RTL SDR scanner scan plot

[ashok shankar das]

Dear All,
I have scanned my location with RTLSDR_scan.py
The exported image is attached.

My question is "Is it possible to Identify the sgnal from the scan ?" I maean type of modulation.

Alternatively can some one point me to appropriate software which can scan a range and Identify the frequency and modulation?

thanks,
Ashok.

[Adam Nielsen]

> My question is "Is it possible to Identify the sgnal from the scan ?"
> I maean type of modulation.

I asked this same question (the idea being a way of automatically
identifying the signal type) and I was told if I could do this I should
patent it because I would become a millionaire ;-)

The only way I have been able to do this so far is by listening to
examples of different modulation types and finding one that sounded
similar to the signal I was interested in. There are a number of web
sites that provide samples, here are two:

http://www.rtl-sdr.com/signal-identification-guide/

http://www.kb9ukd.com/digital/

Cheers,
Adam.

[Ashok Shankar Das]

Hi Adam,

Thanks for your reply.

On Tue, Oct 8, 2013 at 1:23 PM, Adam Nielsen <a.ni...@shikadi.net> wrote:

> My question is "Is it possible to Identify the sgnal from the scan ?"
> I maean type of modulation.

I asked this same question (the idea being a way of automatically
identifying the signal type) and I was told if I could do this I should
patent it because I would become a millionaire ;-)

Oh great... then you are the first one and i am second one ;-)

 

The only way I have been able to do this so far is by listening to
examples of different modulation types and finding one that sounded
similar to the signal I was interested in.  There are a number of web
sites that provide samples, here are two:

  http://www.rtl-sdr.com/signal-identification-guide/

  http://www.kb9ukd.com/digital/

The second site i was not aware of. But it has some more info.

But do you know any site with pictorial info like the first link?

Regards.

Ashok.
 

Cheers,
Adam.

--
You received this message because you are subscribed to a topic in the Google Groups "Ultra Cheap SDR" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ultra-cheap-sdr/y-ZiJTT3f98/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ultra-cheap-s...@googlegroups.com.
To post to this group, send email to ultra-c...@googlegroups.com.
Visit this group at http://groups.google.com/group/ultra-cheap-sdr.
For more options, visit https://groups.google.com/groups/opt_out.

--
Thanks
Ashok.
-------------------------------------------------------
| See My Small Object Oriented OS Kernel       |
| Nanoos at my googlecode page bellow            |
| URL: http://code.google.com/u/ashok.s.das/  |
-------------------------------------------------------

[Adam Nielsen]

> The second site i was not aware of. But it has some more info.
> But do you know any site with pictorial info like the first link?

These are the only two I have used, but if you Google for terms like
"sdr waterfall examples" there are a few other sites. Perhaps someone
else here might post any other good sites they have found.

Cheers,
Adam.

[Leif Asbrink]

Hello Ashok,

From the scan I see center frequency = 94 MHz.
Bandwidth = many MHz.

This is not a signal. It could be computer noise but
more probably it is dongle overload. Your two strongest signals
are at the center of the feature.

Try about 10 dB lower gain. If the wide structure at 94 MHz
drops by more than 10 dB you have an overload problem.

The structure at 38 MHz looks like computer interference.

- Leif -

> --
> You received this message because you are subscribed to the Google Groups "Ultra Cheap SDR" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ultra-cheap-s...@googlegroups.com.

[Ashok Shankar Das]

Hi Leif,

Thanks for the Expert Comments. Definitely I will try and let all Know my Findings.

My Location is very near to GSM Towers and Also crowded by Cable TV network.

So There is a elevated noise Floor( Though I am Not an RF engineer).

Well Can You tell me Why You recommend Going 10dB Lower?

It will help me to enhance my knowledge.

Regards

Ashok.

You received this message because you are subscribed to a topic in the Google Groups "Ultra Cheap SDR" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ultra-cheap-sdr/y-ZiJTT3f98/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ultra-cheap-s...@googlegroups.com.

To post to this group, send email to ultra-c...@googlegroups.com.
Visit this group at http://groups.google.com/group/ultra-cheap-sdr.
For more options, visit https://groups.google.com/groups/opt_out.

[a...@eartoearoak.com]

Hi Ashok,

Glad you are finding my program useful.

As Leif suggests it might be worth dropping the gain as you investigate potential signals, if the gain is too high you risk overloading stages of the tuner.  If this happens the harmonics of the original signal maybe generated.
Harmonics are frequencies which are the integer product or divisor of the original frequency (ie 1/3, 1/2, 2 times, 3 times and etc).  An audio example is an electric guitar, at the flick of a switch you can turn the clean sound of the instrument to a much more rock sound. The heavier sound is produced by overloading and amplifier and generating more harmonics, producing a harsher sound.
So if the tuner in your dongle is overloaded by a strong signal, for example a FM radio station at 100MHz, you may see spikes that are fractions of this (i.e. 50, 25, 12,5MHz and lower) and multiples (200, 300, 400MHz and higher).  The level of a particular harmonic drops with each step away from the original frequency.

As the dongles are wideband the tuners have no input filtering, so if your scan is from 150 to 400MHz you will still see harmonics of that 100MHz signal.

I tend to start scans at a high gain and when I've found something interesting I then drop the gain to see if it's still visible.

To lower your noise floor it might be worth investing in a USB cable with a nice bit of thick shielding.  When I changed from a cheap cable to a branded one | got almost 10dB drop in noise.  I got another 5dB or so by passing the cable 3 times through a ferrite ring.  People have also used a USB hub powered from a 9v battery but I haven't tried this yet.

Wikipedia has some common oscillator frequencies, note USB at 24MHz (48MHz for USB is very common in equipment as well) and your dongle as an internal 28.8MHz crystal.
Your computer's clocks are square waves, without going into the maths a square wave is a sine wave (pure tone, one frequency) that's been massively over driven. This results in lots of harmonics, just odd numbers in this case: 1/7, 1/5, 1/3, 3, 5, 7 and so on, all of which can produce false signals in your spectrum.

Hope that helps,
Al

[Leif Asbrink]

Hello Ashok,

The two signals near 94 MHz are present simultaneously in the
dongle passband which is maybe 2 MHz. I do not know the characteristics
of the dongle you are using, but the two signals at about 93 and 94
MHz are at -4 dB on the scale. That means that the instantaneous amplitude
when the two signals add is twice the amplitude of one signal.
Consequently the level peaks outside the scale at +2dB, 6 dB above a
single signal. (Sometimes the two signals are out of phase and then
they cancel. On the average the two signals are 3 dB above a single
signal.)

One would expect overload to result in the presence of two false
signals at 92 and 95 MHz but such signals are not present.
There could be other signals outside the frequency range that do
unexpected things. It is a standard thing to try 10 dB less gain
to ensure that nothing of what you see is due to non-linearities
in the dongle.

There is also reciprocal mixing. The local oscillator of the dongle
has noise sidebands that are interacting with strong signals to
create noise sidebands around them. One might expect reciprocal
mixing around 100 dB below a signal whenm measured in a bandwidth
of 1 Hz. I do not know what bandwidth you were using, I can guess
it was 100 kHz which means you would see reciprocal mixing at
something like -50 dB. The broad signal you see is 20 dB stronger
than that so it is unlikely to be reciprocal mixing. The way to be sure
is to move the antenna around and look whether the broad signal
follows the signal level for the two signals. I do not think
you will find that they do. Maybe you will find that the broad signal
comes from your computer...

Regards

Leif

[Ashok Shankar Das]

Hi All,

I ama bit late as we are facing Cyclone Phailin.

Well as suggested I have taken a snapshot at 16dB gain frequency range 30 to 147 MHz.

I wish some body should analyse it.

Ashok.

[Leif Asbrink]

Hi All,

The rtl-sdr dongle has only 8 bit resolution. It is essential to run
it at a sampling speed providing Nyquist frequency well above the
-3dB point of the tuner filter.

Simple theory is not applicable however. It seems that RTL2832 does
not sample at 300 kHz when one asks that rate from the dongle.
It seems to sample faster and to add a digital anti-alias filter.
Not a good filter however as you can see in this video:
https://www.youtube.com/watch?v=bVrZ8r7q7cs

The video is intended to give an idea about how to try different
processing parameters to optimise performance of a RTL-SDR dongle
in Linrad on a slow computer.

73

Leif / SM5BSZ

[David J Taylor]

---------------------------------------------------

Thanks for that, Leif. The demonstration of the large interfering signals
was very useful. It might be interesting to repeat that with the popular
R820T tuner as the E4000 is now obsolete, and the R820T may have different
filtering. In my own experience, a really good filter in front of these
DVB-T dongles can do nothing but good, as they are so broad-band.

73,
David GM8ARV
--
SatSignal Software - Quality software written to your requirements
Web: http://www.satsignal.eu
Email: david-...@blueyonder.co.uk

[Leif Asbrink]

Hello David,

It is a pity that Elonics did not survive:-(

The R820T is quite a bit more sensitive to interference
outside the about 2 MHz range of the dongles.

Luckily it is fairly easy to make good filters with 1 to
2 MHz bandwidth to put in front of the dongles, that is
particularly useful for dongles with the R820T.

There is the Mirics MSi3101 (MSi2500 + MSi001.) It can provide
far more bandwidth and a better dynamic range.

I have made a comparison of several SDRs in wideband FM on
90.7 MHz: https://www.youtube.com/watch?v=ka_YORRjldY

The units are:
Elad FDM-S1
Mirics 3101
rtlsdr R820T
rtlsdr E4000 osmocom
rtlsdr E4000 sensitivity
FunCube Pro plus
PCIe-9842 with a down converter

The general conclusion is:
1) Try to buy a dongle with E4000 NOW! (While you still may have
a chance to get one!!
2) Dongles are cheap. Buy a Mirics and a R820T also.
3) Chances are good that one of them will work well in a
difficult environment. False responses are distributed
differently between the dongles.

73

Leif / SM5BSZ

> --
> You received this message because you are subscribed to the Google Groups "Ultra Cheap SDR" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ultra-cheap-s...@googlegroups.com.
> To post to this group, send email to ultra-c...@googlegroups.com.
> Visit this group at http://groups.google.com/group/ultra-cheap-sdr.

> For more options, visit https://groups.google.com/d/optout.