[ADXA] N5J background

Sun Aug 11 19:19:06 EDT 2024

Over the past couple of days, I've seen some of this posted around the
web.  Great to set it all in one place and from the "Horse's mouth" for a
change. I've never decoded them below -16db.  Once they hit the magic point
of -15, it's hit and miss until they build up a few more db, then pretty
solid.

One thing we need to remember is that ALL 9 streams, plus the "Verified"
message are always being transmitted even though some don't contain any QSO
information.  So you are not missing anything if you only see a couple
streams but do see the verified message as that's all they sent.

One thing that I always see as funny, is when some person sees a calling
party that they "need" for one reason or another, and starts calling them
on the EVEN timeslot.  Don't they realize that that station, if running SF
will never see them as their format is not an acceptable Rx format when the
station they are calling is in SH mode.  But with FEC, it doesn't really
cause a problem for the masses.

Another thing that the developers are looking at implementing before the
next SF/SH DXP is something that a few folks may find controversial.  That
is inhibiting "blind calling" of the DXP.  Obviously, if you can't hear
them at the get go, why call.  In order to call the DXP, you must decode
them at least once during each session you click on Enable Tx, which is
currently hard coded at 5 minutes.  If you don't hear them again, then you
can't Enable Tx again until you do.  Sounds good to me when looking at the
DXP operators side of things.  Not so many wasted responses to those that
don't hear them.

Thanks Bill.

73's
George - WB5JJJ
HoIP - 100105

On Sun, Aug 11, 2024 at 6:00 PM Bill Priakos via ADXA <adxa at mailman.qth.net>
wrote:

> Hello, ADXA;
>
> Some info on N5J condx.
>
> https://groups.io/g/wsjtgroup/message/3112
>
> from Glenn, AF8C
>
>
> ------------------------------
>
>  Ned Stearns  <https://groups.io/g/wsjtgroup/message/3112>
>
> Aug 10   #3112    <https://groups.io/g/wsjtgroup/message/3112>
>
>
>
> de AA7A here. I am the digi team lead on N5J and have been behind the
> piles for over 24 hours so far. Let me give everyone some background info.
>
> We manage the PC clocks to balance the DT error between tx and rx . There
> is an annoying tx audio delay caused by DSP activities in SmartSDR. We
> check time error daily and make minor time adjustments at times to keep the
> SuperFox's contribution to timing errors to the minimum on both tc and rx
> intervals.
>
> There is also an issue with TX offsets in the Flex. Again, we run daily
> tests with my home station used as truth in order to keep the Fox sync tone
> as close to 750 Hz on every band. With frequent band changes, it might be
> off for a short while while we re-adjust.
>
> We are concerned about the completion rates. Using my home station as a
> quality control tool, it decodes Fox transmissions 100% for signals down to
> -16. It appears that we are not decoded by Hounds anwhere near as well as
> we can decode Hounds. My thoughts on why. Take your pick
>
> 1. Extremely quiet on Jarvis. We can hear better than Hounds by many dB.
>
> 2. We are running only 10 w at times to dodge interference to other Jarvis
> stations caused by salt spray corrosion on the multiband antennas. Hounds
> are running much more power. We are working procedures to deal with
> corrosion effects, but everyone needs to be patient while they get their
> magic moment. We will bee here another week or so.
>
> 3. Some Hounds are still using version RC5
>
> 4. We are getting reports of DQRM in EU.
>
> We are hoping to get ahead of the corrosion business and try to get to
> full output power on the Fox transmitters. It will take time to sort thru
> all the variables.
>
> 73 and see you all on the screen.
>
> Ned / AA7A
>
> On 8/10/2024 8:49 AM, Joe WB9SBD wrote:
>
> *Hide quoted text*
> <https://groups.io/g/wsjtgroup/message/3112#quoted-241628374>
>
> I have never understood why anyone would not be really close to dead on
> with the clock. EVEN if you do not have access to any good time source.
>
> Just look at the freaking waterfall!
>
> look at the average start and stop times of the hundreds of stations
> transmitting.
>
> are those signal evenly spaced with the TX/RX divider lines? No? adjust
> your clock till it is.
>
> No external clock needed!
>
> Joe WB9SBD
>
> On 8/10/2024 10:42 AM, JP Tucson, AZ wrote:
>
> Hi Mike,
>
>
>
> I can tell you folks the REAL PROBLEM is that there are a bunch of yahoo's
> out there!
>
>
>
> First and worst is the DELIBERATE QRM idiots!  These knuckleheads have
> been flooding both 15m & 20m with MUSIC, which of course is quite illegal!
>
>
>
> On 30m there are folks south of the border who are pirates and they are
> cab drivers who are using VOICE on 30m, which is CW & DATA only! Also very
> illegal, but unfortunately nothing we can do, since the FCC has no control
> down there.
>
>
>
> I have also noticed 1 other possible issue. On the FT8 SF, N5J is
> consistently running with a DT of .3 seconds. Hmmm, you would think they
> would have a GPS disciplined clock.  Now, there may be an issue there of
> this RIB (Remote In Box) having a delay effect... don't know. But somewhere
> I thought one of you dev. team members said the timing was a bit more
> critical with SF... is that true?
>
>
>
> 73
>
>
>
>
>
>
>
> On Sat, Aug 10, 2024 at 8:31 AM Mike Black W9MDB via groups.io <mdblack98=
> yahoo.com at groups.io> wrote:
>
> They were 4.4 QSOs/min on the 1st day for FT8.
>
>
>
> They now show 3.7/min for FT8 (average over all days so their rate now is
> notably less then 3.7/min)
>
>
>
> Over all QSO rate for the dxpedition has dropped a bit from 8.26/min on
> 1st day to 6.5/min on the 8th.
>
> This data comes from the statistics tab here
> https://jarvisisland2024.com/qsl-info-log-search/log-oqrs
>
>
>
> [image: Inline image]
>
> Mike W9MDB
>
>
>
>
>
>
>
> On Saturday, August 10, 2024 at 10:12:56 AM CDT, Saku OH1KH <nylsak at ...>
> wrote:
>
>
>
>
>
> HI!
> I have been watching (and trying) N5J now on Friday and Saturday and qso
> rate seems to be very poor.
> I have seen many N5J periods without any completed qsos (no RR73s) and
> where they exist there are only one or two of them on one period.
>
>
>
> Problem on hound side are tons of tuning stations that break up sensitive
> superfox signal and so RR73 may not be copied by hound that leads to
> unneeded retries over and over again.
>
> Looked from hound side it looks like fox's problem may be too long
> callsign queues.
> Conditions on band are changing and hounds may change their TX frequency
> rolling over other, already queued, callsigns.
>
> For that reason fox should keep the queue as short as possible (maybe
> around 10 calls) so that response for caller could happen quite fast after
> his over is received by fox.
> I do not know how the queue is handled, but as said, from hound's view it
> should not have many calls (old ones) to avoid changes of fox's receiving
> conditions while queue is spooled.
>
>
>
>
>
>
>
> --
>
> 73 - John - N7GHZ
>
>
>
>
>
> trying to catch the crew at OH0AM, too!
>
>
>
> 73, Bill
>
>
>
> *Bill Priakos – W5SJ*
>
> *10 Free Ferry Heights*
>
> *Fort Smith, AR 72903*
>
> *479.461.8368*
>
>
>
>
>
>
> ______________________________________________________________
> ADXA mailing list
> Home: http://mailman.qth.net/mailman/listinfo/adxa
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:ADXA at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.qth.net/pipermail/adxa/attachments/20240811/d34abf33/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.gif
Type: image/gif
Size: 91369 bytes
Desc: not available
URL: <http://mailman.qth.net/pipermail/adxa/attachments/20240811/d34abf33/attachment-0001.gif>