<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
</head><body style="">
<div>
The amplifer could indeed be perfectly linear up to its rated CW output (they aren't perfect of course), but the nature of the DAB modulated signal is such that without pre-correction its power peaks are massively greater than the average power, so the average power has to be kept low enough that it doesn't drive the amplifier beyond its acceptably linear region on the peaks.
</div>
<div>
 
</div>
<div>
Remember we are talking about an SDR that is designed to be used for all manner of modulation systems: AM, FM, DAB, DRM, SSB, GSM, wif-fi - any of these and even some  not yet invented! It's output transceiver is capable of reasonably linear operation up to 100mW continuous carrier power, and if, for example, it were being used as an FM transmitter, then it could indeed be used close to that, because FM is a continuous full power relatively slowly varying sine wave.
</div>
<div>
 
</div>
<div>
On the other hand, the COFDM modulation system used for DAB can include rapidly varying peak modulation levels that are many times higher than the average level (think of it like the '100-year wave' effect where a generally moderate sea can suddenly throw up one huge wave when a whole pile of interfering waves happen to constructively interfere). That isn't a 'fault' - it's an inherent aspect of the design of the COFDM system.
</div>
<div>
 
</div>
<div>
The unfortunate side effect of this is that to avoid the amplifier being driven outside acceptable linearity and producing spurious signals it must be run at a relatively low average power level compared with its peak capability. For the 100mW transceiver on the SDR that Ofcom used they found it had to be kept down to an average power of 5mW for transient peaks to be handled linearly enough (in their case producing a 54dB carrier to noise ratio).
</div>
<div>
 
</div>
<div>
For a high power DAB system where the costs are shared by maybe 10 radio stations, it turns out that it is more economic to use more advanced modulation hardware t hat can 'pre-correct' the signal - ie slightly distort it to reduce the ratio of peaks to average, then run the amplifiers more highly and to use humungous six- or eight-cavity output filters to clean up the act. But for low power transmitters serving just a couple of stations it turns out to be cheaper to noiminally 'under-run' the amplifiers in terms of the average power versus the amplifier's rated CW ability, to allow the use more basic pre- and post- filtering.
</div>
<div>
 
</div>
<div>
Alex
</div>
<blockquote style="position: relative; margin-left: 0px; padding-left: 10px; border-left: solid 1px blue;" type="cite">
On 10 March 2015 at 10:25 Tony Bailey <ravensound@pilgrimsound.co.uk> wrote:
<br />
<br />
<div class="moz-cite-prefix">
I'm not so sure about that analysis, after all, the report implied that it was the SDR that was under-run at 5mW presumably because the output daughterboard was intended for AM/FM modulation?  Also, it was common IBA practice to run a several kilowatt bottle at a hundred watts or so, the excess input power being necessary to ensure class A linear operation.  The point being that the rf amp has to be linear, and it isn't, you can't just turn it down a bit.
<br />
<br /> Tony Bailey
<br />
<br />
<br /> On 09/03/15 22:41,
<a href="mailto:tlr@gairloch.co.uk">tlr@gairloch.co.uk</a> wrote:
</div>
<blockquote type="cite">
<div>
Pete: We know what software and modulator they will probbaly be using, but one of the key features of Ofcom's proposals for low-cost DAB broadcasting is avoiding the use of the conventional large and expensive DAB mask filters (typically 6-cavity high power devices) on th efinal output.
</div>
<div>
 
</div>
<div>
This can be achieved by a combination of tactics, including a very under-run driver stage (eg 100mW stage running at only 5mW) and relatively underrun RF amplifiers (eg maybe using an amplifier potentially capable of say 250-500W continuous carrier power to produce a DAB signal of less than 100W), so that there are fewer spectral regrowth components to be filtered. Plus with the whole kit running at relatively low ERP anyway, the filtering is not as critical as it would be for a high-power transmitter because the absolute level of spurious products is low relative to neighbouring high-power signals and aeronautical limits.
</div>
<div>
 
</div>
<div>
Alex
</div>
<div>
<br /> > On 09 March 2015 at 20:18 Pete
<a href="mailto:peter@engineeringradio.co.uk"><peter@engineeringradio.co.uk></a> wrote:
<br /> >
<br /> >
<br /> > It's obvious what type of exciter they will be using if you have been following them. It's obvious it will be SDR based. The amps may be a different matter. I don't think it is clear what they would be using. They would probably use filtering to the spec set out for DAB.
</div>
<div>
 
</div>
<div>
 
</div>
<br />
<br />
<pre>_______________________________________________
Reply - <a href="mailto:cma-l@commedia.org.uk">cma-l@commedia.org.uk</a>
The cma-l mailing list is a members' service provided by the Community Media Association - <a href="http://www.commedia.org.uk">http://www.commedia.org.uk</a>
Twitter: <a href="http://twitter.com/community_media">http://twitter.com/community_media</a>
<a href="http://www.facebook.com/CommunityMediaAssociation">http://www.facebook.com/CommunityMediaAssociation</a>
Canstream Internet Radio & Video: <a href="http://www.canstream.co.uk/">http://www.canstream.co.uk/</a>
_______________________________________________
Mailing list guidelines: <a href="http://www.commedia.org.uk/about/cma-email-lists/email-list-guidelines/">http://www.commedia.org.uk/about/cma-email-lists/email-list-guidelines/</a>
_______________________________________________
To unsubscribe or manage your CMA-L mailing list subscription please visit:
<a href="http://mailman.commedia.org.uk/mailman/listinfo/cma-l">http://mailman.commedia.org.uk/mailman/listinfo/cma-l</a></pre>
</blockquote>
<br />
<br />
<pre class="moz-signature">--
Local Reports at <a href="http://www.ravensound.pilgrimsound.co.uk">http://www.ravensound.pilgrimsound.co.uk</a></pre>
</blockquote>
<div>
<br /> 
</div>
<blockquote style="position: relative; margin-left: 0px; padding-left: 10px; border-left: solid 1px blue;" type="cite">
_______________________________________________
<br />
<br />Reply - cma-l@commedia.org.uk
<br />
<br />The cma-l mailing list is a members' service provided by the Community Media Association - http://www.commedia.org.uk
<br />Twitter: http://twitter.com/community_media
<br />http://www.facebook.com/CommunityMediaAssociation
<br />Canstream Internet Radio & Video: http://www.canstream.co.uk/
<br />_______________________________________________
<br />
<br />Mailing list guidelines: http://www.commedia.org.uk/about/cma-email-lists/email-list-guidelines/
<br />_______________________________________________
<br />
<br />To unsubscribe or manage your CMA-L mailing list subscription please visit:
<br />http://mailman.commedia.org.uk/mailman/listinfo/cma-l
</blockquote>
<div>
<br /> 
</div>
</body></html>