Once you have measured the actual output frequency, you can calculate the required correction to the
125MHz reference frequency and enter it in the “Ref. Frq.” configuration setting. For example, suppose
you set the output frequency to 12.500000 MHz but you actually measure 12.500075. Your output
frequency is 75 Hz too high. Since your output frequency setting is 10 times less than the 125MHz
reference oscillator frequency, this means that the actual reference oscillator value is 750Hz too high. So
enter 125,000,750 in the “Ref. Frq.” setting.
7. DDS Module Stability
Here's an extremely important tip regarding the stability of the DDS module. If you just wire everything up
and go on air, you would surely be ashamed to see how horrible your signal will probably look:
Obviously the DDS module was designed for something else, not just our QRSS passion, and the stability
at Hz-level wasn't important. My own theory is that there are temperature sensitive components within the
125MHz crystal reference oscillator, which take time to all reach the same temperature. When they do, the
temperature coefficients largely cancel each other out. But in the short term, while the different
components are experiencing slightly different temperatures, the frequency can shift considerably. I
believe that even small air currents over the oscillator case are enough to alter the local temperature and
cause frequency instability.
I have found that in practice there are two easy ways to solve this problem. Preventing the instability is
very important before putting the transmitter on air.
1) If you put the kit into an enclosure, such that you've removed any air currents around the crystal
oscillator, the frequency instability disappears.
2) Attaching (e.g. with glue) a small heatsink to the crystal oscillator can also cures the frequency
instability. I found that almost any heatsink will do. Even take the smallest, most pathetic coin in
your country and place it on top of the crystal oscillator, and the frequency instability is cured. In fact
even non-metal will do it – a piece of plastic, for example – which also adds weight to the air current
theory, since anything you put on top of the crystal oscillator stops air currents altering the
temperature so easily.
Arguably a heatsink may be a good idea in any case, to slow down temperature changes of the crystal
oscillator anyway. Either way: just don't leave the crystal oscillator in the open air.
Below see the result of placing a coin on the crystal oscillator.
13
