As I suggested yesterday, today I have been trying to test a bunch of crystals to see if I can match their resonant frequency. I have set the DDS and 'scope up as the diagram below:
So because the crystal is in parallel with the DDS frequency generator, I have to tune the DDS to get the minimum reading on the scope; this will then be the resonant frequency of the crystal. I think I need to be cautious over introducing capacitance into the circuit through test leads to get an accurate frequency, but as all I am trying to do is match crystals, rather than accurately measure their resonant frequency, I figured it didn't matter that much as they were all tested under the same conditions.
I only have 10 10MHz crystals here, and these are the results I found:
So what I propose to do now is re-make my crystal filter from yesterday using crystal numbers 1, 2, 4 and 8 and then re-test to see if my bandwidth has improved.
So after all that giggery pokery, the chart below shows the old crystal filter in red and the new one in blue:
So matching the crystals didn't make much difference, did it?
I can now calculate where the BFO in an SSB tranciever should be to get the USB and LSB passing through correctly:
So I'm going to go with these values and see how the performance looks.
Watch this space!