THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post information or questions regarding SDRplay products here
glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Tue Jan 08, 2019 8:39 am

3. ANTENNA COUPLING TO THE RECEIVER AND GENERAL PROVISIONS

But there is another sticky point to consider. At this frequency connection of the antenna to the RSP should always be done through a balanced transformer, which removes noisy ground loops and a lot of locally generated noise by isolating the antenna ground from the receiver ground. The balun is also beneficial in protecting the receiver’s input from high voltage induced transients and can also provide impedance matching (even though we need attenuation more than low loss here) between the antenna impedance (in the order of hundreds of Ohms) and the 50 Ohm receiver’s input.

But the balun is a transformer whose input impedance depends on its reactance and hence on its inductance. For a given inductance, which is fixed, the reactance will go down with decreasing frequency, until, at a certain low frequency, the balun will look to the antenna like a short. If one uses a normal HF balun at this frequency it is more than likely that the balun attenuation will increase as frequency goes down, completely distorting the system frequency response and giving misleading results on perceived antenna noise floor, which will appear attenuated with respect to the real level. On the other side of the coin, if balun inductance is too much for the target frequency, its parasitic capacitance will be high and the transformer’s balance will go out of the window, increasing the detrimental effects of local man made noise. Just to have a quantitative idea of the inductance level barely necessary, see the low pass filter post.

Finally all possible LOCAL noise sources feeding the antenna and/or the antenna to receiver connection must be investigated. In general battery operated PC’s are better than mains operated ones; coaxial feed between antenna and receiver should be avoided, if at all possible. Long wire antennas and clean grounds directly connected to the input balun/lowpass filter arrangement are best. With a good antenna an array of attenuators at hand will be very useful: much better receive with least possible loss/mis-match what is available on air and then attenuate BEFORE the front end. This technique has been illustrated here:

https://www.sdrplay.com/community/viewt ... f=5&t=3723

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Tue Jan 08, 2019 8:41 am

4. THE BOTTOM LINE

To go back to the riddle posed by Sabrina1984, surely the following should be investigated:

1) Possible receiver overload in town, where we understand the use of a 160 ft flat top causes severe overloading as indicated by spurious mixing products. Direct feed to the RSP-1A, without balun, seems also likely, a system setup that worsens the situation.
2) We do not know what arrangement is used in the country, so no hypothesis is possible.

Regarding the other important issue being debated, whether or not the RSP-1A is capable of receiving LF signals on par with the RSP2/RSPduo brothers, we can conclude: yes, definitely capable.

On my own RSP’s I have measured the following terminated Noise Floors @ 136 KHz & 1800 Hz Noise Bandwith:

RSPduo 50 Ohm Noise Floor: -108 dBm
RSPduo 1 KOhm Hi Z Noise Floor: -139 dBm
RSP-1A 50 Ohm Noise Floor: -112 dBm

The cited difference in Noise Figure is of no consequence whatsoever in the frequency range under consideration, in the face of a much higher Noise Floor (-70 to -90 dBm) than that of the RSP-1A. This is another confirmation of the competence of the Design Team of SDRplay. The RSP-1A is, in my opinion, a sophisticated and elegant balance between Economy and Design Requirements.




References:
1. from ANNALS OF GEOPHYSICS, VOL. 50, N. 3, June 2007
Natural and man-made terrestrial electromagnetic noise: an outlook
Cesidio Bianchi and Antonio Meloni
Istituto Nazionale di Geofisica e Vulcanologia, Roma, Italy

2. International Telecommunication Union
Recommendation ITU-R P.372-13
(09/2016)
Radio noise
Attachments
RSP's under test.JPG
RSP's under test.JPG (114.66 KiB) Viewed 24494 times

Reason: No reason

sdrom33
Posts: 128
Joined: Sun Dec 09, 2018 4:38 pm

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by sdrom33 » Tue Jan 08, 2019 1:50 pm

Very clear glovisol! Give info about Balun for rsp1a in the photo, thanks

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Tue Jan 08, 2019 7:15 pm

THE PROOF OF THE PUDDING IS IN THE EATING!

- RECEIVER.......................................RSP-1A.
- ANTENNA.......................................135m Beverage.
- IF MODE.........................................Low IF.
- FEED.............................................550 to 50 Ohm Isolation transformer / primary inductance: 3.8 mH.
- ATTENUATOR...................................20 dB type "H", 550/550 Ohm between antenna & Isolation transformer.
- PC................................................IBM X61 laptop with battery supply.


RADIO ALGERIE THIS EVENING ON 531 KHz AND ON 549 KHz. (18:25 GMT)
- RADIO ALGERIE 531 KHz......................Coming in @ - 59.0 dBm + 20 db = -39.0 dBm.
- RADIO ALGERIE 549 KHz......................Coming in @ - 64.9 dBm + 20 dB = -44.9 dBm.
- NOTES...........................................No trace of overload and/or spurious signals.

DDH47 RTTY CALL REPEATER GERMANY ON 147.3 KHz. (20:15 GMT)
- DDH47 @ 147.3.................................Coming in @ -102.9 dBm + 20 dB = - 82.9 dBm.
Attachments
RTTY @ 147.3 KHz.jpg
RTTY @ 147.3 KHz.jpg (178.84 KiB) Viewed 24440 times
Radio Algerie 531 KHz.jpg
Radio Algerie 531 KHz.jpg (181.5 KiB) Viewed 24452 times
Radio Algerie 549 KHz.jpg
Radio Algerie 549 KHz.jpg (190.67 KiB) Viewed 24452 times

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Wed Jan 09, 2019 9:26 am

THE IMPORTANT QUESTION OF RSP OVERLOAD

I must apologise because, being a strong supporter of quantitative information, in this thread I have treated the matter of RSP overload and spurious in generic terms. This could induce groundless alarms. I hope I can be forgiven, because I have already provided comprehensive information on the subject here:

https://www.sdrplay.com/community/viewt ... f=5&t=3926

To summarise the contents of that thread and to restore reasonable reference parameters, it is enough to say that the critical level for all SDRplay RSP processors is in the range of -50 to - 40 dBm and that signals are not cumulative: in other terms, the SDR will whithstand many signals at -48 dBm without overloading. This datum means that the performance of SDRplay RSP processors is quite exceptional, both in theory, considering the wide reception bandwith and in practice, because it does not happen often to meet such signal strengths even at L.F, unless large aperture antennas are used. In this case a 20 dB attenuator in the antenna lead, before the isolation transformer, will suffice, increasing the overload threshold to the safe level of less than - 60 dBm.

The other rules we have seen: no coax lead-in, well designed isolation transformer, low IF mode with low freq. value, low pass filter to reject the Medium Wave/HF 40 m (if local broadcasters are very strong) and battery powered PC. From my experience I can say that such a combination performs as well as conventional front end tuned receivers. At the end of the day the only problem for which there is no remedy is the high level of atmospheric noise at LF, no matter what receiver is used, whether conventional or digital RSP.

Finally, even if the price of SDRplay receivers can be afforded by hobbists in general, never the less the level of specifications and performance delivered is comparable with that of professional equipment: low cost is an added bonus, not a reason to define these products as something more than a toy.
Attachments
Radio Algerie 531(D).jpg
Radio Algerie 531 KHz by day
Radio Algerie 531(D).jpg (185.06 KiB) Viewed 24408 times

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Wed Jan 09, 2019 9:39 am

Hi sdrom33,

Balun data is available here:

https://www.sdrplay.com/community/viewt ... f=5&t=3926

For matching the 500 Ohm long wire antenna impedance to the RSP 50 Ohm input, use 37 turns primary to 8 turns secondary

ENAMELLED COPPER WIRE O.3 MM DIAMETER ALL WINDINGS.


BELOW:
- Radio Algerie @ 252 KHz - 12:40 GMT / left: RSPduo, no attenuation / right: RSP-1A, 20 db attenuation.
- Radio Algerie @ 252 KHz - 12:11 GMT / RSP-1A equipped as in previous posts, 20 dB attenuator included.
- Radio Algerie @ 252 KHz - 12:11 GMT / RSPduo equipped as the RSP-1A in previous posts, 20 dB attenuator included.
Attachments
Radio Algerie 252 comparison.jpg
Comparing RSP-1A to RSPduo
Radio Algerie 252 comparison.jpg (141.56 KiB) Viewed 24378 times
Radio Algerie 252 KHz (D1).jpg
RSP-1A / Radio Algerie 252 KHz
Radio Algerie 252 KHz (D1).jpg (179.75 KiB) Viewed 24382 times
Radio Algerie 252 KHz (D).jpg
RSPduo / Radio Algerie 252 KHz
Radio Algerie 252 KHz (D).jpg (188.93 KiB) Viewed 24384 times

Reason: No reason

sdrom33
Posts: 128
Joined: Sun Dec 09, 2018 4:38 pm

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by sdrom33 » Wed Jan 09, 2019 7:15 pm

Hi glovisol thanks for the info!

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Thu Jan 10, 2019 4:53 pm

- RECEIVER.......................................RSP-1A.
- ANTENNA.......................................135m Beverage.
- IF MODE.........................................Low IF.
- FEED.............................................550 to 50 Ohm Isolation transformer / primary inductance: 3.8 mH.
- ATTENUATOR...................................5 dB
- PC................................................IBM X61 laptop with battery supply.

BELOW

- JJY-40 Time Signal Station Japan @ 40 KHz - 07:15 GMT Today.
Attachments
JJY_Japan, 40 KHz 07GMT.jpg
JJY_Japan, 40 KHz 07GMT.jpg (213.66 KiB) Viewed 24334 times

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Sat Jan 12, 2019 4:35 pm

ASSESSING NOISE LEVELS IN THE LONG WAVE BAND

The LW band is so crowded and some frequencies are so low it is difficult to measure noise floor at a reference bandwith of 1800 Hz, which is the bandwith that has been used for assessing noise floor in HF:

https://www.sdrplay.com/community/viewt ... f=5&t=3685

Assessing the noise floor over a suitable period of time gives us an idea of the quality of our receiving system (receiver + antenna) by comparing the results with data given by the cited ITU publication or measurements by other operators. The bandwith to measure noise floor should be of course a suitably narrow one, such as CW-150 Hz, so we need to convert the measured 150Hz noise floor to the equivalent 1800 Hz noise floor.

The conversion formula is very simple:

Pd = noise floor @ 1800 Hz
P'd = noise floor @ 150 Hz
Pd = P'd + 10.8 [dbm]

The proof is as follows. Rembering equation (5) of the reference cìted above:

Fa = Pd – (10*LOG10(b)) + 173.98 [dB] (5) where Fa is the Noise Figure
Taking:
173.98 dB = a
10*LOG10(150) = h = 21.76
10*LOG10(1800) = r = 32.55

Fa = P'd - h + a
and
Pd = Fa + r - a
substituting
Pd = (P'd - h + a) + r -a
hence
Pd = P'd - h + r
or
Pd = P'd - (10*LOG10(150)) + (10*LOG10(1800)) = P'd - 21.76 + 32.55
but
-h + r = -21.76 + 32.55 = 10.8 [dB]
Finally
Pd = P'd + 10.8 [dBm] Q.E.D.

EXAMPLE: found a noise floor P'd = -121 dBm @ b = 150 Hz; Noise Floor @ 1800 Hz Pd = -121 + 10.8 = -110.2 dBm.

I shall try to record noise floor at the equivalent bandwith of 1800Hz and corrected for the 16 dB negative gain of the Beverage antenna for the next 15 days, three times a day (7:00 / 12:00 / 20:00 GMT) and upload data in this thread. Selected frequencies are (in KHz):

17 - 62 - 120 - 211 - 323

RECEIVER USED: RSP-1A + 550 to 50 Ohm Isolation transformer / primary inductance: 3.8 mH.

I am open to suggestions for alternative/additional frequencies.

Reason: No reason

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

Re: THIS THREAD EXPLAINS WHY IT IS SO DIFFICULT TO UNDERSTAND RECEIVER'S SENSITIVITY IN LF & RELATED MATTERS

Post by glovisol » Sun Jan 13, 2019 7:27 am

ASSESSING NOISE LEVELS IN THE LONG WAVE BAND - FIRST RESULTS


Uploaded below, as an example, table of the measurement results @ 7:00 GMT for the first 3 days. The corrected values will be the most likely Noise Floor values for a rural area for the named LW frequencies, after a relevant number of measurements will be analysed.

ERRATA CORRIGE: the table shows a correction factor of 15 dB for the Beverage antenna. It is a mistake, should be 16 dB. I apologise for the confusion!
Attachments
First LW Noise Floor assssment.jpg
Table of Noise Floor measurements / 3 days
First LW Noise Floor assssment.jpg (67.94 KiB) Viewed 24261 times

Reason: No reason

Post Reply