How to achieve the best PSK31 signal
I've been using PSK31 since I tried it a few years ago. It's one amazing mode if you like doing things digitally, you like keyboard QSOs and you know that five watts can take you anywhere in the world.
I like competitions and long evenings with QSO to find out how people live elsewhere in the world. My problem was, that I was hampered by retuning the signal. Not a walkie-talkie, but the audio settings level of the sound card. From the beginning, I set the audio output to maximum in Widows, to get the most out of the radio. Well, I learned it's not such a good idea- if you overwhelm the device, this can cause a large amount of splater to form and 31 The Hz signal can be up to several hundred Hz wide, which makes it impossible to make further connections. I brought an oscilloscope from work and looked at the RF output at maximum load. It was a caseal case of intermittent sinusoids. The solution to this situation is to keep those sound card settings down, not at maximum. However, there is a question – as far down? With little audio output, nothing or almost nothing comes out of the device. So my dear oscilloscope fitted well with my device and I was able to monitor the RF output and adjust the audio output settings until I got a beautiful sinusoid without interruption.
This is how I was able to calculate my average output (P = E ** / 2R) and of course I got a good signal. The only weakness was, that my oscilloscope served only one purpose. So I got the idea to replace the PIC indicator with a microprocessor. Many PICs have built-in ADCs, however, they are not powerful enough to sample 14 Mhz signal. I was especially interested in the envelope of the curve, which means I have to rectify the signal and go through the RC filter and the ADC. Then I programmed the microprocessor firmware exactly, to allow 64 RF samples on 1 millisecond intervals- exactly two PSK31 bits. This buffer was then uploaded to the PC via the serial port. Then I wrote software based on the Fourier transform, which analyzes harmonic frequencies and calculates IMD (intermodulation distortion). At the beginning, the application showed only a wavy line and recommended to reduce or increase the volume of the sound card to achieve perfect modulation. I found out later, that my program has secure access to the sound card via Windows API and this application allows me to adjust the sound card output settings itself. When I looked back at the program, it worked. My PSK31 signal was permanently perfect without me getting upset about anything again.
The circuit for a small psk meter looks something like this:
I keep it simple: D1 and D2 provide routing and R3 + C2 again filtration. The PIC uses a 2n2202 NPN transistor to run an RS-232 connection to a computer. Although using transistors to switch between + 5V and ground is not technically correct for RS-232 (where is better switching between +12 a –12), it works well and it's cheaper than using a MAXIM interface chip.
You can find the firmware file at: http://www.ssiserver.com/info/pskmeter/pskmeter.hex, so you can build a circuit and program a chip, which is actually hardware.
The application in Windows looks something like this:
I contacted several local radio clubs and then Pacificon 2002 and radio amateurs keep asking me, if I could make a kit. I was a little reluctant, because it took a lot more work: design and invent printed circuit board production, quality control, complete manual etc.. Which convinced me to try it, was an amazing software on http://www.expresspcb.com. This free software allowed me to design a circuit and highlight a printed circuit board. In two days I received connections by e-mail. It's very similar to sending a file to a printer with the difference, that the printer is in another city. Good, Nope?
After several interactions and revisions, I received such a computer board:
George Rothbart (KF6VSG)
Translated by XY of MW3SDO, revised by MW3SDO
