I would say the 2 stage  "Driscoll" oscillator is the
 way to go.  I  have had good luck with it up to 100 MHz.
 The  first stage has the crystal in series with the
  emitter, but is otherwise a grounded emitter stage.

 Rick  Karlquist N6RK  

2 points: First I do not get copies of my own mail... strange . Second a complete treatment including the results is in http://ieeexplore.ieee.org/document/7546729/ These results are superior to other publication , and the paper is complete. Probably only very few do a complete literature search. The Disco oscillator, while applaudable (for its time ) is not used commercially in any product . 73 de Ulrich N1UL In a message dated 2/6/2017 11:00:42 A.M. Eastern Standard Time, dk4xp@arcor.de writes: Am 06.02.2017 um 14:08 schrieb Richard (Rick) Karlquist: > Agreed, for low phase noise FLOOR, it is imperative to > take the signal out through the crystal. However, for > close in noise (say ADEV at t=1), the Driscoll has > worked well for me. I have been able to reach ADEV > = 10^-11 at 100 MHz at using suitable resonators. > > Rick But one won't be able to use the power right out of the crystal for anything. So it will have to be amplified & buffered. If you can do that without lifting the noise floor, then you've got to ask yourself one question :-) Why don't I use that little wonder for the sustaining amplifier, too? And - why do I divide the precious crystal power between the 2 amplifiers at the location where it hurts most: where the level is smallest? When you compare the Driscoll and the Burgoon (sp??) output coupling through the crystal, you see it is exactly the same. One might even apply the current step up trick from Burgoon. The current through the drains/collectors is enforced by the crystal, operating into a near-short. Off-resonance the transistor has complete negative feedback and no gain. On the output side of the buffer, losing a dB or two for sustaining the oscillation does not hurt. A thing I do not like about the typical Colpitts is that it is never on the series resonance of the crystal. That means that the feedback divider is part of the resonance which increases the number of critical parts. In the Driscoll, the sustaining feedback is quite a wideband thing and mostly decoupled from the sharp crystal resonance. regards, Gerhard, DK4XP > > On 2/6/2017 4:35 AM, KA2WEU@aol.com wrote: >> Not quiet, using the crystal also as filter gives much better numbers >> ,,,,, >> 73 de Ulrich N1UL >> >> In a message dated 2/6/2017 7:30:33 A.M. Eastern Standard Time, >> richard@karlquist.com writes: >> >> I would say the 2 stage "Driscoll" oscillator is the >> way to go. I have had good luck with it up to 100 MHz. >> The first stage has the crystal in series with the >> emitter, but is otherwise a grounded emitter stage. >> The second stage is in cascode as a grounded base. >> The important operating condition is that only >> the second stage limits. First publications on >> it were in the early 1970's (search Michael Driscoll). >> >> Rick Karlquist N6RK _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.