[hpsdr] New features in PowerSDR and KISS Konsole

Phil Harman phil at pharman.org
Sat Jan 28 19:55:30 PST 2012


All,

Many thanks to Jeremy and helpers for getting the new SVN in place .

There are a couple of new features both in Doug's, W5WC, PowerSDR(TM) and KISS Konsole that I'd like to highlight.  Both of these are thanks to Warren who took on the challenge of modifying dttSP for us and Doug who added the GUI support for PowerSDR whilst George undertook the tasks for KK.

The first is an new speech processor.  This is a processing technique suggested by Peter, G3PLX, and is the software equivalent of an RF speech clipper. It's by far the best speech processor I've ever used, a view shared by the beta testers.  Whilst the processor has a range of 0-20dB if you are going to use above about 10dB then I suggest getting some reports from a local station before letting loose on the bands.  The setting does vary between different voices so be sure to experiment.

The second feature is one that I've been wanting for many years and that is a really effective AGC system.  Warren has fully implemented some ideas I had back in the SDR1000 days and they work really well. 

The AGC attack, decay and hang settings have all be optimised and the noise pulse removal is exceptional.

One new feature is the addition of Hang AGC.  For those not familiar with this concept it works as follows.  For a strong SSB signal its nice to have an AGC setting with a long decay so that there is no band noise between pauses in speech.  Since the AGC decay is exponential the delay, before the receiver regains full sensitivity, for a very strong signal this can be many seconds.  With Hang AGC a strong signal can be allowed to trigger a large AGC gain reduction but when the signal is removed the AGC will maintain the gain reduction for a pre-determined period ( usually a second or two) and then very quickly return to full gain.  In this way the AGC decay time constant does not alter with signal levels.  It's very effective when working in a net where signal levels can vary greatly.

One issue is how do you set the level at which the Hang is activated?  Doug has implemented a simple visual guide in the form of a horizontal line that appears on the bandscope ( user selectable to be across the width of the screen or just with the receiver bandwidth).  When Hang AGC is available (for Slow, Long and Custom AGC settings) you can move this line up and down the screen with your mouse by placing your cursor in a small square at the left hand side of the line.  If the peak of the signal in the  passband crosses the line then Hang AGC is activated. To deactivate Hang  simply push the line above the signal towards to top of the screen.  

The line has an 'H' next to it for easy identification.  This feature can be turned off using the Setup form.

In addition there is a similar horizontal line to set the AGC 'knee' or threshold.  This line shows the signal level at which the AGC will start to activate. Just like the Hang threshold you can set it by moving the line with your mouse or by using the AGC-T slider.  I suggest setting the knee about 10dB above the average band noise. 

Finally when Fixed AGC gain is selected a horizontal line is displayed that indicates the level of gain being applied.

There could well be some final optimisation to do regarding the AGC dynamics so please report your findings and suggestions via the reflector. 

73 Phil....VK6APH 




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20120129/a52a373e/attachment-0004.htm>


More information about the Hpsdr mailing list