CXO: CXO/overtone for TX

@7/3  2012

I always want to build a transceiver. recently i decide to make one,  at first , play with VXO.  I ever built one  CXO: Low distortion oscillator  but it's not  a real VXO, it's just the CXO, the crystal controlled  Oscillator.

I happens to had some 27.095MH RC crystal.  An  walkie talkie is interesting to built for toy,  fist step to build  transceiver.  I hope the VXO can cover part of the 27Mhz band : 26.957-27.283MHz.

years ago, i  try  several times to built CXO above 20Mhz, but failed , because it oscillate on 1/3 of the Mark frequency.  I never dig to why, till this time, I realized the overtone mode crystal need a little special oscillator topology.

Now let's look the Fig.1, I started with it.  It's a basic Colpitts crystal oscillator.  it work on fundamental frequency, i put the 27.095Mhz crystal on it, and it given 9.04Mhz output.

What really happen there is that the fundamental frequency is activated.  BTW, the waveform on the OUT tip is good.

I  remember  some article talk about the crystal overtone, but never carefully think about it.  I decide to replace the collector load use a tank  resonant around the overtone frequency.  I use a tapped transformer. see Fig 3

And i re-read all i have, finally find this  is actually a multiplier, not  the overtone mode oscillator ,  there is no frequency selection in the feedback loop.  

And why the toy RC transmitter work fine on 3rd overtone mode? but mine did not?.

Overtone

What's wrong in the  Fig1, Fig3?  several article highlight it: DON NOT PUT tuned circuit in the collector LOAD,  that is multiplier not the overtone oscillator, cause there is no frequency enforcement in the feedback loop  [D1, Y1, 3.2uH [U$1], C2,C3 ], the tune circuit on T1 collector just given a different Gain at different .

An overtone oscillator should put the frequency selector to feed back path, refer following overtone oscillator configurations.

 A real Colpitts overtone CXO should like this one:

after trying all of these,  i find the right way to make a  overtone oscillator.   I change Fig 1 to  get the Fig 4. they are almost the same, but wait, the key is: Fig.1 use big L1, Git C1 far away from the overtone frequency.  and In Fig4, L1 series with C1 resonant just at 27Mhz, the 3rd overtone  [27.095/3].    so it works well on overtone mode.

And i  try the miller Common base oscillator  also [at the resonant frequency Y1 is low impedance ], Fig.6.  

Many book point out the T1's input impedance  will damp the tank's Q, make the phase noise performance bad,  trying use the JFET instead.  From Oscilloscope, it seems as good as Colpitts overtone mode oscillator, but I have no SA, damn it.

this configuration is easy to set up overtone VXO, i try play with it: super VXO mode,  only 2Khz shift.  almost useless,  that's why people  suggest use fundamental crystal for VXO.

most article/magazine/books talk about  Pierce Oscillator, but  just  throw few word to point out it is not so good. But you'll found them everywhere in the Toy level application.  Here is an toy car RC transmitter and the output waveform. 

And, all my son's RC car use same chip, the TX2B/RX2B chip,   and use the data-sheet's recommended circuit without change:

The L1, and 47p|| 68p=27pF  resonant at  around 20Mhz,  not 27Mhz.  you need carefully design LC tank for Miller overtone oscillator( Fig 6 ). So i wonder why this one is work?   and I built one my version:

you can see I use the L1 quite large, with C2 33p, resonant around 18Mhz.  IT STILL WORKS, successfully work in overtone mode. the output waveform is not good when compare to the miller Common base overtone mode( Fig.7),  and also worse than Colpitts overtone topology( Fig.5). 

Fig11  Pierce overtone and waveform

very compact layout 36Mhz oscillator design

@2014/8/05

refer to Fig.10  R2 use 24k, C2 use 16.8p(10p || 6.8), L1 use a 2uH primary IF can, more like FCZ28. Primary: secondary ~= 2:1. Pierce is powerful, without buffer, output to 50R 2Vpp, this is almost 10dBm.

By  suppress the fundamental mode,  Pierce oscillator is very easy to tune to overtone mode,  it's then so easy to built and easy start oscillate.  even though it's  full of harmonics, it's still very popular in toy level applications.  And i try the 54Mhz 5rd overtone crystal,  with all the same circuit  as Fig.10,  by just remove the slug, make it work at 5rd overtone mode, if now put back 27.095Mhz 3rd overtone crystal, it start oscillation at 45.15Mhz, the 27.095 crystal's  5rd overtone mode .. 

layout @2016 May/17:

Stepan Novotill reverse engineering many transmitter circuit , here is the one  i  like:  http://www.members.shaw.ca/novotill/TinyRcCars/index.htm

the 2.9uH with 150p , 10p  form the Pi match low pass network, and for pierce oscillator  overtone load.

Reference:

super VXO:         http://www.qsl.net/7n3wvm/supervxo.html

multiplier:            http://www.asahi-net.or.jp/~gt6s-sbic/electro/50m_am_5/vxo.html

butler overtone:  http://www.qrp.pops.net/butler.asp

overtone CXO:   http://hem.passagen.se/communication/txo.html

FCZ coil:             http://www.jarl.or.jp/Japanese/7_Technical/lib1/fcz.htm

27Mhz/40Mhz RC: http://www.members.shaw.ca/novotill/RcCarsSimple/index.htm

TX2B 27/49Mhz   :  http://www.members.shaw.ca/novotill/TinyRcCars/index.htm