bit eXpress

With DRM it’s not all about high power transmitters and international broadcasting.

There are a number of university and low power radio stations experimenting with DRM, including three such stations in Germany and one in France. Transmitters operate in the range 10 - 100 watts and broadcast in the 15 MHz (19m) or 26 MHz (11m) short wave bands.

‘bit eXpress’ is the digital radio station of Friedrich-Alexander University, Erlangen, Germany. Currently testing several aerials to find one suitable for optimal local ground wave coverage. Aerials under evaluation include dipole, magnetic loop antenna, ground plane, and EH-antenna. Future aerials include dipole at a higher level above ground, 5/8 lambda vertical.

The radio station uses an amateur radio linear power amplifier with nominal 2kW power output. DRM signals have a very high ratio between average signal power and statistically occurring signal peaks of about 13 dB (Peak to Average Power Ratio). Also the requirement that the transmitter must be operating in the linear region of the amplifier means only approximately 100 watts output is available from a 2kW amplifier.

The 2kW Arcomm power amplifier with antenna tuner, power and SWR meter.

Various aerials on the roof with the magnetic loop centre of the picture.

I received this Mode A broadcast using my indoor 1.5 metre vertical; which gave a better quality signal (higher SNR) than my external long wire. This reception was during the morning but have previously received this transmission at night. At the time the transmitter power was 45 watts using a dipole aerial.

The signal was just strong enough to decode some audio and multi-media. Not bad for a 900 km path from Erlangen, Germany to the UK and using Mode A which is the least robust DRM mode.

>>>>
DRM Demodulator/Decoder
Software Version 1.1.4
Starttime (UTC)  2004-06-05 09:08:05
Frequency        15896 kHz
Latitude         51°46'N
Longitude        0°56'W
Label            bit eXpress
Bitrate          27.20 kbps
Mode             A
Bandwidth        10 kHz
Comment         

MINUTE SNR    SYNC   AUDIO    TYPE
  0000   14      110   500/10        0
  0001   15      150   200/10        0
  0002   12      120   150/10        0
  0003   9      149     0/10        0
  0004   14      150   390/10        0
  0005   15      150   810/10        0
  0006   15      150   430/10        0
  0007   17      150 1190/10        0
  0008   16      150   890/10        0
  0009   16      150   770/10        0
  0010   16      150   870/10        0
  0011   15      150   80/10        0
  0012   14      150   80/10        0
  0013   11      150   40/10        0
  0014   10      150     0/10        0
  0015   12      150     0/10        0
  0016   12      139   410/10        0
  0017   14      137   180/10        0
  0018   8      127     0/10        0
  0019   8      122     0/10        0
  0020   10      150     0/10        0

SNR min: 0.0, max: 19.1

CRC:
<<<<

Report log shown graphically using DRMcalc analysis software (written by Carsten Knutter), red line is signal strength and blue line is the number of correctly decoded audio frames -

Bit eXpress didn’t have any QSL cards when I sent the reception report, so they sent me a baseball cap, T-shirt and some pens instead!

Received QSL card 23/9/2005.
 

I made a wav recording of this DRM (not the DRM audio but the DRM 12kHz IF signal as seen in the spectrum display screen) and this was analysed by Christopher at Fraunhofer IIS.

Conclusions are that there was no selective fading (as can be seen above) and no multi-path propagation. As the 4-QAM FAC was received correctly any loss of audio was entirely due to the low SNR and changing the transmission mode from Mode A to Mode B would not make much difference.

What would have improved reception is a higher error protection rate. This particular broadcast was using protection level 2 (31.3 kbps total available). Switching to level 1 would improve reception at the expense of audio quality (down to 26.5 kbps total available).

This demonstrates the trade-off that any DRM broadcaster will have to make between audio quality and coverage area, improve one parameter and you make the other parameter worse.

Despite the flat nature of the OFDM spectrum, and no multi-path, reception was via skywave propagation rather than groundwave. It is very unlikely that a groundwave signal could travel this distance.

www.bitexpress.de (for more information about bit eXpress)

[Thanks to Thomas Bauernschmitt / Christopher Laske for the pictures and information]