Testing the SDRplay RSPdx in HDR Mode on NDB and Broadcast AM DX

One of the selling points of the recently released SDRplay RSPdx is it's special High Dynamic Range (HDR) mode which can be used to improve signal performance for frequencies below 2 MHz. This mode should be especially useful in RF environments where there are strong signals that can overload the SDR and desensitize reception on weaker stations.

Over on YouTube Ivan (aka icholakov) has uploaded a video showing comparisons of signals being received with HDR mode turned on and off. He tests it on weak NDB DX signals, and on medium wave broadcast AM. The results do appear to show that using HDR mode results in an improvement in signal strength.

SDRPlay RSPdx HDR mode on and off - testing Non Directional Beacons and Medium Wave

Subscribe
Notify of
guest

10 Comments
Inline Feedbacks
View all comments
snn47

IMHO it is useless to keep discussing if the contributor of a question does not reply/contribute to our discussion.

My question is why is HDR not be available available above 2 MHz unless it is just some kind of additional amplification below 2 MHz?
– In Direct-conversion SDR the signals are directly converted in an A/D-Converter, therefore the performance of the ADC limits sensitivity, max. frequency or dynamic range. The ADC limits the achieveable dynamic range and in an ADC there should not be any compression or intermodulation like in a conventional (multiple) conversion receier
– RTL-SDR utelize additional non linear circuits in addition to the ADC in which compression, desensitation or intermodulation will ocure.

Any comments?

bert

Still getting signal images on this, two signals the same some khz/mhz apart, one false.

Frederick

You clearly don’t understand NDB modulation. Most are DSB AM using morse as the audio for the modulation. The other sideband is NOT a false image!

Frederick

It would help other users if he had been clear as to what bands he is seeing this as he appears to be referencing NDB reception at LW when he talks about images as this is the subject of this video. In my understanding, HDR mode is only available below 2 MHz. I have seen reports of one user having a massive blowtorch of a signal at around -6 dBm in the medium wave AM band and the device was still able to handle that. If he is getting images at LW, I would be interested to hear. If though he is talking about HF or VHF etc, then as far as I am aware little has changed from previous designs.

snn47

“The results do appear to show that using HDR mode results in an improvement in signal strength.”

Unless the receiver sensitivity varies between HDR on and off, why should the singnal strength increase, unless the propagation attenuation fluctuated between the two settings (likely) or the transmitter EIRP varied due to local changes (weather and varying groundconductivity) ?

radiolistener

usual SDR receivers using extended bits resolution to handle decimation, it allows to get process gain. The noise level is reduced due to bandwidth reduction. It is available in any SDR receiver, but it seems SDRplay just ignore it and has dynamic range lose. And now they fixed that bug

radiolistener

10 dB process gain means about 10x oversampling. For frequencies below 2 MHz it means that ADC works at 20 MHz.

There are two questions:
1) why they fixed it for RSPdx devices, but not for old RSPx devices? Old devices have the same issue and it can be fixed in the same way. So why it was fixed for RSPdx only?

2) why it is fixed for frequencies below 2 MHz? Does it means that for these frequencies it works in direct sampling mode with disabled down-converter?

Timmy

In a ideal mixer 3dB of the RF energy goes to the sum f1 + f2 and 3dB at the difference f1 − f2.
2xOversampling ~ 3dB ~ 0.5 bits on the ADC in terms of dynamic range.
4xOversampling ~ 6dB ~ 1 bit on the ADC in terms of dynamic range.

My uninformed guess is that RSPdx has the exact same low pass filter as the RSPduo and RSP1A with a cut-off at 2MHz, but they have now added more RF switches, so that they can fully bypass the Mirics MSi001 150KHz to 1.9GHz RF tuner chip (that bypass would not physically be in older hardware), feeding the RF signal directly (well after the anti-alias filtering and ESD suppression using TVS diodes) into either the I or the Q ADC input of the Mirics MSi2500 (Dual-channel ADC, IQ DSP, 8051 microcontroller, USB 2.0 HS Interface) and then oversample as much as the ADC can handle and then decimate down to 2MHz of bandwidth. There is no block diagram available for the hardware, that I have found, so I’m only guessing.