RSP1A Longwave sensitivity

Post information or questions regarding SDRplay products here
Roger
Posts: 217
Joined: Fri Dec 29, 2017 7:59 pm

Re: RSP1A Longwave sensitivity

Post by Roger » Thu Jan 03, 2019 12:33 am

sdrom33 wrote:Hi z inputs cannot be compared with 50 ohm in of rsp1a. Better compare 50 ohm of all sets to have best idea.
Sabrina1984 was wondering why the performance of the RSP1A was not doing as well as the NRD525 and looking for tips on how to improve performance. I made a few suggestions like using Low IF, Setting the LO to 1000 kHz and using external MW reject filters and preselectors.

However there is only so much you can do with a receiver that has a high noise figure (NF) below 500 kHz when you use a poor antenna. The lower the Noise Figure the better when looking for weak signals under these conditions. SDRplay has published data that shows the 50 ohm ports of the RSP1A, RSP2 and RSPduo have about a 10 dB higher noise floor than the HiZ port of an RSP2 or RSP duo at 1 MHz. The NF will be even worse below 500 kHz. Here are some specs comparing the RSPduo with the RSP1A. This data is from the SDRplay data sheets for each product.


Noise Figure @ 1 MHz. - with maximum gain
RSP1A 21.2 dB
RSPduo Tuner 1 SMA 20.77 dB
RSPduo Tuner 2 SMA 19.33dB
RSPduo Tuner 1 HiZ 11.66 db ****

Sensitivity for AM signal @ 136 kHz, - with 60% Mod - 12 dB SINAD
RSP1A 6.02 microvolts
RSPduo Tuner 1 SMA 6.16 microvolts
RSPduo Tuner 2 SMA 5.36 microvolts
RSPduo Tuner 1 HiZ 0.37 microvolts ****

Sensitivity for SSB/CW signal @136 kHz, - 1.5 kHz. BW - 12 dB SINAD
RSP1A 2.26 microvolts
RSPduo Tuner 1 SMA 2.29 microvolts
RSPduo Tuner 2 SMA 1.99 microvolts
RSPduo Tuner 1 HiZ 0.18 microvolts ****

**** HiZ port is 1000 ohms but microvolt sensitivity has been converted to 50 ohm

SDRplay do not publish detailed NF specs for the RSP2 but in my experience I find slightly worse performance than the RSPduo but better than the RSP1A.

The RSP1A is a good SDR receiver but it has some limitations compared to the RSP2 and the RSPduo. It is the lowest cost member of the family but offers great value. For serious LF DXers the duo is the best choice. The HiZ port and the built-in MW rejection filter give the best weak signal reception below 500 kHz.

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by glovisol » Thu Jan 03, 2019 3:25 pm

Noise Figure @ 1 MHz. - with maximum gain
RSP1A 21.2 dB
RSPduo Tuner 1 SMA 20.77 dB
RSPduo Tuner 2 SMA 19.33dB
RSPduo Tuner 1 HiZ 11.66 db ****
These noise figures are not useful unless the related detection bandwith is given, because they cannot be compared to data below.

Sensitivity for SSB/CW signal @136 kHz, - 1.5 kHz. BW - 12 dB SINAD
RSP1A 2.26 microvolts
RSPduo Tuner 1 SMA 2.29 microvolts
RSPduo Tuner 2 SMA 1.99 microvolts
RSPduo Tuner 1 HiZ 0.18 microvolts ****
RSP1A.............................2.26uV @ 50 Ohm / 1.022*10^-13W / - 99.9 dBm
RSPduo Tuner 1.................2.29uV @ 50 Ohm / 1.049*10^-13W / - 99.8 dBm
RSPduo Tuner 2.................1.99uV@ 50 Ohm / 7.92*10^-14W / -101 dBm
RSPduo Tuner HI Z..............0.18uV @ 1000 Ohm / 3.24*10^-17W / -135 dBm

Looking at these figures one sees that the sensitivity advantage between High Z and 50 Ohm inputs would appear to be in the order of 35 dB and not of 10/12 db, as would look more likely and confirmed by experience. Furthermore I am not sure that the RSPduo or (even more unlikely) the RSP2 should deliver such performance at LF, nor convinced that a sensitivity of -135 dBm would make any difference (tests pending) at a receive frequency of 300 KHz, because of antenna noise, see PC screen below taken from:

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

My preliminary humble opinion is that the main problem of Sabrina lies with antenna coupling to the receiver and spurious generation due to strong MW Broadcasting stations.
Attachments
MW blotted out.jpg
High noise in LF
MW blotted out.jpg (196.33 KiB) Viewed 26353 times

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by Roger » Thu Jan 03, 2019 8:50 pm

In my post above I gave the following specs for sensitivity below 500 kHz. for the RSP1A and RSPduo

Published Sensitivity for SSB/CW signal @136 kHz, - 1.5 kHz. BW - 12 dB SINAD
RSP1A SMA 2.26 microvolts
RSPduo Tuner 1 SMA 2.29 microvolts
RSPduo Tuner 2 SMA 1.99 microvolts
RSPduo Tuner 1 HiZ 0.18 microvolts **** HiZ port is 1000 ohms but microvolt sensitivity has been normalized to 50 ohm

In order to verify these figures I did some tests on my own equipment using a signal generator.
LO locked at 1 MHz. and Low IF enabled. The results are as follows:

Tested Sensitivity for SSB/CW signal @328 kHz, - 1.8 kHz. BW - 12 dB SNR
RSP1A SMA 1.86 microvolts (-101.6 dBm)
RSPduo Tuner 1 SMA 1.97 microvolts (-101.1 dBm)
RSPduo Tuner 1 HiZ 0.16 microvolts (-123.2 dbm) **** Normalized to 50 ohm

Results were very close to those published by SDRplay

You can see screenshots showing test results on my Google Drive >> https://drive.google.com/open?id=1tmoS3 ... NeSfi9Dy78

Tests were also done using on-air signals and these are in the next post...

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by Roger » Thu Jan 03, 2019 9:39 pm

In my previous post I showed that the RSPduo performs better than the RSP1A in the LF band under lab conditions. How do these receivers perform when connected to an outside antenna?

I live in a semi-rural area and the nearest FM or AM radio station is many kilometers away. There is very little man-made noise at this location. So it is a good place to do LF DXing. I do receive a number of aviation non-directional beacons (NDB) on my doublet antenna which is not optimum for LF by any means.

In this test an RF signal combiner was used so that I could connect the antenna and a signal generator to the input of an RSP1A and a RSPduo HiZ port at the same time. This way it was possible to inject a signal from the RF generator while receiving band noise and signal activity. The screenshot below shows some local NDB's being received on an RSP1A. LO was locked at 1 MHz. and Low IF enabled. I tuned to a vacant channel so that I could measure the background noise in 1.8 kHz. Then I turned on the signal generator and adjusted it to get a 12 dB signal-to-noise-ratio (SNR). It required -103.2 dBm (1.58 microvolts) to get the desired SNR. This is shown in the lower spectrum display.
RSP1A - with antenna.png
RSP1A - with antenna.png (378 KiB) Viewed 26338 times
The same test was done with the RSPduo on the HiZ port using a balun designed to convert 50 ohms to the HiZ port 1000 ohms. Then a signal was injected into the RF combiner to get the desired 12 dB SNR. The RSPduo HiZ port required -118 dBm to get an SNR of 12 dB. The results are shown below.
RSPduo - with antenna.png
RSPduo - with antenna.png (381.3 KiB) Viewed 26338 times
Conclusions

- Under these test conditions one can clearly see weaker signals are now being received with the RSPduo.
- The noise floor is lower because the RSPduo has a better Noise Figure on this band.
- The difference in power required for the same SNR is 14.8 dB (-103.2 dBm - -118 dBm) at the test frequency

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by Mike2459 » Fri Jan 04, 2019 2:55 am

Perhaps a RF isolation/impedance matching transformer in the antenna feedline would be a good starting point. The reason I mention this is because I measured a 60VAC 60Hz @ 500 µAmps between my PC and via the USB cable of my RSP2 to the antenna ground. A 4:1 (for the Hi-Z port) isolation transformer stopped the current flow and reduced noise levels as well on my RSP2. I recommend anyone with an SDR that uses an antenna ground to check the voltage between their equipment and earth ground. Leakage currents can be a real pain and even dangerous in some cases.

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by sdrom33 » Fri Jan 04, 2019 6:48 am

Guys I want to buy a Rsp and now I am confused: must I get the 1a or the duo? For glovisol the duo is more sensitive by 35 db, very good. For Roger it is 123-101 = 22 db, good in one post and now is 14 db, not so hot in the last post. But the 1a is far cheaper. Can someone explain and help?
Last edited by sdrom33 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: RSP1A Longwave sensitivity

Post by glovisol » Fri Jan 04, 2019 12:58 pm

Hi Sdrom33,

The sensitivity figure of -135 dBm calculated by me is wrong and is due to a misunderstanding on the receiver's impedance level taken into account when evaluating Roger's data.

The value of 0.18 uV was originally intended by Roger at the value of 50 Ohm input impedance even though the receiver's Hi Z input impedance is 1 KOhm. Therefore:

P = ((0.18*10^-6)^2)/50 = 6.48*10^-16W = -121.8 dBm

Is this technically correct? Yes: if you use a balun transformer with a ratio: 1000/50 = 20, the antenna will provide a voltage of 0.18 uV, and hence a power of -121.8 dBm, as intended by Roger, while the HI Z input will see a voltage of 0.18 *20 = 3.6 uV.

Please understand that the lower sensitivity of the RSP1-A, as compared to the RSP2/RSPduo being discussed in this thread, has perhaps some bearing only if you intend to use the RSP mainly in the LF range, otherwise the RSP-1A provides excellent service, on par with the other units of its class, up to 2 GHz. I hope to have removed confusion and solved your doubts.

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by Roger » Fri Jan 04, 2019 6:11 pm

sdrom33 wrote:Guys I want to buy a Rsp and now I am confused: must I get the 1a or the duo? For glovisol the duo is more sensitive by 35 db, very good. For Roger it is 123-101 = 22 db, good in one post and now is 14 db, not so hot in the last post. But the 1a is far cheaper. Can someone explain and help?
The RSP1A is a good performing SDR receiver but it does not have the sensitivity of the RSPduo and RSP2 when using the HiZ port. The designers of the SDRplay optimized the HiZ port for performance under 30 Mhz. The HiZ port is better under 2 MHz for MW, LF and VLF reception. Sensitivity is somewhat better between 2 MHz and 10 MHz but you will not see the difference because atmospheric and man-made noise is so high optimum sensitivity is not a concern. After 10 MHz the sensitivity of all the devices is very close when operating under 30 MHz. The table below is from SDRplay and gives you the sensitivity specs and only account for thermal noise. You can see that the RSPduo HiZ port is 23 dB more sensitive below 500 kHz under these white noise conditions. When you connect an antenna (like I showed earlier) the increase in external noise will result in the received noise floor getting higher (as shown) and the 23 dB difference will get smaller. In my antenna example the difference was only 14.8 dB due to the external noise. If the external noise was really high (severe man-made noise nearby or summer atmospheric noise) this could drop to 0 and then both receivers would then require the same signal level input to get the required SNR. My tests were done with an antenna not suited for LF DXing. When using a proper LF antenna atmospheric and man-made noise levels will be higher and the sensitivity difference between the different RSP products is not a factor.

If you just want to do all around listening and don't need the 3 switchable antenna ports then the RSP1A is your best bet. If you are a serious MW or VLF/LF DXer then you want the RSPduo because it has more filters for the HiZ port. THe RSP2 is a nice receiver as well and you can always buy an external MW filter if you are interested in LF reception. I will be posting another detailed discussion on dBm, Noise Figures and SNR and how they relate to the RSP in a few days.
.
Attachments
Compare RSP1A and Duo.png
Compare RSP1A and Duo.png (58.96 KiB) Viewed 26251 times

Reason: No reason

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

Re: RSP1A Longwave sensitivity

Post by sdrom33 » Sat Jan 05, 2019 6:56 am

Hi glovisol thanks for the explanation.

Hi Roger, I really appreciate your help & work to give info! I wait for next post before choose it.

Thank again guys

Sdrom33
Last edited by sdrom33 on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

arcosine
Posts: 30
Joined: Wed Jan 02, 2019 6:24 pm
Location: 41.678472, -88.051762

Re: RSP1A Longwave sensitivity

Post by arcosine » Sat Jan 05, 2019 2:37 pm

Hi,

Is the low Z input 50 ohms at LW or is it lower? I have the best signal with one turn secondary and 8 turns primary on a FT-87A-F toroid. That would make my antenna 50*64 = 3200 ohms, if the RSP1A input were 50 ohms. I also have a third winding with 18 turns to a 500 PF variable capacitor for tuning. I don't know how to determine my antenna impedance, but I could model it on 4NEC2.


... Ok I ran the antenna model in 4NEC2 and found it has an impedance strongly influenced by ground conductivity. At 350 khz, For a perfect ground, impedance is in the E-4 range. For a real ground its 1k to 5k ohm from good to poor soil, so 3k ohms sounds reasonable. NEC showed the pattern to be omni, but I know its directional. I also ran a flat top, 60 ft with 2 20 ft hat wires, was about 57 ohm with a good ground soil, and 1 ohm for perfect ground.

The DC resistance of my antenna is about the same as the wire, 700 ft of .030 aluminum welding wire, grounded and the edge of the lake.

Tony
Last edited by arcosine on Sat Jan 05, 2019 4:33 pm, edited 1 time in total.
Tony
AA9CC

Post Reply