[ARC5] Crystal oscillators - the Butler Oscillator.

[Leslie Smith]

Wayne raised the topic of a circuit to check crystal activity.  For some
time I had an interest in the Butler oscillator but I had never had this
circuit on my bench before now.   This gave me the interest to built
'The Butler'.  I'm sure many here will have done that already.

White "bread-boards" are used at times to prototype experimental
circuits - these have IC-like terminal points wired as a buss; 5 common
points are arranged in strips on either side of a 0.3 inch gap.  You
probably have several.  These are often used to knock up small digital
circuits.  I used a 'proto' board to build an experimental Butler xtal
oscillator.

For anyone interested the circuit is in the 'Yahoo' ARC5-radio files. 
Look here:

  https://groups.yahoo.com/neo/groups/ARC-5radio/files/Exptl_%28LNS%29/

(Only 'members' of that 'Yahoo' group can see the (incomplete) sketch,
CNV-1606 etc.  The format is jpg.  Anything like IRFANview will open
it.)  

Looking at this, you'll see the crystal lies between the emitter of two
transistors - Q30 operates as a grounded base, Q31 as an emitter
(cathode) follower.  (Ha!  Had to get some valve terminology in here
somehow!!)  C31 (100pF in the circuit) is in series with the crystal.

A Fluke counter displayed the frequency.  Now to the point of this. 
With the circuit as drawn my counter told me any one of a dozen or so
4MHz crystals oscillated at 3998.8kHz (approx 1kHz low).  So I reduced
the value of C31 (obvious).

     C31        Freq.
     22pF      3999.567  (440Hz low)
     10pF      3999.925  (75Hz low)
  nothing      4000.594  (500Hz high)

What does "nothing" mean?  Between these "common" (not 'gnd') lines some
sort of capacitor forms, and even without a physical capacitor in the
ciruit there is enough 'stray' capacitance to create a feedback path
that includes the crystal.  That's a capacitor of 'nothing' pF.

I get further evidence of the problem of stray capacitance when I
connect a 100pF crystal between one 'leg' of the crystal and the
conductor ADJACENT to the other leg (of the crystal).  There is enough
capacity to pull the crystal frequency some 30-40Hz lower.  What does
this suggest?  First, I need glasses!  Second, that these white 'proto'
boards have a lot of stray "C" between conducting strips. 

It's common knowledge that these proto boards are useful for digital
logic, but not for analog circuits.  This is true for the oscillator I
built, but I also know this after building low-level AF circuits. 
Without a grounded aluminium screen under the protoboard - I hear LOTS
of 50Hz hum.

As for the crystal activity checker - it seems to work.  I built a
single diode RF detector and fed that into a x2 follower (OpAmp) driving
a LED.  The brightness of the LED follows the crystal activity.  Some
say 'toob' based checkers are more reliable.  I can't make a comparison.
 I can say that this circuit I 'pilfered' from the web identified a
number of crystals that didn't/wouldn't oscillate.

Apart from checking crystals, I hope to use this oscillator to 'feed' a
simple converter to 'move' the AM/BC band down to the range of the 'nav'
band receivers (eg BC-453) or 'up' to 4.5 to 5.6MHz.   

73 to all

Les

-- 
http://www.fastmail.com - Same, same, but different...