Series Resonant Crystal Oscillator Circuit
The heat/humidity broke today, a little more conducive to bench work. So I thought I'd get back to a receiver project I was working on. The circuit will be a cut-down version of the Norcal 40 rig using the inexpensive board from the QRP-L Tech group.
I had gotten to the point in the build where the crystals used in the filter (and BFO) need to be sorted/matched/tested. In the Norcal 40 these are 4.9152 MHz crystals. Normally people connect the crystals to an oscillator circuit and then measure the frequency. The crystals which are closest in frequency are used. The oscillator circuits mainly use the parallel mode of the crystal so the frequency read by the counter is going to be a little higher than the series mode.
But I thought that since the crystals are being used in the series mode why not use a series mode oscillator circuit ? I found this circuit on page 125 of Crystal Oscillator Circuits by Robert Matthys.
This is a Common Base or Butler oscillator and I suspect it isn't as popular as others because of the extra transistors. With transistors it isn't too big a deal to add another one but if this was using vacuum tubes I'd see why designers shy away from it.
Anyway, hooked up the circuit on a 3M solderless breadboard
 |
| the white LED is just a pilot light on the input to a 78L05 5Volt regulator |
, got a decent buffered output
 |
| this is the rightmost output of the above circuit, EO |
, fed it to a counter from ebay. Here is the data for the 10 crystals, first as taken and then sorted.
 |
| this is just the last three digits of the counter reading |
As can be seen from the sorted data there aren't any 4 perfectly matched crystals for the filter but 4, 8, 9 and 10 are within 9 Hz so hopefully that should be good enough.
Besides, if the data is retaken on a different day and certainly with a different oscillator the results might not correlate.
We'll see, more to come on this subject.
Best Regards,
Chuck, WB9KZY
http://wb9kzy.com/ham.htm