Topics

Help with AirSpy

michaelkappeyne@...
 

Hi all,

'New to this forum, but not to AirSpy. I bought mine late 2015, never got it to work, and gave up. Now I've resurrected it.

My challenge is this: it works fine on WFM broadcast bands, but is dead just about everywhere else. I am particularly interested in Airband since I'm a student pilot. Using the same antenna that produces good results with my little basic Icom IC-R5 scanner, it shows no usable signal at all--pretty much a flat spectrum on SDR#. I have played with the gain. By way of sanity, should this AirSpy not be at least as good as a simple scanner? Am I overlooking something simple yet essential?

With thanks for your welcome, and many thanks beforehand,
Michael

Airspy US
 

Many users overlook increasing the gain above 0.

We recommend the following settings:

Free mode
IF gain - set to desired. A good baseline is to increase it until the 'noise floor' rises.
Mixer Gain - AUTO
LNA gain - AUTO

HQ has other recommendations but we have found the above to work well.

---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US

NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...

On 6/17/2019 3:30 PM, michaelkappeyne@... wrote:
Hi all,

'New to this forum, but not to AirSpy. I bought mine late 2015, never got it to work, and gave up. Now I've resurrected it.

My challenge is this: it works fine on WFM broadcast bands, but is dead just about everywhere else. I am particularly interested in Airband since I'm a student pilot. Using the same antenna that produces good results with my little basic Icom IC-R5 scanner, it shows no usable signal at all--pretty much a flat spectrum on SDR#. I have played with the gain. By way of sanity, should this AirSpy not be at least as good as a simple scanner? Am I overlooking something simple yet essential?

With thanks for your welcome, and many thanks beforehand,
Michael

<http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient> Virus-free. www.avg.com <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>

Airspy US
 

Also make sure you have selected the correct mode and bandwidth in the RADIO panel.

---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US

NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...

On 6/17/2019 3:36 PM, Airspy US via Groups.Io wrote:
Many users overlook increasing the gain above 0.

We recommend the following settings:

Free mode
IF gain - set to desired. A good baseline is to increase it until the 'noise floor' rises.
Mixer Gain - AUTO
LNA gain - AUTO

HQ has other recommendations but we have found the above to work well.

---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US

NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...

On 6/17/2019 3:30 PM, michaelkappeyne@... wrote:
Hi all,

'New to this forum, but not to AirSpy. I bought mine late 2015, never got it to work, and gave up. Now I've resurrected it.

My challenge is this: it works fine on WFM broadcast bands, but is dead just about everywhere else. I am particularly interested in Airband since I'm a student pilot. Using the same antenna that produces good results with my little basic Icom IC-R5 scanner, it shows no usable signal at all--pretty much a flat spectrum on SDR#. I have played with the gain. By way of sanity, should this AirSpy not be at least as good as a simple scanner? Am I overlooking something simple yet essential?

With thanks for your welcome, and many thanks beforehand,
Michael


<http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient> Virus-free. www.avg.com <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>



---
This email has been checked for viruses by AVG.
https://www.avg.com

michaelkappeyne@...
 

Hi Airspy,

Thank you very much for your quick reply!

Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?

My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.

The settings I use are:

IF gain: 4 - higher gain is not helping
Mixer gain: AUTO
LNA gain: AUTO

Sample rate 10 msps
Tracking filter: ON

AM at 25kHz bandwidth 

The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.

Cheers, and thanks again,
Michael

Airspy US
 

The bandwidth is a little wide. An AM signal will be no more than 10 kHz wide.
You may be confusing bandwidth with channel spacing.

Check the squelch setting, too.

Higher gain will result in hearing weaker signals.

The Airspy R2 sensitivity is about the same as the original, but both should be on par with the Icom.

---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US

NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...

On 6/18/2019 11:45 AM, michaelkappeyne@... wrote:
Hi Airspy,

Thank you very much for your quick reply!

Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?

My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.

The settings I use are:

IF gain: 4 - higher gain is not helping
Mixer gain: AUTO
LNA gain: AUTO

Sample rate 10 msps
Tracking filter: ON

AM at 25kHz bandwidth

The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.

Cheers, and thanks again,
Michael


------------------------------------------------------------------------
AVG logo <https://www.avg.com/internet-security>

This email has been checked for viruses by AVG antivirus software.
www.avg.com <https://www.avg.com/internet-security>

Greg Ella
 

I own an original AirSpy and a surveillance grade Icom R-8500 Communications Receiver.  I have a Radio Shack Discone scanner antenna on my roof.  I live about 4 miles from a regional airport that broadcasts airport weather (AWOS) on 135.075 MHz AM.

My Icom receives this AWOS station at S9 on my Discone antenna.  Inserting a Kay model 131B switchable attenuator and switching in all of the attenuation (41 dB) the Icom shows S0  and the AWOS is still audible but with heavy white noise.

Moving the coax from my Icom to my Airspy, and leaving the 41 dB of attenuation switched in, I have no trouble seeing and hearing the AWOS.  I'm using GQRX software in Linux, but the software and OS should not matter.  I have the LNA gain all the way up at 15 dB, and the mixer and IF gains about 3/4 up at 10 dB,
Normal AM mode with 10 KHz demod bandwidth (high cut 5000 Hz, low cut 5000 Hz).  On the spectrum display I'm seeing the noise floor at -100 and the AWOS at -80.

There is nothing wrong with the sensitivity of the original AirSpy on the VHF Air Band.  I think you have one of the following issues:

1.  Bad AirSpy
2.  Bad antenna or feed line
3.  Poor settings in your software
4.  Squelch is set too high.  Turn it all the way down (open) so you can hear white noise when no signal is present.  When you get everything else sorted out you can turn it up to where you want it.

If you are near Northern Colorado I can test your AirSpy for you.  Otherwise maybe someone near you has a second unit you can compare yours with.

Greg Ella
N0EMP


On Tue, Jun 18, 2019 at 10:16 AM Airspy US via Groups.Io <airspy_us=yahoo.com@groups.io> wrote:
The bandwidth is a little wide. An AM signal will be no more than 10 kHz
wide.
You may be confusing bandwidth with channel spacing.

Check the squelch setting, too.

Higher gain will result in hearing weaker signals.

The Airspy R2 sensitivity is about the same as the original, but both
should be on par with the Icom.

---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US

NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...

On 6/18/2019 11:45 AM, michaelkappeyne@... wrote:
> Hi Airspy,
>
> Thank you very much for your quick reply!
>
> Perhaps I'm misjudging the capabilities of SDR versus traditional
> radio? SDR is of course great in areas where traditional cannot touch,
> and perhaps verse versa. Does the Airspy R2 take its SDR performance
> to a new level compared to the original Airspy I have?
>
> My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM
> voice air band 118 - 136 mHz. This number is established at 10 dB S/N
> for a 1kHz tone at 30% modulation. Sticking with a single channel in
> that range, 120.150 mHz, my Icom is happily chatting away while SDR#
> only registers on occasions--although I can make out the voice on
> stronger signals. The antennas and their placement are nominally the same.
>
> The settings I use are:
>
> IF gain: 4 - higher gain is not helping
> Mixer gain: AUTO
> LNA gain: AUTO
>
> Sample rate 10 msps
> Tracking filter: ON
>
> AM at 25kHz bandwidth
>
> The Airspy is a marvelous piece of equipment--so well made too--and I
> really want to make the most of it and its underlying concepts.
>
> Cheers, and thanks again,
> Michael
>
>
>
> ------------------------------------------------------------------------
> AVG logo <https://www.avg.com/internet-security>     
>
> This email has been checked for viruses by AVG antivirus software.
> www.avg.com <https://www.avg.com/internet-security>
>
>




michaelkappeyne@...
 

Thank you for your help!

Reducing the bandwidth did help a bit. US Airband channel spacing is still at 50kHz, hence my prior choice of using 25 kHz for bandwidth. Reducing it to 10 kHz does indeed improve performance. ' Makes sense too.

I also noticed that SDR appears sensitive to using narrow-band antennas when using it for a narrow band. (I switched to a Slim Jim antenna centered on Airband.) This is of course true for scanners as well, but might SDR be more sensitive to antenna tuning since spectral decomposition occurs in the computational domain rather than in the resonant loop a receiver sets up for each frequency? This is a conceptual question.

Having tweaked gain and bandwidth I now achieve results comparable to my scanner, just as you anticipated. Voice is actually a bit clearer with higher fidelity than what my scanner produces.

One more question: should I set the filter order as high as possible until my computer gets overloaded? Or set it to a nominal value and leave it alone? If so, what value should I use?

Thanks again for your kind assistance!

Michael

Leif Asbrink
 

Hi Michael,

the NF of Airspy R2 and Airspy mini is 3.5 dB according
to specifications. I think the original Airspy 1 is the same.
It is perhaps more sensitive to USB noise so you might
need to be a bit more careful with the setup.

The below is from http://sm5bsz.com/linuxdsp/videos.htm
Modify Airspy part 1. : [Airspy-screening] Here a modification for Airspy One is shown. It is necessary to also do the modification shown in part 2 below, otherwise this modification degrades the noise figure significantly.
Modify Airspy part 2. : [AirspySecondMod] This video shows the second modification that is needed for the first one to not degrade the noise floor.
-------------------------------------------------------------------
https://www.youtube.com/watch?v=fXGMEdeGkbg

https://www.youtube.com/watch?v=cAuNwE69jes


The 0.56 micro-volt specification
can be converted to NF, but I do not know the formula and
I do not have the time to setup a system to measure
AM S/N and compare to the NF.

There is a fundamental aspect you are missing. The 0.56
micro-volt specification must be accompanied by the
actual noise bandwidth of the IF filter. For air band
I would guess a bandwidth of 6 kHz would be appropriate,
but with a 1 kHz tone one could set the bandwidth to
3 kHz and get a 3 dB better result.

With a bandwidth of 25 kHz in SDR# you get very poor results
letting 4 times more noise reaching the AM detector than with
a "normal" AM bandwidth. 25 kHz is for HIFI on mediumwave
with strong local broadcasters modulating with up to 12 kHz
or so.

Reduce the AM bandwidth for optimum reception. With SDR#
you can see what bandwidth the AM signals actually occupy.
Never allow more than that, but sometimes there is an
advantage in making the bandwidth a little narrower.


/ Leif

Hi Airspy,

Thank you very much for your quick reply!

Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?

My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.

The settings I use are:

IF gain: 4 - higher gain is not helping
Mixer gain: AUTO
LNA gain: AUTO

Sample rate 10 msps
Tracking filter: ON

AM at 25kHz bandwidth

The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.

Cheers, and thanks again,
Michael


David Eckhardt
 

Just so all things are equal, try switching the scanner antenna feedline and the AirSpy feedline.  There could be problems with the antenna and/or feedline connected to the AirSpy.  I'd also turn OFF the tracking filter just to get it working.  Also take the Mixer and RF gain off AUTO and set each to around 12.  4 seems a bit low for the IF gain as well.  Try a setting around 10 to 12. 

Dave - WØLEV


On Tue, Jun 18, 2019 at 3:46 PM <michaelkappeyne@...> wrote:
Hi Airspy,

Thank you very much for your quick reply!

Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?

My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.

The settings I use are:

IF gain: 4 - higher gain is not helping
Mixer gain: AUTO
LNA gain: AUTO

Sample rate 10 msps
Tracking filter: ON

AM at 25kHz bandwidth 

The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.

Cheers, and thanks again,
Michael



--
Dave - WØLEV
Just Let Darwin Work
Just Think

Airspy US
 

Actually it is 25 kHz channel spacing in the USA (and in use).

---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US

NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...

On 6/18/2019 2:48 PM, michaelkappeyne@... wrote:

Reducing the bandwidth did help a bit. US Airband channel spacing is still at 50kHz, hence my prior choice of using 25 kHz for bandwidth. Reducing it to 10 kHz does indeed improve performance. ' Makes sense too.

Chris Smolinski
 

Coincidentally, I use two AWOS stations as well as my "weak signal baseline". One is 8 miles away and is fairly weak, the other is 27 miles away and is extremely weak, just above the noise floor. My antenna is also a discone, hoisted up to the top of a tree (which gets it much higher than it would be on my roof, also further away from noise sources inside the house). Comparing between my Icom R-7000, AirSpy and AirSpy HF+ I get about the same signal on each receiver for both stations. As several others noted, using the proper bandwidth for the signal in question is critical for obtaining the best reception.

Chris Smolinski
Black Cat Systems
Westminster, MD USA
http://www.blackcatsystems.com

On Jun 18, 2019, at 1:46 PM, Greg Ella <gregella53@...> wrote:

I own an original AirSpy and a surveillance grade Icom R-8500 Communications Receiver. I have a Radio Shack Discone scanner antenna on my roof. I live about 4 miles from a regional airport that broadcasts airport weather (AWOS) on 135.075 MHz AM.

My Icom receives this AWOS station at S9 on my Discone antenna. Inserting a Kay model 131B switchable attenuator and switching in all of the attenuation (41 dB) the Icom shows S0 and the AWOS is still audible but with heavy white noise.

Moving the coax from my Icom to my Airspy, and leaving the 41 dB of attenuation switched in, I have no trouble seeing and hearing the AWOS. I'm using GQRX software in Linux, but the software and OS should not matter. I have the LNA gain all the way up at 15 dB, and the mixer and IF gains about 3/4 up at 10 dB,
Normal AM mode with 10 KHz demod bandwidth (high cut 5000 Hz, low cut 5000 Hz). On the spectrum display I'm seeing the noise floor at -100 and the AWOS at -80.

There is nothing wrong with the sensitivity of the original AirSpy on the VHF Air Band. I think you have one of the following issues:

1. Bad AirSpy
2. Bad antenna or feed line
3. Poor settings in your software
4. Squelch is set too high. Turn it all the way down (open) so you can hear white noise when no signal is present. When you get everything else sorted out you can turn it up to where you want it.

If you are near Northern Colorado I can test your AirSpy for you. Otherwise maybe someone near you has a second unit you can compare yours with.

Greg Ella
N0EMP

Leif Asbrink
 

Hi Chris and all,

for detecting weak AM stations one can use synchronous
detect to improve S/N by about 3 dB for weak signals.

Rather than rectifying one would generate a carrier
with the correct phase and amplitude by a narrowband
PLL or some other method. Then mixing that carrier
with the IF signal would provide an I/Q pair. I would
contain the AM modulated signal plus 50% of the noise.
Q would contain 50% of the noise only.

Skip Q and listen to I :-)

In Linrad you have this option under the name of "coherent
detect" but you can find it in several other SDR packages also.

/Leif

Coincidentally, I use two AWOS stations as well as my "weak signal baseline". One is 8 miles away and is fairly weak, the other is 27 miles away and is extremely weak, just above the noise floor. My antenna is also a discone, hoisted up to the top of a tree (which gets it much higher than it would be on my roof, also further away from noise sources inside the house). Comparing between my Icom R-7000, AirSpy and AirSpy HF+ I get about the same signal on each receiver for both stations. As several others noted, using the proper bandwidth for the signal in question is critical for obtaining the best reception.

Chris Smolinski
Black Cat Systems
Westminster, MD USA
http://www.blackcatsystems.com




On Jun 18, 2019, at 1:46 PM, Greg Ella <gregella53@...> wrote:

I own an original AirSpy and a surveillance grade Icom R-8500 Communications Receiver. I have a Radio Shack Discone scanner antenna on my roof. I live about 4 miles from a regional airport that broadcasts airport weather (AWOS) on 135.075 MHz AM.

My Icom receives this AWOS station at S9 on my Discone antenna. Inserting a Kay model 131B switchable attenuator and switching in all of the attenuation (41 dB) the Icom shows S0 and the AWOS is still audible but with heavy white noise.

Moving the coax from my Icom to my Airspy, and leaving the 41 dB of attenuation switched in, I have no trouble seeing and hearing the AWOS. I'm using GQRX software in Linux, but the software and OS should not matter. I have the LNA gain all the way up at 15 dB, and the mixer and IF gains about 3/4 up at 10 dB,
Normal AM mode with 10 KHz demod bandwidth (high cut 5000 Hz, low cut 5000 Hz). On the spectrum display I'm seeing the noise floor at -100 and the AWOS at -80.

There is nothing wrong with the sensitivity of the original AirSpy on the VHF Air Band. I think you have one of the following issues:

1. Bad AirSpy
2. Bad antenna or feed line
3. Poor settings in your software
4. Squelch is set too high. Turn it all the way down (open) so you can hear white noise when no signal is present. When you get everything else sorted out you can turn it up to where you want it.

If you are near Northern Colorado I can test your AirSpy for you. Otherwise maybe someone near you has a second unit you can compare yours with.

Greg Ella
N0EMP


David Eckhardt
 

1)  Take 0.56 microvolts as the sensitivity number given by the supplier. 
2)  Assume 10 kHz bandwidth.
3)  Then the noise power (assuming a 50 ohm non-reactive system) is:
         Pn = kTB 
                 where Pn: Noise Power in Watts
                            T:  equivalent noise temperature (Kelvins or K)
                            B is receive bandwidth  in Hz (assumed 10 kHz)
4)  And:  Pn = V^2 / R
                  where V:  0.56 microvolts
                             R:  System impedance (50+/- j0 ohms)
5)  Pn= V^2 / R = [(0.56E-6)^2] / 50 = 6.27E-15 watts
6)  From (3):     T = Pn / kB = 6.27E-15 / [(1.38E-23) X (10E3)] = 45.4E3 degrees
7)  Noise Figure or NF = 10Log{[T / T(Reference)] + 1}
                where T = 45.4E3 degrees
                           T(Reference):  290 K
8)  So:  NF = 10Lof{[45.4E3 / 290]+1} = 22 dB

This is not a very good noise figure (NF) for the scanner at all!!!!!  However, its a bit typical of wide-open front ends of a scanner. 

Take the AirSpy with a published NF of 3.5 dB.  The equivalent sensitivity in noise temperature (T in the above) would be:
          T = T(Reference) X {[10^(NF(dB) / 10)] - 1}  =  359 K

For reference, a noise figure (NF) of 3 dB yields a noise temperature (T) of T(Reference, or 290 K.

Huge difference from the scanner number of 45.4E3 K!!!!!

These formulas can all be found on <RFCafe.com> and elsewhere online or any basic text in receiver and communications theory.

Hope I haven't made any mistakes.....

Dave - WØLEV



On Tue, Jun 18, 2019 at 7:01 PM Leif Asbrink <leif@...> wrote:
Hi Michael,

the NF of Airspy R2 and Airspy mini is 3.5 dB according
to specifications. I think the original Airspy 1 is the same.
It is perhaps more sensitive to USB noise so you might
need to be a bit more careful with the setup.

The below is from http://sm5bsz.com/linuxdsp/videos.htm
Modify Airspy part 1. : [Airspy-screening] Here a modification for Airspy One is shown. It is necessary to also do the modification shown in part 2 below, otherwise this modification degrades the noise figure significantly.
Modify Airspy part 2. : [AirspySecondMod] This video shows the second modification that is needed for the first one to not degrade the noise floor.
-------------------------------------------------------------------
https://www.youtube.com/watch?v=fXGMEdeGkbg

https://www.youtube.com/watch?v=cAuNwE69jes


The 0.56 micro-volt specification
can be converted to NF, but I do not know the formula and
I do not have the time to setup a system to measure
AM S/N and compare to the NF.

There is a fundamental aspect you are missing. The 0.56
micro-volt specification must be accompanied by the
actual noise bandwidth of the IF filter. For air band
I would guess a bandwidth of 6 kHz would be appropriate,
but with a 1 kHz tone one could set the bandwidth to
3 kHz and get a 3 dB better result.

With a bandwidth of 25 kHz in SDR# you get very poor results
letting 4 times more noise reaching the AM detector than with
a "normal" AM bandwidth. 25 kHz is for HIFI on mediumwave
with strong local broadcasters modulating with up to 12 kHz
or so.

Reduce the AM bandwidth for optimum reception. With SDR#
you can see what bandwidth the AM signals actually occupy.
Never allow more than that, but sometimes there is an
advantage in making the bandwidth a little narrower.


/ Leif





> Hi Airspy,
>
> Thank you very much for your quick reply!
>
> Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?
>
> My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.
>
> The settings I use are:
>
> IF gain: 4 - higher gain is not helping
> Mixer gain: AUTO
> LNA gain: AUTO
>
> Sample rate 10 msps
> Tracking filter: ON
>
> AM at 25kHz bandwidth
>
> The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.
>
> Cheers, and thanks again,
> Michael
>
>
>





--
Dave - WØLEV
Just Let Darwin Work
Just Think

David Eckhardt
 

I don't know why this was trimmed from the email.  Here it is again:

1)  Take 0.56 microvolts as the sensitivity number given by the supplier. 
2)  Assume 10 kHz bandwidth.
3)  Then the noise power (assuming a 50 ohm non-reactive system) is:
         Pn = kTB 
                 where Pn: Noise Power in Watts
                            T:  equivalent noise temperature (Kelvins or K)
                            B is receive bandwidth  in Hz (assumed 10 kHz)
4)  And:  Pn = V^2 / R
                  where V:  0.56 microvolts
                             R:  System impedance (50+/- j0 ohms)
5)  Pn= V^2 / R = [(0.56E-6)^2] / 50 = 6.27E-15 watts
6)  From (3):     T = Pn / kB = 6.27E-15 / [(1.38E-23) X (10E3)] = 45.4E3 degrees
7)  Noise Figure or NF = 10Log{[T / T(Reference)] + 1}
                where T = 45.4E3 degrees
                           T(Reference):  290 K
8)  So:  NF = 10Lof{[45.4E3 / 290]+1} = 22 dB

This is not a very good noise figure (NF) for the scanner at all!!!!!  However, its a bit typical of wide-open front ends of a scanner. 

Take the AirSpy with a published NF of 3.5 dB.  The equivalent sensitivity in noise temperature (T in the above) would be:
          T = T(Reference) X {[10^(NF(dB) / 10)] - 1}  =  359 K

For reference, a noise figure (NF) of 3 dB yields a noise temperature (T) of T(Reference, or 290 K.

Huge difference from the scanner number of 45.4E3 K!!!!!

These formulas can all be found on <RFCafe.com> and elsewhere online or any basic text in receiver and communications theory.

Hope I haven't made any mistakes.....

Dave - WØLEV



On Tue, Jun 18, 2019 at 7:52 PM David Eckhardt via Groups.Io <davearea51a=gmail.com@groups.io> wrote:
1)  Take 0.56 microvolts as the sensitivity number given by the supplier. 
2)  Assume 10 kHz bandwidth.
3)  Then the noise power (assuming a 50 ohm non-reactive system) is:
         Pn = kTB 
                 where Pn: Noise Power in Watts
                            T:  equivalent noise temperature (Kelvins or K)
                            B is receive bandwidth  in Hz (assumed 10 kHz)
4)  And:  Pn = V^2 / R
                  where V:  0.56 microvolts
                             R:  System impedance (50+/- j0 ohms)
5)  Pn= V^2 / R = [(0.56E-6)^2] / 50 = 6.27E-15 watts
6)  From (3):     T = Pn / kB = 6.27E-15 / [(1.38E-23) X (10E3)] = 45.4E3 degrees
7)  Noise Figure or NF = 10Log{[T / T(Reference)] + 1}
                where T = 45.4E3 degrees
                           T(Reference):  290 K
8)  So:  NF = 10Lof{[45.4E3 / 290]+1} = 22 dB

This is not a very good noise figure (NF) for the scanner at all!!!!!  However, its a bit typical of wide-open front ends of a scanner. 

Take the AirSpy with a published NF of 3.5 dB.  The equivalent sensitivity in noise temperature (T in the above) would be:
          T = T(Reference) X {[10^(NF(dB) / 10)] - 1}  =  359 K

For reference, a noise figure (NF) of 3 dB yields a noise temperature (T) of T(Reference, or 290 K.

Huge difference from the scanner number of 45.4E3 K!!!!!

These formulas can all be found on <RFCafe.com> and elsewhere online or any basic text in receiver and communications theory.

Hope I haven't made any mistakes.....

Dave - WØLEV



On Tue, Jun 18, 2019 at 7:01 PM Leif Asbrink <leif@...> wrote:
Hi Michael,

the NF of Airspy R2 and Airspy mini is 3.5 dB according
to specifications. I think the original Airspy 1 is the same.
It is perhaps more sensitive to USB noise so you might
need to be a bit more careful with the setup.

The below is from http://sm5bsz.com/linuxdsp/videos.htm
Modify Airspy part 1. : [Airspy-screening] Here a modification for Airspy One is shown. It is necessary to also do the modification shown in part 2 below, otherwise this modification degrades the noise figure significantly.
Modify Airspy part 2. : [AirspySecondMod] This video shows the second modification that is needed for the first one to not degrade the noise floor.
-------------------------------------------------------------------
https://www.youtube.com/watch?v=fXGMEdeGkbg

https://www.youtube.com/watch?v=cAuNwE69jes


The 0.56 micro-volt specification
can be converted to NF, but I do not know the formula and
I do not have the time to setup a system to measure
AM S/N and compare to the NF.

There is a fundamental aspect you are missing. The 0.56
micro-volt specification must be accompanied by the
actual noise bandwidth of the IF filter. For air band
I would guess a bandwidth of 6 kHz would be appropriate,
but with a 1 kHz tone one could set the bandwidth to
3 kHz and get a 3 dB better result.

With a bandwidth of 25 kHz in SDR# you get very poor results
letting 4 times more noise reaching the AM detector than with
a "normal" AM bandwidth. 25 kHz is for HIFI on mediumwave
with strong local broadcasters modulating with up to 12 kHz
or so.

Reduce the AM bandwidth for optimum reception. With SDR#
you can see what bandwidth the AM signals actually occupy.
Never allow more than that, but sometimes there is an
advantage in making the bandwidth a little narrower.


/ Leif





> Hi Airspy,
>
> Thank you very much for your quick reply!
>
> Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?
>
> My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.
>
> The settings I use are:
>
> IF gain: 4 - higher gain is not helping
> Mixer gain: AUTO
> LNA gain: AUTO
>
> Sample rate 10 msps
> Tracking filter: ON
>
> AM at 25kHz bandwidth
>
> The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.
>
> Cheers, and thanks again,
> Michael
>
>
>





--
Dave - WØLEV
Just Let Darwin Work
Just Think



--
Dave - WØLEV
Just Let Darwin Work
Just Think

Marcus D. Leech
 

On 06/18/2019 03:47 PM, David Eckhardt wrote:
1)  Take 0.56 microvolts as the sensitivity number given by the supplier. 
2)  Assume 10 kHz bandwidth.
3)  Then the noise power (assuming a 50 ohm non-reactive system) is:
         Pn = kTB 
                 where Pn: Noise Power in Watts
                            T:  equivalent noise temperature (Kelvins or K)
                            B is receive bandwidth  in Hz (assumed 10 kHz)
4)  And:  Pn = V^2 / R
                  where V:  0.56 microvolts
                             R:  System impedance (50+/- j0 ohms)
5)  Pn= V^2 / R = [(0.56E-6)^2] / 50 = 6.27E-15 watts
6)  From (3):     T = Pn / kB = 6.27E-15 / [(1.38E-23) X (10E3)] = 45.4E3 degrees
7)  Noise Figure or NF = 10Log{[T / T(Reference)] + 1}
                where T = 45.4E3 degrees
                           T(Reference):  290 K
8)  So:  NF = 10Lof{[45.4E3 / 290]+1} = 22 dB

This is not a very good noise figure (NF) for the scanner at all!!!!!  However, its a bit typical of wide-open front ends of a scanner. 

Take the AirSpy with a published NF of 3.5 dB.  The equivalent sensitivity in noise temperature (T in the above) would be:
          T = T(Reference) X {[10^(NF(dB) / 10)] - 1}  =  359 K

For reference, a noise figure (NF) of 3 dB yields a noise temperature (T) of T(Reference, or 290 K.

Huge difference from the scanner number of 45.4E3 K!!!!!

These formulas can all be found on <RFCafe.com> and elsewhere online or any basic text in receiver and communications theory.

Hope I haven't made any mistakes.....

Dave - WØLEV



On Tue, Jun 18, 2019 at 7:01 PM Leif Asbrink <leif@...> wrote:
Hi Michael,

the NF of Airspy R2 and Airspy mini is 3.5 dB according
to specifications. I think the original Airspy 1 is the same.
It is perhaps more sensitive to USB noise so you might
need to be a bit more careful with the setup.

The below is from http://sm5bsz.com/linuxdsp/videos.htm
Modify Airspy part 1. : [Airspy-screening] Here a modification for Airspy One is shown. It is necessary to also do the modification shown in part 2 below, otherwise this modification degrades the noise figure significantly.
Modify Airspy part 2. : [AirspySecondMod] This video shows the second modification that is needed for the first one to not degrade the noise floor.
-------------------------------------------------------------------
https://www.youtube.com/watch?v=fXGMEdeGkbg

https://www.youtube.com/watch?v=cAuNwE69jes


The 0.56 micro-volt specification
can be converted to NF, but I do not know the formula and
I do not have the time to setup a system to measure
AM S/N and compare to the NF.

There is a fundamental aspect you are missing. The 0.56
micro-volt specification must be accompanied by the
actual noise bandwidth of the IF filter. For air band
I would guess a bandwidth of 6 kHz would be appropriate,
but with a 1 kHz tone one could set the bandwidth to
3 kHz and get a 3 dB better result.

With a bandwidth of 25 kHz in SDR# you get very poor results
letting 4 times more noise reaching the AM detector than with
a "normal" AM bandwidth. 25 kHz is for HIFI on mediumwave
with strong local broadcasters modulating with up to 12 kHz
or so.

Reduce the AM bandwidth for optimum reception. With SDR#
you can see what bandwidth the AM signals actually occupy.
Never allow more than that, but sometimes there is an
advantage in making the bandwidth a little narrower.


/ Leif





> Hi Airspy,
>
> Thank you very much for your quick reply!
>
> Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?
>
> My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.
>
> The settings I use are:
>
> IF gain: 4 - higher gain is not helping
> Mixer gain: AUTO
> LNA gain: AUTO
>
> Sample rate 10 msps
> Tracking filter: ON
>
> AM at 25kHz bandwidth
>
> The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.
>
> Cheers, and thanks again,
> Michael
>
>
>





--
Dave - WØLEV
Just Let Darwin Work
Just Think
_._,_._,_
I like to use this table here:

http://wera.cen.uni-hamburg.de/DBM.shtml

To convert from voltages to dBm.

The quoted sensitivity in voltage is roughly -121dBm, or divided by 10kHz gives about -161dBm/Hz.  A matched resistor at room temperature
  is -174dBm/Hz, which gives an equivalent noise figure of something like 16dB for the scanner receiver.  That's not wonderful, but from a
  historical perspective, it isn't horrific, either.

To achieve good final SNR for AM, you really do need to well-match the pre-detector bandwidth to the actual signal bandwidth, or you end
  up admitting a lot of other crap, reducing SNR, and overall audio quality.

Also, an SDR front-end is likely to be fairly wide, so placing an air-band filter in front may further improve things, to reduce the possibility
  of first-stage overload in the SDR.





Greg Ella
 

Hi Michael,

Not sure what you mean by a "narrow band" antenna, but the narrowest of narrow bandwidth antennas you could possibly buy would still have more bandwidth than a 6 KHz wide AM signal.

When you are tuned to a specific frequency, say for example 118.1 MHz, and you are listening to a transmission that is say 6 to 10 KHz wide, your receiver only "cares" about what is happening in that bandwidth, regardless of whether you are using an SDR or a conventional receiver.  The one exception being, if you have a high power signal somewhere else within the passband of your antenna and any pre-filtering in the receiver, the strong signal can overload your receiver.

I think you initially said you were using 10 MSPS.  Unless you need to see that much spectrum all at once or you are recording that much spectrum to your hard drive, go down to 2.5 MSPS.  That will reduce the amount of spectrum the AirSpy is looking at, and reduce the load on your computer.

Don't confuse demodulator bandwidth with channel spacing.  When a signal is transmitted it is engineered to occupy a certain amount of bandwidth, such as 10 KHz.  If you look at a strong AM signal on your waterfall display you can see the width of the signal, which is both left and right of the center carrier frequency.  You want your demodulator bandwidth to be the same as the transmitted signals bandwidth.  if you make it too narrow you are losing audio response.  If you make it too wide you are diluting your signal with additional noise.

Channel spacing is an administrative decision that someone made as to how far apart each carrier frequency is from it's neighbors in a band.  It has to be wider than the signal bandwidth so that the signals don't overlap, but it has nothing to do with demodulating a specific signal.

If you want to optimize your gain levels for a specific band, set up your SDR on the antenna and feed line you will be using.  Then find a signal in the band you are interested in.  Preferably a constant signal like AWOS, or at least a busy channel like a major control tower.  Now look on your spectrum display.  You will see the noise floor between signals and you will see the carrier for the signal you.  Manually adjust all of the gain levels to raise the peak of the carrier as high as you can without significantly raising the noise floor.  Work back and forth between the different gain controls.  you are trying to get the greatest DIFFERENCE between the noise floor and the peak of the carrier.

I would use a normal filter for listening to AM VHF Airband.

Greg Ella
N0EMP


On Tue, Jun 18, 2019 at 1:52 PM David Eckhardt <davearea51a@...> wrote:
1)  Take 0.56 microvolts as the sensitivity number given by the supplier. 
2)  Assume 10 kHz bandwidth.
3)  Then the noise power (assuming a 50 ohm non-reactive system) is:
         Pn = kTB 
                 where Pn: Noise Power in Watts
                            T:  equivalent noise temperature (Kelvins or K)
                            B is receive bandwidth  in Hz (assumed 10 kHz)
4)  And:  Pn = V^2 / R
                  where V:  0.56 microvolts
                             R:  System impedance (50+/- j0 ohms)
5)  Pn= V^2 / R = [(0.56E-6)^2] / 50 = 6.27E-15 watts
6)  From (3):     T = Pn / kB = 6.27E-15 / [(1.38E-23) X (10E3)] = 45.4E3 degrees
7)  Noise Figure or NF = 10Log{[T / T(Reference)] + 1}
                where T = 45.4E3 degrees
                           T(Reference):  290 K
8)  So:  NF = 10Lof{[45.4E3 / 290]+1} = 22 dB

This is not a very good noise figure (NF) for the scanner at all!!!!!  However, its a bit typical of wide-open front ends of a scanner. 

Take the AirSpy with a published NF of 3.5 dB.  The equivalent sensitivity in noise temperature (T in the above) would be:
          T = T(Reference) X {[10^(NF(dB) / 10)] - 1}  =  359 K

For reference, a noise figure (NF) of 3 dB yields a noise temperature (T) of T(Reference, or 290 K.

Huge difference from the scanner number of 45.4E3 K!!!!!

These formulas can all be found on <RFCafe.com> and elsewhere online or any basic text in receiver and communications theory.

Hope I haven't made any mistakes.....

Dave - WØLEV



On Tue, Jun 18, 2019 at 7:01 PM Leif Asbrink <leif@...> wrote:
Hi Michael,

the NF of Airspy R2 and Airspy mini is 3.5 dB according
to specifications. I think the original Airspy 1 is the same.
It is perhaps more sensitive to USB noise so you might
need to be a bit more careful with the setup.

The below is from http://sm5bsz.com/linuxdsp/videos.htm
Modify Airspy part 1. : [Airspy-screening] Here a modification for Airspy One is shown. It is necessary to also do the modification shown in part 2 below, otherwise this modification degrades the noise figure significantly.
Modify Airspy part 2. : [AirspySecondMod] This video shows the second modification that is needed for the first one to not degrade the noise floor.
-------------------------------------------------------------------
https://www.youtube.com/watch?v=fXGMEdeGkbg

https://www.youtube.com/watch?v=cAuNwE69jes


The 0.56 micro-volt specification
can be converted to NF, but I do not know the formula and
I do not have the time to setup a system to measure
AM S/N and compare to the NF.

There is a fundamental aspect you are missing. The 0.56
micro-volt specification must be accompanied by the
actual noise bandwidth of the IF filter. For air band
I would guess a bandwidth of 6 kHz would be appropriate,
but with a 1 kHz tone one could set the bandwidth to
3 kHz and get a 3 dB better result.

With a bandwidth of 25 kHz in SDR# you get very poor results
letting 4 times more noise reaching the AM detector than with
a "normal" AM bandwidth. 25 kHz is for HIFI on mediumwave
with strong local broadcasters modulating with up to 12 kHz
or so.

Reduce the AM bandwidth for optimum reception. With SDR#
you can see what bandwidth the AM signals actually occupy.
Never allow more than that, but sometimes there is an
advantage in making the bandwidth a little narrower.


/ Leif





> Hi Airspy,
>
> Thank you very much for your quick reply!
>
> Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?
>
> My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.
>
> The settings I use are:
>
> IF gain: 4 - higher gain is not helping
> Mixer gain: AUTO
> LNA gain: AUTO
>
> Sample rate 10 msps
> Tracking filter: ON
>
> AM at 25kHz bandwidth
>
> The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.
>
> Cheers, and thanks again,
> Michael
>
>
>





--
Dave - WØLEV
Just Let Darwin Work
Just Think

jdow
 

Perhaps instead of trying to diddle around with controls best used by people who understand noise figure and dynamic range he would be better served by selecting sensitivity mode and setting the gain up around 9 to 15. Once that works find a VOR signal that is moderately weak and starting with a gain setting of about 7 step the gain up until the apparent SNR on the VOR signal stops improving. That should give him an ideal dynamic range. If he receives particularly poorly even with a gain setting of 15 he may have his front end compromised. However, he should switch antennas and see what difference he sees. In 60 years of this stuff I've had my share of numb radios due to shorted antenna feed lines.

{o.o} Joanne

On 20190618 09:16:08, Airspy US via Groups.Io wrote:
The bandwidth is a little wide. An AM signal will be no more than 10 kHz wide.
You may be confusing bandwidth with channel spacing.
Check the squelch setting, too.
Higher gain will result in hearing weaker signals.
The Airspy R2 sensitivity is about the same as the original, but both should be on par with the Icom.
---------
Airspy.US
Your USA source for quality SDR products!
www.Airspy.US
NOTE! This email address is not routinely monitored.
If you have an issue, please contact us at airspy@...
On 6/18/2019 11:45 AM, michaelkappeyne@... wrote:
Hi Airspy,

Thank you very much for your quick reply!

Perhaps I'm misjudging the capabilities of SDR versus traditional radio? SDR is of course great in areas where traditional cannot touch, and perhaps verse versa. Does the Airspy R2 take its SDR performance to a new level compared to the original Airspy I have?

My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz tone at 30% modulation. Sticking with a single channel in that range, 120.150 mHz, my Icom is happily chatting away while SDR# only registers on occasions--although I can make out the voice on stronger signals. The antennas and their placement are nominally the same.

The settings I use are:

IF gain: 4 - higher gain is not helping
Mixer gain: AUTO
LNA gain: AUTO

Sample rate 10 msps
Tracking filter: ON

AM at 25kHz bandwidth

The Airspy is a marvelous piece of equipment--so well made too--and I really want to make the most of it and its underlying concepts.

Cheers, and thanks again,
Michael



------------------------------------------------------------------------
AVG logo <https://www.avg.com/internet-security>

This email has been checked for viruses by AVG antivirus software.
www.avg.com <https://www.avg.com/internet-security>

jdow
 

Say what? US civilian aircraft VHF frequencies have had 25 kHz spacing for decades now.

If you face dynamic range problems then you want to set the gain as low as practical. There is little or nothing to be gained past the point at which increases in gain do not increase signal to noise ratio. So start low and step it up watching the distance between a signal peak and the noise fuzz. At first signal will pop up higher above noise. Eventually they start going up together. At that point back off one step on gain setting and you have probably the optimum setting. I am speaking in terms of the "Sensitivity" mode setting. (Yuossef understands noise figure and has documentation. So his settings are pretty close to as good as it gets.) If large signals are still a problem you can use linearity mode for a slight compromise in sensitivity and a useful, if not huge, increase in dynamic range.

{^_^}

On 20190618 11:48:00, michaelkappeyne@... wrote:
Thank you for your help!
Reducing the bandwidth did help a bit. US Airband channel spacing is still at 50kHz, hence my prior choice of using 25 kHz for bandwidth. Reducing it to 10 kHz does indeed improve performance. ' Makes sense too.
I also noticed that SDR appears sensitive to using narrow-band antennas when using it for a narrow band. (I switched to a Slim Jim antenna centered on Airband.) This is of course true for scanners as well, but might SDR be more sensitive to antenna tuning since spectral decomposition occurs in the computational domain rather than in the resonant loop a receiver sets up for each frequency? This is a conceptual question.
Having tweaked gain and bandwidth I now achieve results comparable to my scanner, just as you anticipated. Voice is actually a bit clearer with higher fidelity than what my scanner produces.
One more question: should I set the filter order as high as possible until my computer gets overloaded? Or set it to a nominal value and leave it alone? If so, what value should I use?
Thanks again for your kind assistance!
Michael

michaelkappeyne@...
 

Thank you all for your kind and detailed replies! Your willingness to help as well as level of expertise is awesome,

I'm up and running with AirSpy plus SDR# and getting results that are now better than my Icom R5, just as Dave's analysis anticipated. Continuing to adjust the various gains per his and AirSpy's advice should garner even better results with more experimentation. And yes, Airband spacing is 25 kHz, my mistake, and signals are captured in a 6 to 8 kHz bandwidth so I made that change.

A small challenge is the highly intermittent nature of Airband signals I am able to receive. No continuous weather AWOS for me. Living in a 24th floor apartment on Manhattan's West side I can receive most of Newark air traffic, plus Ground on a good day, but that's it. I could use LiveATC of course but that's no fun at all. Doing things the hard way is. So I hang my Slim Jim out the window and do the best I can!

'Very glad that I resurrected my AirSpy after my initial frustration. It is amazing and most intriguing technology.

Many thanks again to AirSpy, Chris, David, Greg, and Leif!

Cheers,
Michiel Kappeyne

jdow
 

It takes a really obnoxious front end design to produce a noise figure over 10 dB. You also lost 10 dB in there, it appears. He noted a 10 dB SNR for 0.56 uV modulated 30%. So that last may account for another 5 dB. That puts it into a range of noise figure I'd expect.

{^_-}

On 20190618 12:47:14, David Eckhardt wrote:
1)  Take 0.56 microvolts as the sensitivity number given by the supplier.
2)  Assume 10 kHz bandwidth.
3)  Then the noise power (assuming a 50 ohm non-reactive system) is:
         Pn = kTB
                 where Pn: Noise Power in Watts
                            T:  equivalent noise temperature (Kelvins or K)
                            B is receive bandwidth  in Hz (assumed 10 kHz)
4)  And:  Pn = V^2 / R
                  where V:  0.56 microvolts
                             R:  System impedance (50+/- j0 ohms)
5)  Pn= V^2 / R = [(0.56E-6)^2] / 50 = 6.27E-15 watts
6)  From (3):     T = Pn / kB = 6.27E-15 / [(1.38E-23) X (10E3)] = 45.4E3 degrees
7)  Noise Figure or NF = 10Log{[T / T(Reference)] + 1}
                where T = 45.4E3 degrees
                           T(Reference):  290 K
8)  So:  NF = 10Lof{[45.4E3 / 290]+1} = 22 dB
*This is not a very good noise figure (NF) for the scanner at all!!!!! *However, its a bit typical of wide-open front ends of a scanner.
**
Take the AirSpy with a published NF of 3.5 dB.  The equivalent sensitivity in noise temperature (T in the above) would be:
          T = T(Reference) X {[10^(NF(dB) / 10)] - 1}  =  359 K
For reference, a noise figure (NF) of 3 dB yields a noise temperature (T) of T(Reference, or 290 K.
Huge difference from the scanner number of 45.4E3 K!!!!!
These formulas can all be found on <RFCafe.com> and elsewhere online or any basic text in receiver and communications theory.
Hope I haven't made any mistakes.....
Dave - WØLEV
On Tue, Jun 18, 2019 at 7:01 PM Leif Asbrink <leif@... <mailto:leif@...>> wrote:
Hi Michael,
the NF of Airspy R2 and Airspy mini is 3.5 dB according
to specifications. I think the original Airspy 1 is the same.
It is perhaps more sensitive to USB noise so you might
need to be a bit more careful with the setup.
The below is from http://sm5bsz.com/linuxdsp/videos.htm
Modify Airspy part 1. : [Airspy-screening] Here a modification for Airspy
One is shown. It is necessary to also do the modification shown in part 2
below, otherwise this modification degrades the noise figure significantly.
Modify Airspy part 2. : [AirspySecondMod] This video shows the second
modification that is needed for the first one to not degrade the noise floor.
-------------------------------------------------------------------
https://www.youtube.com/watch?v=fXGMEdeGkbg
https://www.youtube.com/watch?v=cAuNwE69jes
The 0.56 micro-volt specification
can be converted to NF, but I do not know the formula and
I do not have the time to setup a system to measure
AM S/N and compare to the NF.
There is a fundamental aspect you are missing. The 0.56
micro-volt specification must be accompanied by the
actual noise bandwidth of the IF filter. For air band
I would guess a bandwidth of 6 kHz would be appropriate,
but with a 1 kHz tone one could set the bandwidth to
3 kHz and get a 3 dB better result.
With a bandwidth of 25 kHz in SDR# you get very poor results
letting 4 times more noise reaching the AM detector than with
a "normal" AM bandwidth. 25 kHz is for HIFI on mediumwave
with strong local broadcasters modulating with up to 12 kHz
or so.
Reduce the AM bandwidth for optimum reception. With SDR#
you can see what bandwidth the AM signals actually occupy.
Never allow more than that, but sometimes there is an
advantage in making the bandwidth a little narrower.
/ Leif

> Hi Airspy,
>
> Thank you very much for your quick reply!
>
> Perhaps I'm misjudging the capabilities of SDR versus traditional radio?
SDR is of course great in areas where traditional cannot touch, and perhaps
verse versa. Does the Airspy R2 take its SDR performance to a new level
compared to the original Airspy I have?
>
> My Icom scanner's sensitivity is given as 0.56 micro-volt on the AM voice
air band 118 - 136 mHz. This number is established at 10 dB S/N for a 1kHz
tone at 30% modulation. Sticking with a single channel in that range,
120.150 mHz, my Icom is happily chatting away while SDR# only registers on
occasions--although I can make out the voice on stronger signals. The
antennas and their placement are nominally the same.
>
> The settings I use are:
>
> IF gain: 4 - higher gain is not helping
> Mixer gain: AUTO
> LNA gain: AUTO
>
> Sample rate 10 msps
> Tracking filter: ON
>
> AM at 25kHz bandwidth
>
> The Airspy is a marvelous piece of equipment--so well made too--and I
really want to make the most of it and its underlying concepts.
>
> Cheers, and thanks again,
> Michael
>
>
>
--
*Dave - WØLEV
*
*/Just Let Darwin Work/*
*/Just Think/*