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

Re: [MT63] 20m Olivia experiment results

Expand Messages
  • Mark Miller
    Dave, Your results are interesting. ... This is interesting. I wonder if either station was able to reduce receiver bandwidth to 1000Hz, 500Hz, and 250 Hz?
    Message 1 of 3 , Mar 7, 2005
    • 0 Attachment
      Dave,

      Your results are interesting.

      At 12:46 AM 3/7/2005, you wrote:
      >Steve and I started by establishing communications in Olivia's default
      >mode of 32
      >tones within 1,000 Hz bandwidth. We both experienced 100 percent print at
      >both 50
      >W and 5 W, notably after we moved up 500Hz to avoid a packet station.
      >
      >Settings of 16 tones and 500 Hz produced the same results at both power
      >settings,
      >just with slightly slower print. At 50 W, Steve reported that I was S7 on
      >his meter,
      >while I sent him an S6. The noise level at my QTH was sitting right around
      >S3, and
      >when we went QRP, Steve's signals hovered between S3 and S4, but I never
      >lost a
      >character.
      >
      >Then we went to a setting of 32 tones in 250 Hz of bandwidth. What a
      >change! The
      >text, while running between 75 to 100 percent copy, slowed to a crawl at
      >7.81 baud
      >and 0.6 characters per second. The spacing of the tones was quite narrow.

      This is interesting. I wonder if either station was able to reduce
      receiver bandwidth to 1000Hz, 500Hz, and 250 Hz? There is a penalty for
      increasing the number of tones, but that can be offset by reducing the
      bandwidth thus increasing the S/N ratio. The other affect with Olivia is
      increasing the interleaving depth, which seems to have had little affect
      during your test. I would guess that the QSB was not too bad (boy talk
      about relative terms). Another way of saying this is that when your fade
      margin is higher, increasing the number of tones has little constructive
      affect.


      >Next we went to 8/250 and the print was much better, near perfect at both
      >at 50 and
      >5 W. As a tuning aid, I was sampling the audio out of the back of the 706
      >and feeding
      >that to a old laptop running DigiPan, so I could use the panoramic
      >display. The
      >difference between 32 tones and 8 tones in the same bandwidth was visually
      >appreciable.

      I don't think interleaving was helping you that much, and I suspect the
      receiver bandwidth did not change, so you were paying less penalty for the
      number of tones.

      >Much wider tones made for a more robust contact. While a tad slower
      >throughput than with the default 32/1000 (2.4 cps at 32/1000 vs. 1.5 cps
      >at 8/250),
      >we were having a solid QSO and taking up one-quarter of the bandwidth,
      >which all
      >digital operators can appreciate.

      I think you are saying that the tone spacing was wider, as the tone
      duration is a function of baud rate.

      >Then Steve suggested we go for broke and set up at 4 tones and 125 Hz
      >bandwidth.
      >Surprisingly, this set-up was as good or better than 8/250. The baud rate
      >was the
      >same (31.25), although we lost one-third of the character rate, dropping
      >to 1.0 cps.

      This shows again that the interleaving depth was not a factor.

      >We decided to go to low power again, and at the end of Steve's
      >transmission, he
      >revealed he'd been at 200 mW, as low as the Mk V Field would go. Whoa! At
      >times his
      >signal faded out altogether, but on the receiving end, I only lost one
      >character.

      This is an example where interleaving depth and Viterbi decoding did help

      >The
      >software was actually copying below the noise!

      I doubt that it was copying below the noise. Even on an AGWN channel when
      the average noise and average signal are equal, 50 % of the time the signal
      is above the noise, and 50% of the time the signal is below the noise. If
      we spread out our bit errors by interleaving, and we have FEC, the protocol
      can "fill in the gaps" during the time that the signal falls below the
      noise. I hear this "below the noise" claim from PSKers all of the
      time. The average signal may be below the average noise, but there are
      times when that signal pops its head above the noise. Otherwise you would
      not copy.

      Very nice report Dave. It certainly gives us some things to think about.

      73,

      Mark N5RFX
    • Steve Friis
      Hi mark, and all others interested, My comments here. There were times when the audio was below the threshold of the ear (mine for what that is worth) to be
      Message 2 of 3 , Mar 7, 2005
      • 0 Attachment
        Hi mark, and all others interested,

        My comments here. There were times when the audio was below the
        threshold of the ear (mine for what that is worth) to be able to
        distinguish from the noise. My bandwidth was 1.8 Khz for all tests. We
        did discuss over the air not being able to hear the audio signals. When
        Dave reduced his power level to the 1 watt level, I actually could not
        hear his audio for ALL of the test, which was at 4/125 setting. There
        was just vissible difference within the "crosshairs" in the waterfall. I
        would say that at these low levels, and with the slow responce of the
        decoder, precise receiver tuning via a schedule would be needed. Not a
        DX chasing mode like PSK31 can be. My obsevation is that this would be a
        fantastic tool for ARES and RACES traffic handling where 100% copy is
        needed.

        Steve/WM5Z
        Las Cruces, NM USA



        Mark Miller wrote:

        > Dave,
        >
        > Your results are interesting.
        >
        > At 12:46 AM 3/7/2005, you wrote:
        > >Steve and I started by establishing communications in Olivia's default
        > >mode of 32
        > >tones within 1,000 Hz bandwidth. We both experienced 100 percent
        > print at
        > >both 50
        > >W and 5 W, notably after we moved up 500Hz to avoid a packet station.
        > >
        > >Settings of 16 tones and 500 Hz produced the same results at both power
        > >settings,
        > >just with slightly slower print. At 50 W, Steve reported that I was
        > S7 on
        > >his meter,
        > >while I sent him an S6. The noise level at my QTH was sitting right
        > around
        > >S3, and
        > >when we went QRP, Steve's signals hovered between S3 and S4, but I never
        > >lost a
        > >character.
        > >
        > >Then we went to a setting of 32 tones in 250 Hz of bandwidth. What a
        > >change! The
        > >text, while running between 75 to 100 percent copy, slowed to a crawl at
        > >7.81 baud
        > >and 0.6 characters per second. The spacing of the tones was quite narrow.
        >
        > This is interesting. I wonder if either station was able to reduce
        > receiver bandwidth to 1000Hz, 500Hz, and 250 Hz? There is a penalty for
        > increasing the number of tones, but that can be offset by reducing the
        > bandwidth thus increasing the S/N ratio. The other affect with Olivia is
        > increasing the interleaving depth, which seems to have had little affect
        > during your test. I would guess that the QSB was not too bad (boy talk
        > about relative terms). Another way of saying this is that when your fade
        > margin is higher, increasing the number of tones has little constructive
        > affect.
        >
        >
        > >Next we went to 8/250 and the print was much better, near perfect at
        > both
        > >at 50 and
        > >5 W. As a tuning aid, I was sampling the audio out of the back of the
        > 706
        > >and feeding
        > >that to a old laptop running DigiPan, so I could use the panoramic
        > >display. The
        > >difference between 32 tones and 8 tones in the same bandwidth was
        > visually
        > >appreciable.
        >
        > I don't think interleaving was helping you that much, and I suspect the
        > receiver bandwidth did not change, so you were paying less penalty for
        > the
        > number of tones.
        >
        > >Much wider tones made for a more robust contact. While a tad slower
        > >throughput than with the default 32/1000 (2.4 cps at 32/1000 vs. 1.5 cps
        > >at 8/250),
        > >we were having a solid QSO and taking up one-quarter of the bandwidth,
        > >which all
        > >digital operators can appreciate.
        >
        > I think you are saying that the tone spacing was wider, as the tone
        > duration is a function of baud rate.
        >
        > >Then Steve suggested we go for broke and set up at 4 tones and 125 Hz
        > >bandwidth.
        > >Surprisingly, this set-up was as good or better than 8/250. The baud
        > rate
        > >was the
        > >same (31.25), although we lost one-third of the character rate, dropping
        > >to 1.0 cps.
        >
        > This shows again that the interleaving depth was not a factor.
        >
        > >We decided to go to low power again, and at the end of Steve's
        > >transmission, he
        > >revealed he'd been at 200 mW, as low as the Mk V Field would go.
        > Whoa! At
        > >times his
        > >signal faded out altogether, but on the receiving end, I only lost one
        > >character.
        >
        > This is an example where interleaving depth and Viterbi decoding did help
        >
        > >The
        > >software was actually copying below the noise!
        >
        > I doubt that it was copying below the noise. Even on an AGWN channel
        > when
        > the average noise and average signal are equal, 50 % of the time the
        > signal
        > is above the noise, and 50% of the time the signal is below the
        > noise. If
        > we spread out our bit errors by interleaving, and we have FEC, the
        > protocol
        > can "fill in the gaps" during the time that the signal falls below the
        > noise. I hear this "below the noise" claim from PSKers all of the
        > time. The average signal may be below the average noise, but there are
        > times when that signal pops its head above the noise. Otherwise you
        > would
        > not copy.
        >
        > Very nice report Dave. It certainly gives us some things to think about.
        >
        > 73,
        >
        > Mark N5RFX
        >
        >
        >
        >
        >
        > << Try MT63 on 80m - great fun!>>
        >
        > - The MT63 Reflector -
        > MT63@egroups.com
        >
        > (To unsubscribe. send email to
        > MT63-unsubscribe@onelist.com)
        >
        >
        >
        > *Yahoo! Groups Sponsor*
        > ADVERTISEMENT
        > click here
        > <http://us.ard.yahoo.com/SIG=129jhsrjj/M=298184.6018725.7038619.3001176/D=groups/S=1705034690:HM/EXP=1110286816/A=2593423/R=0/SIG=11el9gslf/*http://www.netflix.com/Default?mqso=60190075>
        >
        >
        >
        > ------------------------------------------------------------------------
        > *Yahoo! Groups Links*
        >
        > * To visit your group on the web, go to:
        > http://groups.yahoo.com/group/MT63/
        >
        > * To unsubscribe from this group, send an email to:
        > MT63-unsubscribe@yahoogroups.com
        > <mailto:MT63-unsubscribe@yahoogroups.com?subject=Unsubscribe>
        >
        > * Your use of Yahoo! Groups is subject to the Yahoo! Terms of
        > Service <http://docs.yahoo.com/info/terms/>.
        >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.