Receive Antenna & Impedance "Matching"

Useful information regarding antennas for SDR products.
glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Receive Antenna & Impedance "Matching"

Post by glovisol » Thu Jul 25, 2019 6:20 pm

CLARIFICATION ON SENSITIVITY PERFORMANCE COMPARISON RSP-1A VS. RSPduo & FOCUS ON NEGATIVE S/N

9. Looking at WSPR reception @ 7 MHz from the perspective of RSP Processors - an example
3-3. Noise meas. 7.2 MHz.jpg
15 days Noise floor measurements @ 7 MHz in December 2018
3-3. Noise meas. 7.2 MHz.jpg (146.82 KiB) Viewed 47433 times


This post is an explanation of the calculation/detection mechanics involved in very weak signal reception such as that of the WSPR signals.The above diagram shows the average noise floor of my 140 m Beverage antenna for 15 days in December 2018. In June 2019, as described here:

https://www.sdrplay.com/community/viewt ... f=8&t=4493

the antenna was modified, brought to 160 m and terminations optimised, so that an average of 6 dB noise improvement on Noise Floor was obtained. Thus I could expect to have an average antenna Noise Floor of - 111 dBm @ 7 MHz.

Keeping in mind the specified Noise Figures shown in the previous post, measurement of the terminated noise floors of my own RSP-1A and RSPduo yelded the following results @ 1800 Hz (LSB) detection bandwidth:

Detection bandwith factor = 10*LOG10(1800) = 32.5 dB
RSP-1A, 50 Ohm input: N. Floor = -123.5 dBm....Noise Figure = 174 - (123.5 + 32.5) = 18 dB (as per spec.)
RSPduo, HI Z input: N. Floor = -127.5 dBm........Noise Figure = 174 - (127.5 + 32.5) = 14 dB (better than rated 16.65 dB)

The reference bandwith for the WSPR, e.g. the reference at which the S/N ratio is measured, is 2500 Hz, thus we may have a measurement error of approx. 1.5 dB by using the reference value of 1800 Hz. This is not so important, considering that the final autocorrelated detection bandwidth of the WSPR signal is 8 Hz. The worst case S/N value of WSPR signal is -28 dB, that is the WSPR signal that will be detected with an S/N value of -28 dB or better, referred to the bandwidth of 2500 Hz.

If now our antenna Noise Floor is -111 dBm, the minimum detectable WSPR signal (MDS) will be:

MDS = -111 - 28 = - 139 dBm

Our RSP Spectrum Processor, in order to be able to receive the Minimum Detectable WSPR Signal, will need, at the antenna input a level of:

MDS + NF = -139 + 18 = -121 dBm in case of the RSP-1A
and
MDS + NF = -139 + 14 = -125 dBm in case of the RSPduo

Considering that our antenna noise floor is -111 dBm, the above calculated WSPR signals will be far buried into the noise of the SDRuno dial and therefore invisible. This is why, for WSPR operation, the frequency precision and stabilty of the SDR Processors are essential and indispensable for successful reception.

All the above we have seen is just an example based on my Beverage antenna performance data. If the antenna Noise Floor was, for example, -125 dBm, then with an RSPduo one could detect a WSPR signal at a level of -139 dBm.

In any case, as shown in the thread about "Getting ready for Diversity reception", with RSPduo in Diversity operation, I have uploaded an example of full copy of an SSB signal at - 117 dbm with the same Beverage plus an auxiliary 60 m long wire antenna.

https://www.sdrplay.com/community/viewt ... f=7&t=4731

This example promises a significant advance in WSPR reception using Diversity operating mode.
Last edited by glovisol on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

SDRguy22
Posts: 12
Joined: Fri Jul 19, 2019 4:34 pm

Re: Receive Antenna & Impedance "Matching"

Post by SDRguy22 » Thu Jul 25, 2019 6:43 pm

I have some links to reference material on the concept of Noise Figure (NF) and how atmospheric and man-made noise is measured. I hope readers find it useful for understanding some of the issues being discussed here.

There are many tutorial articles and books that discuss the concept of Noise Figure and how it is used. The following article from from Wikipedia is quite informative and I suggest readers unfamiliar with this subject take a look at this article. https://en.wikipedia.org/wiki/Noise_figure

The term Antenna Noise Figure (Fa) is a related concept and is very useful when discussing noise that is received on any antenna. Natural noise (galactic, lightning, atmospheric etc.) and man-made noise can significantly affect reception. Characterizing this noise and measuring the level at different locations, time and in different seasons has been a subject of study for many years. An updated version of ITU Publication 372 which discusses this topic is freely available for download here: https://www.itu.int/rec/R-REC-P.372-13-201609-I/en


Bill
Last edited by SDRguy22 on Fri Jul 26, 2019 3:41 pm, edited 1 time in total.

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Receive Antenna & Impedance "Matching"

Post by glovisol » Thu Jul 25, 2019 7:22 pm

Dear Bill,

I indeed welcome your tractation and I am glad to have found a Forum member founding his Antenna Noise analysis on the notorious ITU publication 372. I really would like you to examine my one year old work and Antenna measurement cycle based on the mentioned publication here:

https://www.sdrplay.com/community/viewt ... ing#p12602

Kind regards,

Gianfranco
Last edited by glovisol on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Receive Antenna & Impedance "Matching"

Post by glovisol » Thu Jul 25, 2019 8:17 pm

CLARIFICATION ON SENSITIVITY PERFORMANCE COMPARISON RSP-1A VS. RSPduo & FOCUS ON NEGATIVE S/N

10. Conclusion

The theory and practice of the performance analyses we have seen in the above titled series of posts can be found in these two basic studies available on this Forum:

ASSESSING THE PERFORMANCE OF THE RSP RECEIVER SYSTEM IN HF
https://www.sdrplay.com/community/viewt ... ing#p12602

and the companion, showing, among a wealth of additional data, practical long wire antenna noise floors measured for extended periods of time:

RSP RECEIVER LOW NOISE HF OPERATION WITH LONG WIRE ANTENNAS
https://www.sdrplay.com/community/viewt ... f=5&t=3436

These two studies, based on the important ITU Recommendation ITU-R P.372-13, are the foundation for modern HF receiving system performance assessment, based on field measurements and data analyses, which yeld significant and immediate practical results.
Last edited by glovisol on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: Receive Antenna & Impedance "Matching"

Post by glovisol » Sat Jul 27, 2019 4:04 pm

CLARIFICATION ON SENSITIVITY PERFORMANCE COMPARISON RSP-1A VS. RSPduo & FOCUS ON NEGATIVE S/N

11. APPENDIX - Looking at WSPR reception @ 14 MHz from the perspective of RSP Processors
4-3 Noise meas. 14 MHz.jpg
15 days Noise floor measurements @ 14 MHz in December 2018
4-3 Noise meas. 14 MHz.jpg (111.22 KiB) Viewed 47292 times

With the cited modification the Beverage antenna Noise Floor increased less than at the lower frequencies, but still an improvement of 3 dB was obtained, bringing the expected average Noise Flor at - 115 dBm. Terminated Noise Floor at 14 MHz of my own RSP-1A and RSPduo yelded the following results @ 1800 Hz (LSB) detection bandwidth:

Detection bandwith factor = 10*LOG10(1800) = 32.5 dB
RSP-1A, 50 Ohm input: N. Floor = -126.1 dBm....Noise Figure = 174 - (126.1 + 32.5) = 15.4 dB (Spec. 15.2 dB)
RSPduo, HI Z input: N. Floor = -126.9 dBm........Noise Figure = 174 - (126.9 + 32.5) = 14.6 dB (Spec. 14.8 dB)

We repeat here: reference bandwith for the WSPR, e.g. the reference at which the S/N ratio is measured, is 2500 Hz, thus we may have a measurement error of approx. 1.5 dB by using the reference value of 1800 Hz. This is not so important, considering that the final autocorrelated detection bandwidth of the WSPR signal is 8 Hz. The worst case S/N value of WSPR signal is -28 dB, that is the WSPR signal that will be detected with an S/N value of -28 dB or better, referred to the bandwidth of 2500 Hz.

If now our antenna Noise Floor is -115 dBm, the minimum detectable WSPR signal (MDS) will be:

MDS = -115 - 28 = - 143 dBm

Our RSP Spectrum Processor, in order to be able to receive the Minimum Detectable WSPR Signal, will need, at the antenna input a level of:

MDS + NF = -143 + 15.4 = -127.6 dBm in case of the RSP-1A
and
MDS + NF = -143 + 14.6 = -128.4 dBm in case of the RSPduo

We can conclude that long distance, WSPR reception of QRP signals has better chances of success in the 20 m band, because, while Noise Floor level is generally lower than at 40 m, the antenna dimensions are halved and it is easier for the average operator to approach this new mode. Additionally the Noise Figure advantage of the RSPduo over the RSP-1A on 20 m amounts to 0.8 dB only, so there is practically no difference between the two processors' performance.
Last edited by glovisol on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Post Reply