Tagged: srm


Receiving DRM broadcast with SDR radio receiver

Follow Quadrus SDR

Digital broadcast on HF bands

Due to significant progress in Digital Signal Processing (DSP) technology, it is now possible to use spectrum efficient digital transmissions for high quality audio broadcast in HF bands. The current and already stable standard is called Digital Radio Mondial (DRM). This transmission is using the same 10 kHz bandwidth as a traditional AM broadcast station, although with a multi-carrier modulation format, and delivers content as source coded streams. The spectrum of such transmissions is typically noise-like and rectangular, as you can see on the screen of an SRM-3000 SDR radio receiver.

DRM signal on SDR radio receiver

Demodulation and decoding of DRM signals

We need a special DSP tool to demodulate such waveforms with several sub-carriers and QAM modulation of each carrier. Fortunately, we have an open source community project delivering such a software, called Dream DRM receiver.  You can download the latest version of their software from its SourceForge repository. The digital stream is provided by the demodulator. However, we need to use the decoder to generate an audio sample and other meta info associated with the transmission, i.e., station name, program characteristics, etc. The DRM stations use AAC+ codec. It is built in the Dream software stack as a dynamic library.

Connecting the SDR radio receiver station components

We can use the DRU-244A digitizer card with SRM-3000 sdr radio receiver software to receive the signals from the air. The Dream software is used to demodulate the stream and to decode the audio signal. Both of them are able to feed signals to and consume signals from a sound device. A jumper cable can be used to connect the two applications. However, that’s not recommended for analog conversion and back. A virtual audio cable may be used instead, which directly forwards samples from one app to the other. I’ve used the VB-Audio Virtual Cable for my test runs. See their web site for more information.


Enjoy DRM programs form world wide providers

Finally, we need to find some DRM signals on the air, or look after program guides on the net for given stations. I simply search the spectrum. It was easy to recognize some DRM signals using the SRM-3000 SDR radio receiver. One of the first transmissions I was able to find was from All India Radio, which is pretty good DX from Europe with a simple long wire antenna. We have a enough receiver station sensitivity with the DRU card and the antenna for such a DX reception.

drm01 drm02

Share Quadrus SDR
RF input response for a -120dBm input signal level

Direct digital UHF SDR radio receiver with DRU-244

Follow Quadrus SDR

Direct digital receiver

A direct digital receiver has the Analog-to-Digital Converter (ADC) directly connected to the incoming RF signal without any frequency translation. This is contrary to the superheterodyne or direct conversion methods, which translate the RF signal to an Intermediate Frequency (IF) or the BaseBand (BB) respectively before digitizing. The direct digital receiver concept can be regarded as an example of the ultimate software defined radio.

Instantaneous versus input bandwidth

Instantaneous bandwidth is the frequency band that is continuously processable by the digitizer device. This bandwidth is determined by the sampling frequency; half of the sampling frequency is often called the Nyquist-frequency. The maximum bandwidth of the processable signal should be less than this Nyquist-frequency in accordance with the Nyquist-Shannon sampling theory. However, the input bandwidth is determined mainly by the the analog front-end and the sample-and-hold circuit in the ADC. If the limit imposed by these circuits is higher than the Nyquist-frequency, we have a chance to sample higher frequency signals as well. This is usually called under sampling or sub-Nyuist sampling.

Input bandwidth of the DRU-244

I have an older version of the DRU-244 digitizer board. The input bandwidth was not specified, but it should be up to a few hundred MHz. Maybe even as high as 144 MHz or 432 MHz. I’ve connected the 432 MHz output of the signal generator to the input of the SRM SDR radio receiver, and I’ve tuned to the same frequency to get a good look at the signal. I’ve observed that the level is more than 20 dB less than in the HF band. So, I need at least 20 dB preamplification to maintain the sensitivity in the UHF band.

Sensitivity testing in the 432 MHz (70 cm) band

First, I’ve checked the sensitivity with my FT-897 transceiver. I’ve connected a signal with a -120 dBm output power, which had a well audible sound level employing the Singel SideBand (SSB) demodulator.


Next, I’ve connected two MiniCircuits ZX60-6013E amplifiers in cascade providing 30 dB amplification with a reasonable noise figure.


I’ve checked the input noise level of the receiver without the connected preamp.


Then, I’ve checked again with the preamp. The noise floor increased by a couple dBs.


This was a good indication that the system sensitivity was determined by the preamp as opposed to the digitizer in the receiver.

I’ve connect the -120 dBm signal to the input, which was unfortunately less audible after the SSB demodulation then with the FT-897.


The Signal-to-Noise Ratio (SNR) at AF level was not sufficient.


437 MHz falls into the 11th Nyquist band of the converter. My idea was that all of the preamp output noise ended up getting aliased into the baseband, and consequently reduced the SNR. So, I’ve tried a Band-Pass Filter (BPF) at the preamp output, before the digitizer input. It helped a lot, and the SNR increased substantially.


I had a really good, clear audio signal without any noise. The next picture shows the audio output spectrum of the direct digital SDR radio receiver with DRU-244 running the SRM radio receiver software for the -120 dBm input signal at 437 MHz.




Usually, modern ADCs have significant input bandwidth, and allow sampling in higher Nyquist bands. This way direct digital VHF/UHF radio receivers can be built with simple architectures. However, input signal degradation should be mitigated with input preamplifiers. Although, we loose some dynamic range, this is an acceptable price for a very simple receiver architecture.

PS: Why 437-450MHz? This is the down-link of the MASAT-1 satellite. So, I guess, now you know my next plan… :-)

Share Quadrus SDR

Audio output from the SRM SDR receiver

Follow Quadrus SDR

Audio output

The SRM-3000 SDR receiver can handle up to 16 independent receiver channels, but usually, we only have one sound card output in the PC. The radio receiver software defines one channel as monitored, and sends its audio output to the default audio out device. This way, you can listen to an active channel with your speakers or headphone.

Processing the audio output

Usually, after the filtering and demodulation, we want some additional post-processing of the audio output. The post-processing software can run on the same or on a separate computer. In the latter case, we can directly connect the audio out to the microphone input of the other PC using a simple jumper cable with two standard 3.5 mm jack plugs. In case we use the same computer for post-processing, we can use a virtual audio cable. This is a software that defines virtual output and input devices. It can be set as the default audio device, and thereby act as a gateway: the SRM SDR receiver can send its audio output to it, which the post-processing software can directly receive.


Testing with different post processors and decoders

First, I used an external recording software to save audio output for later processing. I’ve employed Audacity, which is a free, open-source audio editor and recorder software for a variety of platforms.

srmin auda

After that, I’ve successfully tested the connection with the well-known Spectrum Lab post-processor developed by DL4YHF.


Finally, I’ve tested the demodulation possibility for the audio output by utilizing the Code300 software package from Hoka Electronic. The audio signals contained a real FSK transmission, and it was successfully demodulated and processed.




Share Quadrus SDR

Wideband SDR reception in the SRM SDR software receiver

Follow Quadrus SDR

[Wideband SDR available bandwidth settings]

The SRM-3000 radio receiver SDR software supports the following six built-in bandwidths: +/- 6.25, 12.5, 25, 50, 100, 200 kHz. These are Single Side Band (SSB) specifications. So, for example, the widest (200 kHz) setup actually enables the reception of 400 kHz wide transmissions, as all the signals are complex. This is seen on the spectrum/waterfall display, which has a frequency range going from -200 kHz to +200 kHz in this case. The wideband SDR software radio receiver SRM automatically sets the Digital Down Converts (DDCs) on the DRU digitizer to the desired Nyquist frequency. But this DDC output sample rate is NOT equal to the bandwidth specification above! In fact, the sample rate is much higher, thus the system can deliver unattended performance within the useful bandwidth. So, for example, in case of the 200 kHz bandwidth setup, the actual DDC output sample rate is 398 kSPS – almost double. This overhead allows to have a magnitude drop of less than 1 dB within the useful ±200 kHz bandwidth.

200 kHz is wide enough to display and receive WB FM broadcast transmissions. As the input frequency range of the DRU-244A digitizer hardware can go up to 320 MHz, we just need to tune the radio receiver to the desired station.

FM_broadcast_if_spectrum400 FM_broadcast_if_waterfall

Thanks to the high SNR the large amplitude 19 kHz stereo pilot is easily identifiable in the audio spectrum. The audio processing runs at 48 kSPS, thus the difference channel is not processed for the time being. Full stereo decoding is an upcoming feature.


Recall that the display only contains the 1 dB band of the DDC output, which runs in the example at 398 kSPS if IQ pairs are counted as one sample, or at 796 kSPS if counted as two. This is a significant burden on the subsequent DSP processing, which is entirely performed by the x86 Intel GPP in the PC. The processor workload is very high, so you need a powerful machine (i5 or preferably i7) to process one or more wideband radio channels. On the other hand, WB FM broadcasts may be processed in the 100 or 50 kHz bandwidth mode as well due to the robust nature of FM modulation.

FM_broadcast_if_spectrum200 FM_broadcast_if_spectrum100a

The wideband radio receiver is not only for demodulating WB FM broadcasts. The receiver may be tuned to any band of interest. E.g., the following images show the 20 m HAM radio band; the center of the ±50KHz window is at 14.050MHz. In this case, the demodulator offset functionality may be used to select and demodulate the narrowband signal of interest within the available bandwidth. This involves mixing, band filtering, and CW/USB demodulation – all performed on the PC.

20m_band_if_spectrum 20m_band_if_waterfall

Finally, a note on safe device handling. Direct digital radio reception at wide input bandwidths (i.e., 320 MHz for the DRU-244A digitizer board) necessitates the use of input preselection filters, in order to avoid overload conditions for the sensitive receiver input circuitry. To design and simulate such filters, you may use the excellent free filter design program DIONYSUS by ComNav Engineering.

Share Quadrus SDR