Topics

CW dBm measurement

Dennis Drakopoulos
 

Hello all!
I am using an AIRSPY R2 along with latest community SDR#. I also have a CW RF generator for up to 70MHz.
Having the RF generator output a CW carrier at 50MHz I tune to it in SDR# and write down the dBm shown. Now if I do not modify anything in hardware and software but only tune SDR to a new carrier at frequency of VHF or even UHF bands, would the dBm measurement returned be valid to use? Keeping the same reference obtained at 50MHz by the generator.
Idea behind is to have AIRSPY - SDR# measure a carrier at VU bands, where I do not have any instrument to do so.

Thank you!
73 de SV1CDN, Dennis!

Steve Sykes
 

I have asked this exact question before when I wanted to see the absolute noise floor. I am sure the SDR# coder will reply but SDR# reads dBFS.  Full scale is not an absolute value. I have been told that I must use SNR instead of dBm. To get an absolute value you will have to use another program.

73,
Steve KD2OM

On Tue, Feb 4, 2020 at 10:11 Dennis Drakopoulos via Groups.Io <sv1cdn=yahoo.com@groups.io> wrote:
Hello all!
I am using an AIRSPY R2 along with latest community SDR#. I also have a CW RF generator for up to 70MHz.
Having the RF generator output a CW carrier at 50MHz I tune to it in SDR# and write down the dBm shown. Now if I do not modify anything in hardware and software but only tune SDR to a new carrier at frequency of VHF or even UHF bands, would the dBm measurement returned be valid to use? Keeping the same reference obtained at 50MHz by the generator.
Idea behind is to have AIRSPY - SDR# measure a carrier at VU bands, where I do not have any instrument to do so.

Thank you!
73 de SV1CDN, Dennis!

jdow
 

SDRSharp does not measure in dBm. It measures dB relative to full scale which varies with a great number of factors. Without you making an effort, greater or lesser, to calibrate the system you cannot even measure signal power accurately. Then the noise aspect of the equation brings in more potential sources for error. I'd not use this system for any precision absolute level work. For relative levels of either noise or signal, not both, the accuracy should be very good.

{^_^}

On 20200204 06:14:02, Dennis Drakopoulos via Groups.Io wrote:
Hello all!
I am using an AIRSPY R2 along with latest community SDR#. I also have a CW RF generator for up to 70MHz.
Having the RF generator output a CW carrier at 50MHz I tune to it in SDR# and write down the dBm shown. Now if I do not modify anything in hardware and software but only tune SDR to a new carrier at frequency of VHF or even UHF bands, would the dBm measurement returned be valid to use? Keeping the same reference obtained at 50MHz by the generator.
Idea behind is to have AIRSPY - SDR# measure a carrier at VU bands, where I do not have any instrument to do so.
Thank you!
73 de SV1CDN, Dennis!

David Eckhardt
 

Included with the download of SDR#, toward the bottom of the list, resides "SpectrumSpy"  You can use that spectrum analyzer application to directly measure power in dBm (assumed 50-ohm non-reactive system).  It is relatively accurate in measuring the peak power of emissions.

Dave - WØLEV  


On Tue, Feb 4, 2020 at 6:31 PM Steve Sykes <sesykes71@...> wrote:
I have asked this exact question before when I wanted to see the absolute noise floor. I am sure the SDR# coder will reply but SDR# reads dBFS.  Full scale is not an absolute value. I have been told that I must use SNR instead of dBm. To get an absolute value you will have to use another program.

73,
Steve KD2OM

On Tue, Feb 4, 2020 at 10:11 Dennis Drakopoulos via Groups.Io <sv1cdn=yahoo.com@groups.io> wrote:
Hello all!
I am using an AIRSPY R2 along with latest community SDR#. I also have a CW RF generator for up to 70MHz.
Having the RF generator output a CW carrier at 50MHz I tune to it in SDR# and write down the dBm shown. Now if I do not modify anything in hardware and software but only tune SDR to a new carrier at frequency of VHF or even UHF bands, would the dBm measurement returned be valid to use? Keeping the same reference obtained at 50MHz by the generator.
Idea behind is to have AIRSPY - SDR# measure a carrier at VU bands, where I do not have any instrument to do so.

Thank you!
73 de SV1CDN, Dennis!



--
Dave - WØLEV
Just Let Darwin Work
Just Think

Dionisis “Dennis” Drakopoulos
 

Thank you all! 
So Dave, suppose Spectrum Spy shows me that at 50MHz a -20dBm signal is measured, mapped to a say input of 0dBm signal from the RF generator. And then at 450MHz a same signal of -20dBm is measured, would it be reliable to assume that the input is again 0dBm? Give it a 3 dBm error margin or perhaps 5 if you wish. I am not trying to replace a professional instrument, just get a close enough measurement - for ham use. 

73 de SV1CDN, Dennis!

David Eckhardt
 

I do not understand your statement regarding mapping -20 dBm to 0 dBm from a generator.  If SpectrumSpy indicates -20 dBm, that is the power (in a 50-ohm non-reactive system).  Same for 0 dBm, no matter where the signal originates.  The AirSpy is relatively flat to 1.8 GHz, so whether the signal is -20 dBm at 50 MHz or -20 dBm at 1.2 GHz, the measurement is correct (within published error).   You need not match the level to that from a generator.  Of course, as you increase frequency with the same coax, the loss will go upward due to the coax and the measured level will decrease.

Dave - WØLEV


On Thu, Feb 6, 2020 at 10:14 PM Dionisis “Dennis” Drakopoulos <sv1cdn@...> wrote:
Thank you all! 
So Dave, suppose Spectrum Spy shows me that at 50MHz a -20dBm signal is measured, mapped to a say input of 0dBm signal from the RF generator. And then at 450MHz a same signal of -20dBm is measured, would it be reliable to assume that the input is again 0dBm? Give it a 3 dBm error margin or perhaps 5 if you wish. I am not trying to replace a professional instrument, just get a close enough measurement - for ham use. 

73 de SV1CDN, Dennis!



--
Dave - WØLEV
Just Let Darwin Work
Just Think

jdow
 

A tool can read anything it wants. Whether it is to be believed refers back to the last time it was fed through a formal calibration and certification process. SDRSharp, at least, reads in dBFS. I would expect the same of SpectrumSpy unless somebody provides it with a calibration chart.

{^_^}

On 20200206 16:48:59, David Eckhardt wrote:
I do not understand your statement regarding mapping -20 dBm to 0 dBm from a generator.  If SpectrumSpy indicates -20 dBm, that is the power (in a 50-ohm non-reactive system).  Same for 0 dBm, no matter where the signal originates. The AirSpy is relatively flat to 1.8 GHz, so whether the signal is -20 dBm at 50 MHz or -20 dBm at 1.2 GHz, the measurement is correct (within published error).   You need not match the level to that from a generator.  Of course, as you increase frequency with the same coax, the loss will go upward due to the coax and the measured level will decrease.
Dave - WØLEV
On Thu, Feb 6, 2020 at 10:14 PM Dionisis “Dennis” Drakopoulos <@sv1cdn <mailto:@sv1cdn>> wrote:
Thank you all!
So Dave, suppose Spectrum Spy shows me that at 50MHz a -20dBm signal is
measured, mapped to a say input of 0dBm signal from the RF generator. And
then at 450MHz a same signal of -20dBm is measured, would it be reliable to
assume that the input is again 0dBm? Give it a 3 dBm error margin or perhaps
5 if you wish. I am not trying to replace a professional instrument, just
get a close enough measurement - for ham use.
73 de SV1CDN, Dennis!
--
*Dave - WØLEV
*
*/Just Let Darwin Work/*
*/Just Think/*

Bob Dengler
 

At 2/4/2020 11:46 AM, you wrote:
SDRSharp does not measure in dBm. It measures dB relative to full scale which
varies with a great number of factors. Without you making an effort, greater or
lesser, to calibrate the system you cannot even measure signal power accurately.
Then the noise aspect of the equation brings in more potential sources for
error. I'd not use this system for any precision absolute level work. For
relative levels of either noise or signal, not both, the accuracy should be very
good.
FWIW SDRuno does have an absolute channel power indicator & I checked it using an RSP1a against the signal generator in my E8285A service monitor, which was professionally calibrated before I purchased it. SDRuno's readings were always within +/- 2 dB of the sig. gen's indicated output.

If it's possible to achieve this level of accuracy in another SDR product, it should be possible to (semi-) calibrate an Airspy against a known standard provided you have a linear relative channel power indicator. If you want to measure noise power (e.g. radiometry) then your indicator needs to read R.M.S. or average power, not peak. Not sure how SDR software calculates total channel power but when using diode detectors this can be a major issue - why I stuck to thermocouple-type power meters for many years.

Bob NO6B

David Eckhardt
 

I have not used it in years, but HDSDR allows for calibration of the "S-Meter".  I usually selected 6 dB per S-Unit and set S-9 from the Sig Gen to -73 dBm (50 uV in a 50-ohm non-reactive system).

Dave - WØLEV


On Fri, Feb 7, 2020 at 5:31 PM Bob Dengler <no6b@...> wrote:
At 2/4/2020 11:46 AM, you wrote:
>SDRSharp does not measure in dBm. It measures dB relative to full scale which
>varies with a great number of factors. Without you making an effort, greater or
>lesser, to calibrate the system you cannot even measure signal power accurately.
>Then the noise aspect of the equation brings in more potential sources for
>error. I'd not use this system for any precision absolute level work. For
>relative levels of either noise or signal, not both, the accuracy should be very
>good.

FWIW SDRuno does have an absolute channel power indicator & I checked it using an RSP1a against the signal generator in my E8285A service monitor, which was professionally calibrated before I purchased it.  SDRuno's readings were always within +/- 2 dB of the sig. gen's indicated output.

If it's possible to achieve this level of accuracy in another SDR product, it should be possible to (semi-) calibrate an Airspy against a known standard provided you have a linear relative channel power indicator.  If you want to measure noise power (e.g. radiometry) then your indicator needs to read R.M.S. or average power, not peak.  Not sure how SDR software calculates total channel power but when using diode detectors this can be a major issue - why I stuck to thermocouple-type power meters for many years.

Bob NO6B






--
Dave - WØLEV
Just Let Darwin Work
Just Think

jdow
 

On 20200207 09:25:24, Bob Dengler wrote:
At 2/4/2020 11:46 AM, you wrote:
SDRSharp does not measure in dBm. It measures dB relative to full scale which
varies with a great number of factors. Without you making an effort, greater or
lesser, to calibrate the system you cannot even measure signal power accurately.
Then the noise aspect of the equation brings in more potential sources for
error. I'd not use this system for any precision absolute level work. For
relative levels of either noise or signal, not both, the accuracy should be very
good.
FWIW SDRuno does have an absolute channel power indicator & I checked it using an RSP1a against the signal generator in my E8285A service monitor, which was professionally calibrated before I purchased it. SDRuno's readings were always within +/- 2 dB of the sig. gen's indicated output.
If it's possible to achieve this level of accuracy in another SDR product, it should be possible to (semi-) calibrate an Airspy against a known standard provided you have a linear relative channel power indicator. If you want to measure noise power (e.g. radiometry) then your indicator needs to read R.M.S. or average power, not peak. Not sure how SDR software calculates total channel power but when using diode detectors this can be a major issue - why I stuck to thermocouple-type power meters for many years.
Bob NO6B
I never said you could not calibrate it. I said that the readings on SDRSharp's spectrum display are in dBFS. It does not measure and display dBm until you figure out a cross-calibration for it and the various gain settings you might use. There is no out of the box calibration.

{^_^}

jdow
 

You calibrated it. It did not come out of the box calibrated, did it?
{^_-}

On 20200207 11:14:04, David Eckhardt wrote:
I have not used it in years, but HDSDR allows for calibration of the "S-Meter". I usually selected 6 dB per S-Unit and set S-9 from the Sig Gen to -73 dBm (50 uV in a 50-ohm non-reactive system).
Dave - WØLEV
On Fri, Feb 7, 2020 at 5:31 PM Bob Dengler <no6b@... <mailto:no6b@...>> wrote:
At 2/4/2020 11:46 AM, you wrote:
>SDRSharp does not measure in dBm. It measures dB relative to full scale which
>varies with a great number of factors. Without you making an effort,
greater or
>lesser, to calibrate the system you cannot even measure signal power
accurately.
>Then the noise aspect of the equation brings in more potential sources for
>error. I'd not use this system for any precision absolute level work. For
>relative levels of either noise or signal, not both, the accuracy should
be very
>good.
FWIW SDRuno does have an absolute channel power indicator & I checked it
using an RSP1a against the signal generator in my E8285A service monitor,
which was professionally calibrated before I purchased it.  SDRuno's
readings were always within +/- 2 dB of the sig. gen's indicated output.
If it's possible to achieve this level of accuracy in another SDR product,
it should be possible to (semi-) calibrate an Airspy against a known
standard provided you have a linear relative channel power indicator.  If
you want to measure noise power (e.g. radiometry) then your indicator needs
to read R.M.S. or average power, not peak.  Not sure how SDR software
calculates total channel power but when using diode detectors this can be a
major issue - why I stuck to thermocouple-type power meters for many years.
Bob NO6B
--
*Dave - WØLEV
*
*/Just Let Darwin Work/*
*/Just Think/*

Bob Dengler
 

At 2/7/2020 02:41 PM, you wrote:
You calibrated it. It did not come out of the box calibrated, did it?
{^_-}
No I did not calibrate the SDRuno/RSP1a. I verified that it's out-of-the-box calibration was within +/- 2 dB.

Bob NO6B

Dennis Drakopoulos
 

Dave, suppose that RF generator registers a CW output of 0dBm into 50Ohm load. Disregard any connector and cable losses. Then AirSpy R2 is tuned to a specific QRG with certain demodulator - FFT - front end settings provisioned. Then SDR# measures say -10dBFS or Spectrum Spy say -15dBm. This way I map that true 0dBm are corresponding to -10dBFS / -15dBm.
If I tune SDR to another frequency, perhaps even 100 or more MHz away - but always below 500MHz, and SDR# returns -20dBFS then I will deduce that actual CW input power is - 10dBm.
Right?

David Eckhardt
 

If your sig gen is set for 0 dBm output and the coax between it and the SDR has insignificant loss, a reading of -10 dBFS should be good about anywhere the SDR covers.  I have tested the AirSpy R1 and R2 and found them to be relatively flat in response from some 20 MHz through 1.5 GHz.  That indicates if you 'calibrate' as described above and as you are doing, -10 dBFS is -10 dBFS and 0 dBm from 20 MHz through 1.5 GHz.

Dave - WØLEV


On Mon, Feb 10, 2020 at 8:22 PM Dennis Drakopoulos via Groups.Io <sv1cdn=yahoo.com@groups.io> wrote:
Dave, suppose that RF generator registers a CW output of 0dBm into 50Ohm load. Disregard any connector and cable losses. Then AirSpy R2 is tuned to a specific QRG with certain demodulator - FFT - front end settings provisioned. Then SDR# measures say -10dBFS or Spectrum Spy say -15dBm. This way I map that true 0dBm are corresponding to -10dBFS / -15dBm.
If I tune SDR to another frequency, perhaps even 100 or more MHz away - but always below 500MHz, and SDR# returns -20dBFS then I will deduce that actual CW input power is - 10dBm.
Right?



--
Dave - WØLEV
Just Let Darwin Work
Just Think

Dionisis “Dennis” Drakopoulos
 

TKS dr Dave fer ur insights sharing!
All the best!

73 de SV1CDN, Dennis!