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

Re: Sorting Frequencies...

Expand Messages
  • Alex
    It s possible the microprocessor could wait some increment of time to allow the VFO to stabilize. However, I believe it s highly unlikely because the wait
    Message 1 of 10 , Jun 11, 2011
    • 0 Attachment
      It's possible the microprocessor could wait some increment of time to allow the VFO to stabilize. However, I believe it's highly unlikely because the wait time would have to be a length of time equal to the largest settling time required by the VFO which would be the time required to change from the lowest possible frequency to the highest.

      If, on the other hand, the microprocessor waits for PLL lock then the delay time will vary because a smaller VFO frequency step will occur faster than a larger step. I'm fairly confident that the 386XT uses PLL lock indication to listen on a frequency before moving on to the next programmed frequency. In that case search time is impacted by frequency sorting.

      However, whether sorted or unsorted, the reality is that the difference is small or unnoticeable for most users if the number of individual frequencies scanned at any given time is small (low tens of frequencies). For users who are routinely scanning hundreds of frequencies then the difference between sorted and unsorted may manifest.

      --- In BCD396XT@yahoogroups.com, MCH <mch@...> wrote:
      >
      > If the micro is using a delay of X to wait for a VCO change, no amount
      > of VFO change will have any effect on the delay, as the micro will
      > always wait X mS. Do you not agree that the micro could operate this way?
      >
      > It's the microprocessor that controls all the timing regardless of any
      > physical components involved. (feedback line issues aside)
      >
      > If you ignore the micro's action, any timing issues are moot. It's the
      > micro that is in charge of everything that happens.
      >
      > Joe M.
      >
      > Alex wrote:
      > > In what I described the microprocessor has nothing to do with the speed of the VFO changing frequency so therefore the step size does matter.
      > >
      > > A VFO can only change frequency at a certain speed because of the way it's controlled (usually by a varactor diode). It's not instant frequency changes from one frequency to another. That is where the majority of the delay comes from. I specifically ignored PLL lock because that's constant for any frequency change. But the actual frequency change delay is not.
      > >
      > > --- In BCD396XT@yahoogroups.com, MCH <mch@> wrote:
      > >> All great info, but you also have to look at the microprocessor and how
      > >> it is programmed.
      > >>
      > >> Scenario A: The micro knows when the PLL is locked (through a "PLL
      > >> locked" feedback line), and then checks the COS for activity. If none is
      > >> found, it moves on to the next channel. If activity is found, it
      > >> evaluates the channel for proper CTCSS/CDCSS/NAC/DataStream/whatever.
      > >>
      > >> Scenario B: The micro simply waits X mS (or uS) for the PLL to lock, and
      > >> then performs its evaluation. (where X is a maximum tolerance)
      > >>
      > >> If scenario B is the case, there would be no difference whether the VCO
      > >> has moved 10 Hz or 10 MHz or 1000 MHz, as the timing would be all the same.
      > >>
      > >> It is likely that Uniden IS using a PLL Lock feedback line in which case
      > >> there may be some advantage to sequential frequency tuning.
      > >>
      > >> In the real world, the time you are saving in sequential frequency
      > >> scanning is likely not worth the aggravation you would have in finding a
      > >> specific channel when your channels are arranged in the order: Town A PD
      > >> 1, Town B FD 3, Town E EMS 2, Town A EMS 2, Etc.
      > >>
      > >> Yes, you may be able to scan your channels in 9.7 seconds as opposed to
      > >> 10 seconds, but your channels would be in no logical order whatsoever.
      > >> Is that worth saving 0.3 seconds in scan time?
      > >>
      > >> Joe M.
      > >>
      > >> Alex wrote:
      > >>> DMA doesn't affect how the VFO tunes. DMA is simply a way of organizing the storage memory most efficiently for storage of items. Non-DMA scanners remembered channels in fixed banks with no ability to group them logically. The DMA removes those limitations. But there is still only one VFO in the radio and it must tune each frequency you scan.
      > >>>
      > >>> There is a small advantage to tuning the VFO from one adjacent frequency to the next because there is a finite slewing rate for the tuning and a period of time for frequency/phase locking (in FM reception). The frequency/phase locking is about the same no matter how tuning occurs but it does take the tuner some time to tune from one frequency to the next. If you're tuning from say 100 MHz to 110 MHz (a step of 10 MHz) then perhaps that takes 2.3 milliseconds. But a jump from 100 MHz to 200 MHz (a 100 MHz step) would take 23.0 milliseconds (assuming the slew rate is constant).
      > >>>
      > >>> So run hypothetical numbers and compare the results. Say you have eleven channels spaced 10 MHz apart (100 MHz to 200 MHz) and the radio takes 2 milliseconds to change by 10 MHz (rounding off to make the numbers nicer). We're not including the lock time required to acquire the signal and play audio. That is assumed here to be constant regardless of the jump in frequency.
      > >>>
      > >>> If the channels are in order (100, 110, 120, 130...200) then that would be ten steps of 10 MHz or 20 ms total time to scan up. There is also the one last part which is the 'wrap-around' when you hit 200 and jump back to 100 to start the scan again. That's a jump of 100 MHz which is another 20 ms. So looping through one complete scan and starting off where you were takes 40 ms if the steps are consecutive.
      > >>>
      > >>> Now let's put the channels out of order:
      > >>> 100, 130, 200, 180, 150, 160, 120, 110, 140, 170, 190
      > >>>
      > >>> It's still the same eleven channels spanning 100 MHz to 200 MHz but they're now jumbled up so the steps aren't uniform. In this case the steps are:
      > >>>
      > >>> 30, 70, 20, 30, 10, 40, 10, 30, 30, 20
      > >>>
      > >>> So now the time delays for each of those steps are in milliseconds (2 ms per 10 MHz):
      > >>> 6, 14, 4, 6, 2, 8, 2, 6, 6, 4
      > >>>
      > >>> That's a grand total of 58 ms for a one-way trip up the list. The wrap-around for this is a 90 MHz step (end of the list is 190, start of the list is 100, the difference is 90). A 90 MHz step is 18 ms to add to the 58 ms for a grand total of 76 ms for a complete scan.
      > >>>
      > >>> So you have 40 ms for a frequency sorted list and 76 ms for an unsorted random list. You gain a 47% improvement in speed by sorting over the given random list. Note that this isn't the worst case list since there are some small steps, it's just a random list.
      > >>>
      > >>>
      > >>> Now, the reality is that the VFO in the radio is most likely faster than 2ms but the result still applies. As you scan larger and larger sets of frequencies, the total time required to scan a sorted vs. unsorted list will diverge.
      > >>>
      > >>>
      > >>> --- In BCD396XT@yahoogroups.com, Lance <milcom_chaser@> wrote:
      > >>>> I have the Alpha Tags as the reference to the way objects are organized in this radio.
      > >>>>
      > >>>> As an example, down at a group level within a conventional system, I might list Sheriff, then Sheriff Input, Co. Fire, then Co. Fire Input, Police, Police Input, etc...
      > >>>>
      > >>>> However, I used the "Scanner Optimizer" tool in FreeSCAN. Which sorts frequencies in numerical order.
      > >>>>
      > >>>> So, given the speed at which this radio scans, and if that even applies here in a DMA based radio architecture,
      > >>>>
      > >>>> Does sorting really effect the time it takes for this radio to tune a programmed frequency and then play the audio?
      > >>>>
      > >>>> How much time in Milliseconds are we talking here between the radio demodulating audio between sorted and unsorted
      > >>>>
      > >>>> frequencies? Is this even valid in this type of Architecture?
      > >>>>
      > >>>> Thanks,
      > >>>>
      > >>>> -Lance
      > >>>>
      > >>>
      > >>>
      > >>>
      > >>> ------------------------------------
      > >>>
      > >>> Yahoo! Groups Links
      > >>>
      > >>>
      > >>>
      > >>>
      > >>> ------------------------------------------------------------------------
      > >>>
      > >>>
      > >>> Internal Virus Database is out of date.
      > >>> Checked by AVG - www.avg.com
      > >>> Version: 9.0.783 / Virus Database: 271.1.1/2746 - Release Date: 03/14/10 03:33:00
      > >>>
      > >
      > >
      > >
      > >
      > > ------------------------------------
      > >
      > > Yahoo! Groups Links
      > >
      > >
      > >
      > >
      >
    • mch@nb.net
      Bingo. That s exactly my point. The micro could be waiting the full time no matter what the jump. I agree they are likely using the lock indication, as Uniden
      Message 2 of 10 , Jun 11, 2011
      • 0 Attachment
        Bingo. That's exactly my point. The micro could be waiting the full time no matter what the jump.

        I agree they are likely using the lock indication, as Uniden rarely takes shortcuts in their products.

        Still, I maintain that any advantage in scanning speed would be offset by the annoyance in finding channels which will be in no logical order at all (as they would only be in frequency order).

        Joe M.

        On Sat 11/06/11 7:37 PM , "Alex" agcme2002@... sent:
        > It's possible the microprocessor could wait some increment of time to allow
        > the VFO to stabilize. However, I believe it's highly unlikely because the
        > wait time would have to be a length of time equal to the largest settling
        > time required by the VFO which would be the time required to change from
        > the lowest possible frequency to the highest.
        > If, on the other hand, the microprocessor waits for PLL lock then the delay
        > time will vary because a smaller VFO frequency step will occur faster than
        > a larger step. I'm fairly confident that the 386XT uses PLL lock
        > indication to listen on a frequency before moving on to the next programmed
        > frequency. In that case search time is impacted by frequency sorting.
        > However, whether sorted or unsorted, the reality is that the difference is
        > small or unnoticeable for most users if the number of individual
        > frequencies scanned at any given time is small (low tens of frequencies).
        > For users who are routinely scanning hundreds of frequencies then the
        > difference between sorted and unsorted may manifest.
        > --- In BCD3
        > 96XT@yahoogroups.com, MCH wrote:>
        > > If the micro is using a delay of X to wait for a
        > VCO change, no amount > of VFO change will have any effect on the delay,
        > as the micro will > always wait X mS. Do you not agree that the
        > micro could operate this way?>
        > > It's the microprocessor that controls all the
        > timing regardless of any > physical components involved. (feedback line
        > issues aside)>
        > > If you ignore the micro's action, any timing
        > issues are moot. It's the > micro that is in charge of everything that
        > happens.>
        > > Joe M.
        > >
        > > Alex wrote:
        > > > In what I described the microprocessor has
        > nothing to do with the speed of the VFO changing frequency so therefore the
        > step size does matter.> >
        > > > A VFO can only change frequency at a
        > certain speed because of the way it's controlled (usually by a varactor
        > diode). It's not instant frequency changes from one frequency to another.
        > That is where the majority of the delay comes from. I specifically ignored
        > PLL lock because that's constant for any frequency change. But the actual
        > frequency change delay is not.> >
        > > > --- In BCD3
        > 96XT@yahoogroups.com, MCH wrote:> >> All great info, but you also have to
        > look at the microprocessor and how > >> it is programmed.
        > > >>
        > > >> Scenario A: The micro knows when the
        > PLL is locked (through a "PLL > >> locked" feedback line), and then
        > checks the COS for activity. If none is > >> found, it moves on to the next channel.
        > If activity is found, it > >> evaluates the channel for proper
        > CTCSS/CDCSS/NAC/DataStream/whatever.> >>
        > > >> Scenario B: The micro simply waits X mS
        > (or uS) for the PLL to lock, and > >> then performs its evaluation. (where X
        > is a maximum tolerance)> >>
        > > >> If scenario B is the case, there would
        > be no difference whether the VCO > >> has moved 10 Hz or 10 MHz or 1000 MHz,
        > as the timing would be all the same.> >>
        > > >> It is likely that Uniden IS using a PLL
        > Lock feedback line in which case > >> there may be some advantage to
        > sequential frequency tuning.> >>
        > > >> In the real world, the time you are
        > saving in sequential frequency > >> scanning is likely not worth the
        > aggravation you would have in finding a > >> specific channel when your channels are
        > arranged in the order: Town A PD > >> 1, Town B FD 3, Town E EMS 2, Town A
        > EMS 2, Etc.> >>
        > > >> Yes, you may be able to scan your
        > channels in 9.7 seconds as opposed to > >> 10 seconds, but your channels would be
        > in no logical order whatsoever. > >> Is that worth saving 0.3 seconds in
        > scan time?> >>
        > > >> Joe M.
        > > >>
        > > >> Alex wrote:
        > > >>> DMA doesn't affect how the VFO
        > tunes. DMA is simply a way of organizing the storage memory most
        > efficiently for storage of items. Non-DMA scanners remembered channels in
        > fixed banks with no ability to group them logically. The DMA removes those
        > limitations. But there is still only one VFO in the radio and it must tune
        > each frequency you scan.> >>>
        > > >>> There is a small advantage to
        > tuning the VFO from one adjacent frequency to the next because there is a
        > finite slewing rate for the tuning and a period of time for frequency/phase
        > locking (in FM reception). The frequency/phase locking is about the same
        > no matter how tuning occurs but it does take the tuner some time to tune
        > from one frequency to the next. If you're tuning from say 100 MHz to 110
        > MHz (a step of 10 MHz) then perhaps that takes 2.3 milliseconds. But a
        > jump from 100 MHz to 200 MHz (a 100 MHz step) would take 23.0 milliseconds
        > (assuming the slew rate is constant).> >>>
        > > >>> So run hypothetical numbers and
        > compare the results. Say you have eleven channels spaced 10 MHz apart (100
        > MHz to 200 MHz) and the radio takes 2 milliseconds to change by 10 MHz
        > (rounding off to make the numbers nicer). We're not including the lock
        > time required to acquire the signal and play audio. That is assumed here
        > to be constant regardless of the jump in frequency.> >>>
        > > >>> If the channels are in order (100,
        > 110, 120, 130...200) then that would be ten steps of 10 MHz or 20 ms total
        > time to scan up. There is also the one last part which is the
        > 'wrap-around' when you hit 200 and jump back to 100 to start the scan
        > again. That's a jump of 100 MHz which is another 20 ms. So looping
        > through one complete scan and starting off where you were takes 40 ms if
        > the steps are consecutive.> >>>
        > > >>> Now let's put the channels out of
        > order:> >>> 100, 130, 200, 180, 150, 160, 120,
        > 110, 140, 170, 190> >>>
        > > >>> It's still the same eleven channels
        > spanning 100 MHz to 200 MHz but they're now jumbled up so the steps aren't
        > uniform. In this case the steps are:> >>>
        > > >>> 30, 70, 20, 30, 10, 40, 10, 30, 30,
        > 20> >>>
        > > >>> So now the time delays for each of
        > those steps are in milliseconds (2 ms per 10 MHz):> >>> 6, 14, 4, 6, 2, 8, 2, 6, 6,
        > 4> >>>
        > > >>> That's a grand total of 58 ms for a
        > one-way trip up the list. The wrap-around for this is a 90 MHz step (end
        > of the list is 190, start of the list is 100, the difference is 90). A 90
        > MHz step is 18 ms to add to the 58 ms for a grand total of 76 ms for a
        > complete scan.> >>>
        > > >>> So you have 40 ms for a frequency
        > sorted list and 76 ms for an unsorted random list. You gain a 47%
        > improvement in speed by sorting over the given random list. Note that this
        > isn't the worst case list since there are some small steps, it's just a
        > random list.> >>>
        > > >>>
        > > >>> Now, the reality is that the VFO in
        > the radio is most likely faster than 2ms but the result still applies. As
        > you scan larger and larger sets of frequencies, the total time required to
        > scan a sorted vs. unsorted list will diverge.> >>>
        > > >>>
        > > >>> --- In BCD3
        > 96XT@yahoogroups.com, Lance wrote:> >>>> I have the Alpha Tags as the
        > reference to the way objects are organized in this radio.> >>>>
        > > >>>> As an example, down at a group
        > level within a conventional system, I might list Sheriff, then Sheriff
        > Input, Co. Fire, then Co. Fire Input, Police, Police Input,
        > etc...> >>>>
        > > >>>> However, I used the
        > "Scanner Optimizer" tool in FreeSCAN. Which sorts frequencies in
        > numerical order.> >>>>
        > > >>>> So, given the speed at which
        > this radio scans, and if that even applies here in a DMA based radio
        > architecture,> >>>>
        > > >>>> Does sorting really effect the
        > time it takes for this radio to tune a programmed frequency and then play
        > the audio?> >>>>
        > > >>>> How much time in Milliseconds
        > are we talking here between the radio demodulating audio between sorted and
        > unsorted> >>>>
        > > >>>> frequencies? Is this even
        > valid in this type of Architecture?> >>>>
        > > >>>> Thanks,
        > > >>>>
        > > >>>> -Lance
        > > >>>>
        > > >>>
        > > >>>
        > > >>>
        > > >>>
        > ------------------------------------> >>>
        > > >>> Yahoo! Groups Links
        > > >>>
        > > >>>
        > > >>>
        > > >>>
        > > >>>
        > ------------------------------------------------------------------------> >>>
        > > >>>
        > > >>> Internal Virus Database is out of
        > date.> >>> Checked by AVG - www.avg.com > >>> Version: 9.0.783 / Virus Database:
        > 271.1.1/2746 - Release Date: 03/14/10 03:33:00> >>>
        > > >
        > > >
        > > >
        > > >
        > > >
        > ------------------------------------> >
        > > > Yahoo! Groups Links
        > > >
        > > >
        > > >
        > > >
        > >
        >
        >
        >
        >
        > ------------------------------------
        >
        > Yahoo! Groups Links
        >
        > To visit your group on the web, go to:
        > http://groups.yahoo.com/group/BCD396XT/
        > Your email settings:
        > Individual Email | Traditional
        >
        > To change settings online go to:
        > http://groups.yahoo.com/group/BCD396XT/join(Yahoo! ID required)
        >
        > To change settings via email:
        > BCD396XT-digest@yahoogroups.com BCD396XT-fullfeatured@yahoogroups.com
        > To unsubscribe from this group, send an email to:
        > BCD396XT-unsubscribe@yahoogroups.com
        > Your use of Yahoo! Groups is subject to:
        > http://docs.yahoo.com/info/terms/
        >
        >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.