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

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

Post by glovisol » Sun Jan 06, 2019 3:56 pm

I have followed with interest the debate on the preceding thread:

"RSP1-A Longwave sensitivity" started by Sabrina1984

and have also given some contribution, based on my experience in designing a Low Frequency Low Pass Filter, attenuating the AM Broadcast band, see:

https://www.sdrplay.com/community/viewt ... f=5&t=3943

just to be able to explore this "alien" frequency band. I say alien, because this is what it is: a strange domain where the so called "common sense" and conventional know how in many cases leads one up a wrong path, or worse, to incorrect conclusions. This is so true that several experienced and eminently capable forum dvellers have been unable, not only to explain, but not even to come near to the various problems at hand. The more so, because nobody was able to really explain the probable reason why Sabrina1984 was unable to operate the RSP-1A in the LF band, while, remarkably, two important hints had been given:

1) My uploaded illustration: "High noise in LF" and
2) The post of ON5HB, who, I am happy to say, HIT AT LEAST ONE NAIL SQUARELY ON THE HEAD.

We must surely be grateful to Sabrina1984, who started this because it gave us all not only food for thought, but a chance to increase our knowledge and experience. Surely if Sabrina1984 should intervene again, better describing his or her receivers' plight, we would receive a great help.

We shall start looking (Diogene's lamp!) with the next post....

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

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

Post by Mike2459 » Sun Jan 06, 2019 11:40 pm

From Germany DDH47 on 147.3 khz sending call tape.
Received in Ohio::

<<2019-01-06T23:35Z RTTY @ 147300+0700>>
Q CQ CQ DE DDH47 DDH9:DDH8
FREQUENCIES 147.3 KHZ 11039 KTZH 0&4#7#3 KHZ
RYRYRYRYRYRYJYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRY
CQ CQ CQ DE DDH47 DDH9 DDH8
FREQUENCIOIUBIE KHZ 21039 KHT 14467.3 KHZ
RYRYRZRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRY
CQ CQ CQ DE DDH47DDH9 ($$#8
FREQUENCIES 147.3 (#" 11039 HZ 14:67.3 KHZ
RYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRYRY
CQCQ CQ DE DDH47 DDH9 EDH8
FREQUENCIES 147.3 KHZ 11039 KHZ 14467.3 KHZ
Attachments
DDH47.JPG
DDH47.JPG (113.44 KiB) Viewed 28053 times

Reason: No reason

Roger
Posts: 217
Joined: Fri Dec 29, 2017 7:59 pm

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

Post by Roger » Mon Jan 07, 2019 12:59 am

Mike2459 wrote:From Germany DDH47 on 147.3 khz sending call tape.
Received in Ohio::
<<2019-01-06T23:35Z RTTY @ 147300+0700>>
Q CQ CQ DE DDH47 DDH9:DDH8
Mike -- That is a good catch in Ohio. I seem to recall that you can hear Radio Algeria on longwave once in awhile. These winter DX conditions on LF are interesting.

What antenna, receiver and external devices (trap, preselector, MW filter etc.) did you use to snag DDH47? Did you need much RF gain on the RSP?

Roger

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

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

Post by Mike2459 » Mon Jan 07, 2019 3:54 am

Thanks, Roger

Rcvr: RSP2 on HiZ port

75Ω to 1kΩ wideband RF transformer (6T Pri:24T Sec on a high Permeability core (Al about 12500) salvaged from an old switching power supply. Calculations done with this online calculator: https://www.changpuak.ch/electronics/br ... tching.php
Receiver ground is isolated from antenna ground by the transformer to minimize ground loops.

Antenna: Home built miniwhip (the PA0RDT design with coax shield connected to a good ground at entry to the premises).

I have a 3 pole elliptical lowpass filter but it wasn't connected at the time.

RF gain was 1 notch from max.

These European stations start coming in about 5:30 local time and last till sunrise in the EU. I am also receiving DCF77 on 77.5, DCF49 Manflingen 129.1, HGA22 Lakihegy, HU 135.6 and DCF39 Burg 139.0. The last 3 are used by the German electric power grid management system (SmartGrid) and several of the Longwave broadcasters. Right now I can still hear DDH47 but thunderstorm QRN in the eastern Atlantic and the Caribbean is getting very frequent, tearing up the signal.

Reason: No reason

Roger
Posts: 217
Joined: Fri Dec 29, 2017 7:59 pm

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

Post by Roger » Mon Jan 07, 2019 4:44 pm

Mike,

Thanks for the reply. Sounds like you have a quiet site like mine if you can run the RF gain 1 notch below max on the HiZ port.

I heard that the PA0RDT works well at low frequencies. Do you pickup much man-made noise on the coax shield? Maybe your outside ground helps to lower the level.

Roger

Reason: No reason

fotoralf
Posts: 38
Joined: Wed Aug 08, 2018 10:13 pm
Location: Western Germany, near Belgian border
Contact:

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

Post by fotoralf » Mon Jan 07, 2019 5:38 pm

Roger wrote:I heard that the PA0RDT works well at low frequencies. Do you pickup much man-made noise on the coax shield? Maybe your outside ground helps to lower the level.
I've just spent two weeks near Dunkirk with a RSP2 connected to my Macbook Pro and an original PA0RDT Miniwhip in the garden on a 4 m fiberglass rod. I've had good LF reception while neither the computer nor the antenna were mains fed.

Had the computer running on its own battery and the Miniwhip on a 12 V lead/gel battery. Perfect reception. As soon as the computer was connected to its mains power supply most of the low LF spectrum was drowned in noise.

Ralf

Reason: No reason

Mike2459
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm

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

Post by Mike2459 » Mon Jan 07, 2019 9:36 pm

Roger: Yes, the noise levels were high until I grounded the coax shield. When I did that I saw a 58 volt AC potential between the RSP2 and the coax. The RF isolation transformer blocks the 60Hz leakage current (measured as 500uA) from flowing to ground through the antenna shield by way of the RSP2's chassis through the USB cable to the PC. The home does not have a grounding conductor in the mains wiring, just a hot and a neutral.

fotoralf: I notice no difference in general noise levels when using my laptop just different spurious signals. The biggest improvement in noise reduction was getting the antenna as far away from the house and the PC as possible and grounding the coax shield.

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 8:35 am

PRELIMINARY CONSIDERATIONS

The LOW FREQUENCY band extends from 30 to 300 KHz, covering FOUR OCTAVES or ONE DECADE. To successfully receive signals in this full band, just to see this issue in the proper light, it is like to pretend to successfully receive signals in the range 30 to 300 MHz using exactly the SAME EQUIPMENT AND THE SAME ANTENNA: obviously not impossible, but surely very difficult technically.

To gain an even clearer perspective, the received signal wavelength ranges from 10 kilometers to 1 kilometer, a 1000 percent variation! If we compare this range with the popular 40 m band, 7000 to 7300 KHz, we have to accommodate a wavelength variation of only 4.2%, or from 41 to 42.8 m.
Within the LF band we have the 2200 m puny Radio Amateur Segment from 135.7 to 137.8 KHz, or a wavelength variation of 0.023% from 2177 to 2182 m. A small segment, really, but a very long wavelength for sure, which would logically require a receive antenna length in proportion.

If we seriously wish to understand what is happening when we attempt to set up reception at LF, we must look at the readily available physics and data not in a generic and qualitative way, but in a precise and QUANTITATIVE WAY. Just to make an example, if we generically say that “noise in LF” is very high in general, but we fail to state THE ACTUAL LEVEL of that noise, we are just deluding ourselves about our “expertise”, but in reality we should admit we know nothing or very little. In the same train of thought, if we take an RSP-1A and we stick a piece of wire to the input connector and by sheer chance we get a signal in the LF band, then we immediately declare that that receiver is tops, while, on the contrary, if we get only noise and squeaks, we classify the receiver as a brick and publicly complain.

The fact is that the devil is not only in the details, which are seldom mentioned and considered, but also in the lack of willingness to look at the huge mass of knowledge which is generally available and which can help us in setting our path.

To really understand what is going on in LF, we need to QUANTITATIVELY consider and analyse at least the following issues pertaining exclusively to this band. This we are going to do in the following sections, concentrating our attention to the frequency range 100 to 300 KHz.

1. PROPAGATION, ATMOSPHERIC NOISE and MAN MADE NOISE
2. LF & MW TRANSMITTER POWERS AND ANTENNAS
3. ANTENNA COUPLING TO THE RECEIVER & GENERAL PROVISIONS
4. THE BOTTOM LINE

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 8:37 am

1. PROPAGATION AND ATMOSPHERIC NOISE

What is Atmospheric Noise? Mr. Marconi conceived and built his first radio transmitter as a SPARK GENERATOR. Atmospheric noise is also produced….by sparks, these being lightning flashes all over the world. This text taken from "references" in the last post.
Lightning flashes are the main source of energy for the electromagnetic background inside the ionospheric cavity. Starting from the lower band ELF (few Hz) up to several VHF (hundreds MHz) the noise originates mainly from the energy radiated by lightning strokes (see for example: Cummer and Inan, 2000; Mika et al., 2005). Several million lightning strokes occur daily from an estimated 2000 storms worldwide, and the Earth is hit about 100 times a second by lightning. The discharge is very violent and can easily reach 10000 A. The amount of energy released by each discharge can vary from units to tenths of GJ. Hence, for the duration of the discharge (less than 1 s), the power involved in this phenomenon is of the order of 1-10 GW. The annual total released energy is in the order of 1019J. If only 10% of this energy is radiated as electromagnetic energy, it would be comparable to the energy produced in 1970 by the electric power stations in the world.


Figure 1 shows a typical PC screen taken with the RSPduo equipped with the cited Medium Wave elimination low pass filter and the 135 m Beverage antenna connected to the Hi Z input. Referring to the graph on Beverage Antenna Gain Vs. wavelength ratio, uploaded here:

https://www.sdrplay.com/community/viewt ... 5&start=20

we see that for the very low ratio of: 135/2200 we can expect a negative gain of approx. 16 dB. This means that the Noise Floor of -96.1 dBm shown in Figure 1 must be corrected by 16 dB to an approximate value of –80 dBm. This is what I get, day in, day out in wintertime and with good weather and shown in Figures 1 and 2 below. But this measurements would mean very little with no general confirmation.
And confirmation is given by our old friend, ITU publication ITU-R P372-13. This publication provides the quantitative data about LF band noise in the form of Figure 3, which I have scripted for immediate evaluation. Figure 3 shows that at 140 KHz the Noise Floor to be expected is in the range of -81 dBm (best value) to -6 dBm (worst value) with a most likely value of -61 dBm in a quiet rural area. Looking at this data it is immediately evident that my measured value of -80 dBm drops smack into the ballpark. In figure 3 the External Noise Figure Fa in dB is converted to Noise Floor in dBm by the expression:

Pd = Fa + (10*LOG10(b)) – 173.98 [dBm] with detection bandwith b = 1800 Hz

Already mentioned in the post: "ASSESSING THE PERFORMANCE OF THE RSP RECEIVER SYSTEM IN HF"

https://www.sdrplay.com/community/viewf ... 5&start=25

But what about Man Made noise? Again looking at Figure 3, curve C, we see that its contribution, compared to that of Atmospheric Noise, is far less important, being very local, nearby noise relatively easy to eliminate, provided you use a balanced balun at the input and other provisions, as covered in PART 3 following. I can see many of the readers shaking their heads, thinking I have gone out of mine. They are thinking: “how about my own experience? How about all the guys on the Forum claiming they experience very low noise levels (-130 dBm perhaps) because they live in remote, low noise, rural areas, etc.? Surely glovisol and the ITU engineers have taken a big lemon….”.


The story is very simple: if you experience a low noise floor not in accordance with the ITU data it means that your receiving system has some kind of loss that attenuates the atmospheric noise pervading the entire ionospheric cavity; in other words you are deluding yourself with a pretty deaf receiving system. How come then, that useful signals can still be received? This apparently inexplicable paradox will be explained in the next two chapters.
Attachments
F 1 night.jpg
Figure 1 - LF band noise at night
F 1 night.jpg (185.87 KiB) Viewed 27902 times
F 2 day.jpg
Figure 2 - LF band noise during the day
F 2 day.jpg (190.98 KiB) Viewed 27902 times
ITU LF noise graph F3.jpg
Figure 3 - ITU noise data graph
ITU LF noise graph F3.jpg (88.19 KiB) Viewed 27902 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 » Tue Jan 08, 2019 8:38 am

2. LF & MW TRANSMITTER POWERS AND ANTENNAS

In the beginning I mentioned that at LF we are in a strange, more, in an alien territory, where often a cause does not have the expected effect. We must realize that in LF are operating the highest power radio transmitters available in the world. In no other frequency range such powers are used and we are talking of many transmitters with antenna powers of up to 200 kW between 180 and 280 KHz. These powers are used in order to serve listeners, equipped with cheap portable receivers with diminutive antennas, in order to overcome an all pervasive noise cloud travelling around the Earth. See Figure 4 below.

Now imagine what can happen in such an environment to a sensitive RSP receiver connected to a substantial antenna: the receiver overloads, no matter what engineering arrangements it is based on and the only remedy is ATTENUATION IN THE ANTENNA LEAD (yes, you could use a crystal filter, but unfortunately tuneable crystal filters have still to be invented).

This environmental shortcoming has been very well described three days ago by ON5HB here:

https://www.sdrplay.com/community/viewt ... 8&start=20

The reverse of the coin is to use a minimal antenna. We have a very recent example on this thread contributed by Mike2459 who received German DDH47 on 147.3 KHz using a home made whip antenna,consisting of a loop printed on a PCB followed by a transistor amplifier. So we come to the paradox that the lossier the receiving system the more chances to successful reception. On the other hand at these frequencies radio signals propagate with minimal loss, as calculated by the frequency dependent propagation loss formula, this being the reason why very low power can do a lot of road, provided it is connected to a big transmit antenna.

When all is said and done, we still must not forget we are operating an extremely wideband receiver in a very high level signal environment. Conventional receivers normally have single or double tuned bandpass filters to protect the front end from interference/overload by near frequency strong signals. The same result can be obtained with RSP’s in the 2200 m band by inserting a simple fixed bandpass filter with a 135 to 138 KHz bandpass. I plan to post the complete design for such a filter soon.

We must not overlook the issue of the powerful Medium Wave transmitters, which are more likely to cause receiver overload and spurious for the simple reason that the probability of being near a MW transmitter site in metropolitan areas is much higher than that of being near a Shortwave one. This issue has been covered already in the cited thread on an ad hoc Lowpass Filter: all RSP’s missing one should be equipped with an external unit and operated in Low IF mode at LF.
Attachments
F 4 Powers.jpg
Figure 4: World transmitter powers from cited reference
F 4 Powers.jpg (37.29 KiB) Viewed 27901 times

Reason: No reason

Post Reply