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

Re: [wsjtgroup] Re: K3 noise blanker with ISCAT/FSK

Expand Messages
  • Bill VanAlstyne W5WVO
    Hi Dan, Thanks for the data point. Theoretically, I wouldn’t expect any problems decoding ISCAT with DSP NB on, because it is a much slower protocol, and
    Message 1 of 11 , Jul 25, 2011
    • 0 Attachment
      Hi Dan,
       
      Thanks for the data point. Theoretically, I wouldn’t expect any problems decoding ISCAT with DSP NB on, because it is a much slower protocol, and thus much less likely to trigger the DSP algorithms. However, I use ISCAT only occasionally, since meteor scatter has been my main pursuit, and ISCAT is typically too slow for brief (< 200 ms) m/s pings.
       
      Your K3 setup is pretty much the way I have mine set up also for WSJT – AGC OFF, PR-6 inline, internal preamp off. Sometimes lightning crashes are too high in amplitude to run with AGC OFF, but in those cases, the overall noise profile is such that operating at all is unproductive, not to mention somewhat risky because of the proximity of lightning discharges, so I pull the switch and disconnect. :-)
       
      Bill W5WVO
       
       
      From: Dan Bates
      Sent: Monday, July 25, 2011 18:18
      Subject: RE: [wsjtgroup] Re: K3 noise blanker with ISCAT/FSK
       

      Just as a reference, I have a very nasty power line buzz that starts popping on hot afternoons until it becomes solid then goes away after sundown.

       

      I have the K3 NB set to DSP t3-6, and IF Med4.  This completely knocks out the noise, and I notice no problems with decoding ISCAT. The other thing I do is turn off AGC which keeps the noise pulses from shutting down the rf gain.  I also only run the external preamp and leave the internal preamp off.  I notice no decrease in SNR, but an improved noise floor..

       

      I’d appreciate anyone else’s experience.

       

      Dan n5tm

       

      From: wsjtgroup@yahoogroups.com [mailto:wsjtgroup@yahoogroups.com] On Behalf Of Bill VanAlstyne W5WVO
      Sent: Sunday, July 24, 2011 5:07 PM
      To: wsjtgroup@yahoogroups.com; TLC_Carney
      Subject: Re: [wsjtgroup] Re: K3 noise blanker with ISCAT/FSK

       

       

      Tom,

       

      Analog noise blankers tend to work best on “impulse”-type noise – the kind of noise you get from sparks (ignition systems, power line arcing, etc.). Electronically generated broadband noise is typically not reduced a lot, if any, by this kind of NB, which is why Elecraft put in a DSP-based NB as well. The noise I tried the analog NB on sounded like power-line noise to me.

       

      Bill W5WVO

       

       

      From: TLC_Carney

      Sent: Sunday, July 24, 2011 16:28

      Subject: [wsjtgroup] Re: K3 noise blanker with ISCAT/FSK

       

       

      Thanks for the follow up Bill. My experience is exactly the same as your with the DSP based noise blanker. I didn't try the IF blanker as, even cranked all the way up, it had very little effect on the noise in my neighborhood. I tried it again this morning and while it still has little effect on the worst noise source, it does reduce the noise of other sources in the neighborhood.

      I'll try some experimenting over the next week and post the results.

      73,

      Tom K6EU

      --- In mailto:wsjtgroup%40yahoogroups.com, "Bill VanAlstyne W5WVO" <w5wvo@...> wrote:

      >
      > I’m
      resurrecting this old thread as I have some new anecdotal data on it.
      >
      > As suggested by Joe Taylor (see thread below), I’ve been
      experimenting with the K3 noise blankers (DSP and conventional analog) to see which one performs the best with FSK441 on meteor scatter signals.
      >
      > Over and over, I’ve found that running the DSP noise blanker on the
      K3 at any useful level creates artifacts or behaviors that seem to interfere with FSK441 decoding of weak to moderate signals. Strong signals that are well above the ambient noise level are not affected by the DSP noise blanker, but of course the NB is, in such a case, unnecessary anyway. But any signal that would be no more than a dB or two above the noise without the NB on does not decode cleanly with the DSP NB set to any useful noise-blanking level.
      >
      >
      To my amazement, however, the analog NB seems to have no such problems. While it reduces the dynamic range of the receiver chain (as does any analog IF NB), this is typically not a problem when working weak-signal modes. I have cleanly decoded weak FSK441 sigs that would have been in the noise had the K3 analog NB not been invoked.
      >
      > So tentatively, if you use a K3, turn OFF the
      DSP NB and turn ON the analog NB. Set it to the least aggressive position that produces a maximum of noise reduction. K3 m/s ops, let me know your experiences with this.
      >
      > Again, to emphasize, this is with regard to FSK441
      VHF meteor-scatter operation with the Elecraft K3. Not anything else.
      >
      > Bill W5WVO
      >
      >
      >
      > From: Bill VanAlstyne
      W5WVO
      > Sent: Friday, February 11, 2011 03:50
      > To: 'TLC_Carney' ;
      mailto:wsjtgroup%40yahoogroups.com ; Barry Garratt
      > Subject: Re: [wsjtgroup] noise blanker with
      ISCAT/FSK
      >
      >
      >
      > Tom, Barry, et al.,
      >
      > This question about noise blankers and FSK441 has occurred to me as
      well. I wrote to Joe Taylor about it, in fact. This is what he said:
      >
      > “ In general I have not seen bad effects of a noise blanker on FSK441
      signals. I think it generally unlikely, but in detail it must depend on how the blanker is implemented and configured. I'd say you should experiment and decide for your own system.”
      >
      > Well, I’d already been doing that, of
      course.
      >
      > Barry and I both run K3s, and the K3 actually has two
      independent noise blankers -- one being a standard-design IF hardware noise blanker (albeit with numerous available threshold and time-slice settings), which like all such NBs introduces considerable non-linearity into the IF path; and a second DSP-implemented noise blanker which generally preserves receiver linearity and dynamic range, but still (as it must) “cuts out holes” in the received tones. So there is a lot of opportunity in the K3 for “experimenting.”
      >
      > The problem with “experimenting” with
      meteor scatter, of course, is that it’s rather difficult to try different setups on the same m/s signal because they don’t last long enough, and in any case are subject to propagation vagaries of their own during the few seconds (at best) that they are heard. The only way to really test NB implementations would be to make use of a soundcard-and-software generator that can be set to emulate meteor pings at the radio frequencies being used. The audio thus generated is fed into a transmitter, the RF output attenuated to very low levels, and finally fed into a WSJT-equipped receiver.
      >
      > Now, I don’t have such a
      “ping generator”, but I believe Joe has something like this that he uses during decoder development. One of these days I’m going to ask Joe if he would be willing to share that code and setup with the rest of the world. :-)
      >
      > But as far as unscientific, subjective experience is concerned â€" I
      believe that running a noise blanker at aggressive levels can have a distorting effect on weak FSK441 pings that can render them difficult to decode. (I haven’t sensed any problems using more conservative NB settings.) The catch, of course, is that in many cases of severe man-made QRN, without the aggressively-set NB, you wouldn’t see or hear or decode the ping in the first place, because it would be way below the pumped-up noise floor. :-) Remember that FSK441 does take up a sizeable portion of the SSB reference passband, due both to its spread of audio tones and to its high speed. Because of these factors, encoded signals even slightly below the noise level will not decode cleanly â€" nothing at all like JT65 or even ISCAT, which are much slower (and in the case of JT65A, much narrower). So to speak to the other implied question â€" no, I don’t think using a NB would interfere significantly or at all with decoding ISCAT. And it definitely has no effect on decoding JT65 that I’ve ever observed.
      >
      > Bill W5WVO
      >
      >
      > From: Barry
      Garratt
      > Sent: Friday, February 11, 2011 02:14
      > To: 'TLC_Carney'
      ; mailto:wsjtgroup%40yahoogroups.com
      > Subject: RE: [wsjtgroup] noise blanker with ISCAT/FSK
      >
      >
      >
      > Hey Tom,
      >
      >
      > I’m in a very quiet
      location but occasionally get some noise when I’m pointing west because of a mall out that way. I run my K3 with both preamps on all the time by the way. Unless I’m pointing west my S meter barely flickers the S1 led at times. Even though I have that kind of noise level I will usually run the NB and also I have a Timewave ANC-4. It’s the 1st thing in the chain and it feeds the PRE-6 GasFet. That pre-amp then feeds a DCI 10 pole, 300 KHz bandpass filter the output of which then goes back into the K3.
      >
      >
      > I run it
      like that all the time usually just so I don’t forget to switch things on if I turn west. When it comes to weak signals every little bit helps. I also limit the audio to the sound card with an audio DSP, also by Timewave, to 700 â€" 2700 Hz and run the K3 with the 1800 Hz filter.
      >
      >
      > Just my way
      of doing things Tom. As the saying goes your mileage may vary.
      >
      >
      > 73,
      >
      >
      > Barry KS7DX
      >
      >
      >
      From: mailto:wsjtgroup%40yahoogroups.com [mailto:mailto:wsjtgroup%40yahoogroups.com] On Behalf Of TLC_Carney
      > Sent: Thursday, February 10, 2011 5:40
      PM
      > To:
      href="mailto:wsjtgroup%40yahoogroups.com">mailto:wsjtgroup%40yahoogroups.com
      >
      Subject: [wsjtgroup] noise blanker with ISCAT/FSK
      >
      >
      >
      >
      > Greetings all
      >
      > I'm in one of those
      neighborhoods where there is frequently some form of man made noise on 6M. I've noticed that turning on the noise blanker (K3) will usually reduce the noise floor by 2 to 3 db in SpecJT.
      >
      > I wonder, does it really do any
      good? IE can decode ISCAT/FSK signals that I couldn't decode with the noise blanker off?
      >
      > 73,
      >
      > Tom
      K6EU
      >

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