LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Unwanted Signals appear on the spectrum using RFSA 5665 (3.6GHz)

Solved!
Go to solution

Hello everyone,

 

I'm using NI PXIe 5665 (3.6GHz) RFSA which uses 5603 downconverter, 5653 synthesizer and 5622 IF digitizer. I used the NI-RFSA Soft Front Panel initially to acquire the RF spectrum with Reference level: 0dBm, Center Frequency: 100MHz, Span 50MHz and RBW: 3kHz. There was no input given to the downconverter RF In port. The received spectrum is shown below which is the expected spectrum because there was no signal present at the input of the downconverter.

 

RFSA Soft Panel.png

 

Then I opened up the example 'RFSA Getting Started Spectrum.vi' from the NI example finder. This example configures RFSA to acquire data in spectrum acquisition mode. With Reference Level: 0dBm, Start Freq: 20Hz, Stop Frequency: 3.6GHz, RBW: 3kHz and no input signal connected to the downconverter RF In port, the received spectrum shows peaks after every 50MHz bandwidth distance which is strange as these peaks don't appear when aquisition is done using NI-RFSA Soft Front Panel . This spectrum is shown in the image below.

 

Ref0_RFSA Spectrum Acquisition Mode.png

 

Zoomed Spectrum..

Ref0_Zoom3.png

 

I checked this with another example 'SMT IQ Spectrum for niRFSA.vi' from the NI Example Finder. This example configures RFSA to acquire IQ data and compute its spectrum using SMT VIs. I ran the example with the following settings; Reference Level: 0dBm, Center Frequency: 100MHz, IQ rate: 62.5MHz, RBW: 3kHz and spectrum plot from -fs/2 to fs/2. Again there was no input signal connected to the downconverter RF In port. The received spectrum again showed that peak which was appearing after every 50MHz span in the previous example plus it also showed two really high level signals near fs/2 and -fs/2. The figure below shows this

 

RFSA IQ Spectrum.png

 

Same result was produced by the RFSA Demo Panel.vi example which is shown below

 

RFSA Demo Panel.png

 

I'm not getting why are these signals appearing in the spectrum when there is no input signal connected to the downconverter? And why do they not appear on the spectrum displayed by RFSA Soft Front Panel ? Is there any configuration setting that is missing in all of these examples?

 

Thanks

0 Kudos
Message 1 of 7
(3,225 Views)

Hello,

 

Try to disable the preamplifier and set the IF Filter to through using the property nodes in RFSA.

 

 

Thanks

0 Kudos
Message 2 of 7
(3,184 Views)

What you are seeing is called "Local Oscillator Leakage."  The Soft Front panels use a technique called "Spectral Stitching" to mask the LO Bleedthrough.  Esentially they take a span a little above and another span a litte below Fc and "Stitch" them together affer cutting out the LO.  At least that's what I encountered a few years ago with the 5644R VSTs


"Should be" isn't "Is" -Jay
0 Kudos
Message 3 of 7
(3,172 Views)

Hi Adnan,

 

I have checked. The Pre-amplifier is already disabled and IF Filter is by default set to Through.

0 Kudos
Message 4 of 7
(3,151 Views)
Solution
Accepted by topic author SunJab

I think we have figured out the issue. The unwanted signals appearing on the spectrum are basically residual leakage of the dither signal. This residual leakage can be eliminated by disabling the digitizer dithering through the niscope property node. Refer to the following links

 

http://zone.ni.com/reference/en-XX/help/372058N-01/rfsapropref/pnirfsa_digitizerditherenabled/

 

http://dsp.stackexchange.com/questions/19782/improving-digitization-by-adding-white-noise-before-ove...

 

Regards

Message 5 of 7
(3,130 Views)

pl ignore this email

0 Kudos
Message 6 of 7
(2,235 Views)

@SunJab wrote:

pl ignore this email


Ignoring 


"Should be" isn't "Is" -Jay
0 Kudos
Message 7 of 7
(2,219 Views)