Minimizing the Dreaded Drift of MF or LF
Rick W7RNB / WI2XJQ
Drift – That is frequency drift of your transmitted signal on either the LF or MF bands. It happens, whenever your running a transmitting downconverter. There always the possibility that you will drift a few cycles. The experience encountered was that it could the transceiver, the down converter, or both. But the good news is that it is solvable. Methodical consideration and troubleshooting will show you what needs to be fixed.
In my case it was both the transceiver and the down converter. Not bad mind you but for WSPR to be the best it can be one needs to have a stable signal. First was to fix the drifting of the transceiver; it was solved by the purchase of a low priced radio that has the stability of 1ppm. So and Alinco DX-SR8T was purchased. However, when looking at the output from WSPR on another receiver I could still see 4 or 5 hertz drift on my signal over several transmitting periods. It was also learned from this that WSPR can’t interpret drift, so spots can be missed, even if they stay within the WSPR window. This issue had to be fixed.
The downconverter that I use is a great little unit, and when I purchased the kit I ordered it with the crystal heater. This works quite well actually and if one was just going to do CW it would not make any difference in contacting others. However with the WSJT-x modes it might.
Using Arduino processors have come a long way and are quite easy to build and program. Basically they are all building blocks that one just wires together. All kinds of DDS boards are made that can be controlled with the Arduino UNO board; and the DDS chip stability is + or – 1ppm. As a matter of cost, a DDS can be built for about $30US, including some kind of box to put it in. They run on 9 to 12Vdc , I recommend the 9Vdc as it keeps the voltage regulator cool.
I used the AD9850 DDS generator chip that runs on 5Vdc, it is easy to control and is very stable. It has an output of about 50 to 70 millivolts P-p and with a bandpass filter plus a 2n5179 amplifier board you can get up to about 10volt P-p of clean output, if need be. In my case I needed less output so I used an -8db pad to reduce the amplifiers output to the SA612 recommended input level. All of these aforementioned items are available on the web, EBAY and / or KITS and PARTS and a healthy junk box is helps too. Once the build was complete I programmed it to be a 3.2Mhz VFO.
It’s a little bit of work, but once you have the DDS built and running it is pretty easy to hook it to your down converter. In my case, I removed the 3.2Mhz crystal from its socket, and the parallel capacitor on pins 6 and 7 of the SA612. Added a .01uf cap 50Volts to pin 6 of the chip and injected the DDS at 350millivolts. Then I added my 3.674.2Mhz signal at pin1 and the output was on 475. xxxKhz. A little over 11Volts P-p square wave to drive my class ‘D’ amplifier.
That’s it –