ground, I think with a
resistor, and took the signal from there. The crystal was
then acting as a filter between the oscillator and the rest of
the world.

Michael

That would certainly reduce phase noise. If it were me I'd look at
implementing the best oscillator I could (I'd probably check the Butler
oscillator for starters). If that weren't enough I'd follow it with a
filter and accept the inevitable tradeoff between filter bandwidth and
the difficulty of matching the filter passband with the oscillator
frequency.

There is literature out there on low phase noise oscillators, but
crystal and otherwise.

--

Tim Wescott

doh, too late done it already, managed to get it working @ 20mhz with 2
bfe520 transistor inverters stages, would rather use series resonant sc cut
crystals rather than the at cut parallel resonance mode crystals wich is all
i seem to be able to get hold of easily. was thinking a 3 stage might be
better still as the phase needed for each stage would more convenient.

I figured a second one inside of the loop will be beter than on the outside
due of the effect of positive feedback, although of course one (or more) on
the outside is a posibility too. ive not tried it in circuit yet as i need
to make a second one, also the colector waveform has a weird kink in it
almost like its trying to oscillate at 2nd harmonic (wich it shldnt) but i
need to multiply it anyway. (and the ocxos have turned up now too

Ive seen the circuit refered to in a couple of places, ideal if you need a
clean sinewave. there are many low phase noise circuits out there wich apear
to be as good as the (single) crystal used.
my circuit feedback is similar in nature to a butler type.


Colin. =^.^=