Different usages of Linrad.
(Sept 25 2009

A CW receiver with a small time delay and a fast waterfall graph covering a bandwidth of 90 kHz.

When Linrad is run on a modern computer (Pentium 4) the time for the actual processing is very small. The processing delay is determined by the size of the FFTs that do the filtering and by the delay caused by the input and output buffering. By setting small transforms one makes a compromise with the filter shape but the performance is still quite respectable. This link: low delay with fast waterfall shows parameters and performance for running Linrad with a bandwidth of about 1.5 kHz and a processing delay down to about 10 milliseconds while loading the CPU with a fast waterfall with svgalib. Performance under X11 and Windows 200 is also shown.

A "conventional" SSB or CW receiver with good filters and normal speed waterfall graphs covering a bandwidth of 220 kHz.

The SDR-14 hardware is the first entirely digital hardware available to radio amateurs. The maximum nominal sampling speed is about 160 kHz but under Linux-2.6 kernels it can run up to about 240 kHz with a Pentium 4 and a modern USB PCI board. This link: very wide waterfall shows parameters and performance for running Linrad under Linux 2.6.xx with svgalib and X11. The drive routines for Microsoft Windows and Linux-2.4.xx do not allow running the SDR-14 50% faster than the nominal sampling speed. The link shows some screen dumps on how Linrad behaves if an attempt is made.

A signal processor for a conventional receiver to improve readability of weak EME CW signals

Linrad can be used to process the loudspeaker output of a conventional ssb receiver. This link Linrad at SSB bandwidth shows the usage of Linrad as an add on to a conventional EME receive system for improved performance. The usage as a post-processor to a conventional receiver was one of the originally intended usages and this Linrad at SSB bandwidth in 2001 shows the same thing five years earlier.

Using the S-meter to measure the amplitude of a weak and frequency unstable signal and its variation with time

When a linearly polarized wave is reflected off the moon the reflected signal will be depolarised to some extent. Depolarisation on 10 GHz EME shows the evaluation of the signals from the 20 m parabola at Bochum received by OK1KIR April 2 2006.

The S-meter with 144 MHz EME signals

Here are graphs showing the typical fading due to libration on 144 MHz. Signal variation with time on 144 MHz EME. A strong wsjt signal is also processed and the Linrad screen is compared to the wsjt screen.

The Linrad S-meter showing peak to average power on SSB signals

Since the Linrad S-meter graph can be set to show both peak power and average power it can be used to judge the quality of whatever speech processing done at the transmit side. Peak to average power on SSB. Here are a few examples of typical curves.

Using Linrad as a spectrum analyzer with extreme resolution

Linrad both allows large fourier transforms and a zooming function that allows a narrow frequency range to be studied at extreme resolution. Linrad at mHz (milliHertz) resolution bandwidths gives a few examples. Radio signals that propagate via the ionosphere are affected by turbulence and movements in the ionosphere. Calibration at 10 MHz. shows some screens that illustrate the limited accuracy of calibration on 10 MHz.

Using a wide waterfall to find extremely weak signals

The typical computer screen has about 1000 pixels across the screen. When looking for really weak signals like EME transmissions in the JT65 mode one needs a bandwidth of about 5 Hz to optimize S/N. In case one wants to display more than about 5 kHz of the RF spectrum some S/N is lost. If however the waterfall is allowed to run very slowly the loss is small. High sensitivity wide waterfall shows an evaluation of the sensitivity of a 96 kHz wide waterfall that is optimised for JT65.

To SM 5 BSZ Main Page