Loading ...
Sorry, an error occurred while loading the content.
 

Re: [spectrumanalyzer] Re: Spurs

Expand Messages
  • william sprowls
    Hi Jurgen and All,  Spurs within the MSA (and any spectrum analyzer) are mathematically predictable. For the MSA, any time a combination of signals create
    Message 1 of 47 , Nov 25 10:57 AM
      Hi Jurgen and All,
       Spurs within the MSA (and any spectrum analyzer) are mathematically predictable. For the MSA, any time a combination of signals create 10.7 MHz (final I.F.) or 1013.3 MHz (1st I.F.) a potential spur will exist. Whether or not the spurs become problematic (visible on the graph) depends on the original design and the implementation of that design (filtering).  The number of predictable spurs are related to the number of signals and the number of local oscillators and mixers. This is why my original design of the Spectrum Analyzer uses a dual conversion, rather than more multiple conversions like the Big Boys use. They create more spurs, then spend a lot of money on extra filtering to get rid of those spurs. (dumb)

       The Verification MSA (in SA Mode) does not display any of your spurs except the one at 10.7 MHz (displayed as -112 dBm). They may be created in the MSA but they are below the noise floor at -120 dBm.

       Sam guessed that the spur at 12.038 MHz is a product of the two PLO's (8*LO1-8*LO2). He is right on.

       Here are the mathematics for known spurs.
       (m*LO1 - n*LO2 = 10.7 MHz) or
      (n*LO2 - m*LO1 = 10.7 MHz), where m and n are any whole numbers.
       These formulas "assume" that the Master oscillator is "right on" at 64 MHz and the Final IF Filter is exactly 10.7 MHz. This dictates that LO1 will always be exactly 1013.3 MHz above the Commanded Frequency (where the spur shows up).

      Here are some calculated known potential spurs:
      Spur at 8.025 MHz: 4*LO2 - 4*LO1=10.7 MHz = 4*1024 - (4*1021.325) = 10.7
      Spur at 9.3625 MHz: 8*LO2 - 8*LO1=10.7 MHz = 8*1024 - (8*1022.6625) = 10.7
      Spur at 10.7 MHz : DDS 1 spurious (see below)
      Spur at 12.0375 MHz : 8*LO1 - 8*LO2
      Spur at 16.05 MHz : 2*LO1 - 2*LO2
      Spur at 21.4 MHz : 1*LO1 - 1*LO2 (this is the image frequency that the cavity filter is designed to eliminate)
       As you can see, these spurs are the ones you have, but are a tiny bit different in frequency. This is due to the fact that your Final Filter center is not exactly 10.700 MHz.

        A little more on the DDS 1 spur at 10.7. If you expand your scale at this spur, you will probably find that the actual center frequency of the spur is the same frequency where DDS 1 is commanded to. Verify this by opening the Variables window and note the "dds1output = xx". Another method to verify that this spur is actually the DDS 1 frequency, halt the sweep, open the Sweep Parameters Window and click the Spur Test button. Continue the sweep and notice the spur has moved. The Spur Test changes the PLO 1 phase detector frequency, which changes the DDS 1 frequency.
       The DDS 1 spur is created by the DDS 1 modulating the PLO 1 and creating sidebands. You can also see this spur by commanding the MSA Center Frequency to -10.7 MHz. Yes, this is allowed, although there is no such thing as a negative frequency. However, this commands the PLO 1 to 1002.6 MHz. DDS 1 (10.7) creates an upper sideband (although small) at (1002.6 MHz + 10.7 MHz) = 1013.3 MHz. This sideband zips right through the Cavity Filter and mixes with PLO 2 in mixer 2, creating a signal at 10.7 MHz. Bingo, the spur.

       Any time you see a spur in the MSA, expand the sweep to measure the 3 dB points of the spur. If the spur BW is less than the BW of your Final Resolution Filter, then you are dealing with a harmonic. For example, your Final Filter BW is 7 KHz. Your spur shows a BW to be 1.4 KHz. This is one fifth. So you are dealing with a 5th harmonic.
      Scotty

      --- On Thu, 11/24/11, Sam <swetterlin@...> wrote:

      From: Sam <swetterlin@...>
      Subject: [spectrumanalyzer] Re: Spurs
      To: spectrumanalyzer@yahoogroups.com
      Date: Thursday, November 24, 2011, 5:05 PM

       

      You have a smaller number of spurs than I do, but a couple of big ones. Your 12.038 MHz (8*LO1-8*LO2, I think) spur probably results from LO1 leaking through Mixer 1 and the cavity filter to Mixer 2, and mixing with LO2. Maybe you have high leakage of LO1 through Mixer 1. If you have a way to check the Mixer 1 output for LO2 content, it would be worth doing.
      Sam W.

      --- In spectrumanalyzer@yahoogroups.com, Jurgen <khj221@...> wrote:
      >
      > My MSA is almost complete. Started testing.
      > Looking at spur's I found the following:
      >
      > 8.0264 MHz -73 dBm
      > 9.3576 MHz -100 dBm
      > 10.7 MHz -102 dBM
      > 12.038 MHz -68 dBm
      > 16.05 MHz -98 dBm
      > 21.4 MHz -102 dBm
      >
      > There may be more, but I was concentrating on peaks above -105 dBm.
      >
      > All modules are fenced and properly mounted.
      > PLO's, Mixer's, Comp amp, IF amp, filter, VGA, are sealed.
      >
      > Oscillator, phase detector, A/D converter fenced but still open.
      > All RF connections use UT141, solid copper tubing.
      > 10.7 MHz section of DDS sealed, rest open.
      >
      > I manually sealed the above modules with no effect.
      >
      > How does this result compare to the rest of you.
      >
      > Thanks, Jurgen
      >

    • James Hontoria
      Look in the msa msin web page (not the upgrade), I believe there is a reflection bridge in Sam Weterlin section Jim G. Hontoria, W1JGH Sent from my iPhone
      Message 47 of 47 , Oct 7, 2015
        Look in the msa msin web page (not the upgrade), I believe there is a reflection bridge in  Sam Weterlin section

        Jim G. Hontoria, W1JGH
        Sent from my iPhone

        On Oct 7, 2015, at 5:46 PM, ivancic.bojan@... [spectrumanalyzer] <spectrumanalyzer@yahoogroups.com> wrote:

         

        Hi,


        Please, I am kindly asking what additional hardware You are using for antenna measurements. As I understand this should be done in reflection mode, so a kind of a directional coupler must be used.
        Any idea to cover the all three bands up to 3GHz?

        best regards  Bojan

      Your message has been successfully submitted and would be delivered to recipients shortly.