Ed, in one of the single-hour "contest mode trials" of WSJT 2.0 release
candidates, I hit 36 FT8 Q's in an hour. The abbreviated contest exchange
and the fact that most of the guys on had some contest-leanings, really
helped me in that hour.
I did that by going almnost 100% S&P where I get the one-per-minute cycle
time. It also really helped, that the trial hour band segment was not
completely jam-packed with signals.
In the big FT8 test weekend two weeks ago, I did not hit that rate. Even
when I focused and tried hard, I was usually only in the 20's for rate over
a while hour. I attribute this to many of the participants being less
contest-oriented.
The 120-per-hour run rate with the "TU, NOW" feature you mention, I think
you might be able to exploit that from P4, with callers spreading
themselves nicely and being patient to work a rare mult, but my feeling is
that us more common mults will hardly ever use that.
Again, I do not plan to do any FT8 during RTTY RU. However, I do appreciate
that the WSJT developers are learning some things about contesting to
improve rates and that non-contest FT8 rates have improved noticeably as a
result of this learning.
Ed, what rates were you able to hit in the FT8 weekend two weeks ago? I
think when the logs are collected that AA5AU may be able to advise us as to
peak rates hit by participants.
I did a quick-looksee at 3830 comments for FT8 roundup and see two
experienced contesters talking about a best hour of 38Q's, and more typical
"good hour" rates in the 20-30Q per hour range.
Tim N3QE
On Sat, Dec 15, 2018 at 6:53 PM Ed Muns <ed@w0yk.com> wrote:
> The discussion below is based on the WSJT-X non-contest protocol. Did you
> achieve any higher rates using the ARRL RTTY Roundup contest protocol in
> the
> FT8 Roundup two weeks ago? If so, probably not by much.
>
> That protocol will eventually have a "TU, NW" feature added which will get
> instantaneous rate to 120/hour. SO2R double that. Of course, average rate
> over a full hour will be less, but it will be much faster than then the
> current WSJT-X contest version.
>
> Ed W0YK
>
> -----Original Message-----
> From: CQ-Contest [mailto:cq-contest-bounces@contesting.com] On Behalf Of
> Tim
> Shoppa
> Sent: 14 December, 2018 11:49
> To: PVRC; cq-contest@contesting.com
> Subject: [CQ-Contest] Contesters review: new FT8, WSJT 2.0.0
>
> Keep in mind that I write the below from a "Contesters perspective" but
> maybe toned down for lower FT8 expectations.
>
> The official flag day for the new 77-bit FT8 messages was Monday Dec 14
> with the release of WSJT 2.0.0. Within a day almost all the activity had
> flipped over the the new version.
>
> I spent many hours (20+ hours) trying it out, with "RR73" turned on to get
> max rates, and it is an improvement. I had also been involved a little bit
> with pre-release WSJT software in the mock-contest hours and the "FT8 mock
> test" the first weekend of December.
>
> With the new WSJT 2.0.0 FT8, my peak S&P rate approaches one QSO a minute,
> my peak run rate approaches one QSO every 90 seconds, and my average rate
> over an hour is s 20 or more. It is hard to achieve peak rates on a super
> packed busy band like 20M FT8 but on a WARC band or "off peak" band it's
> quite achievable.
>
> Contrast that with the old WSJT where my typical FT8 rate in an hour was 8
> to 12 QSO's an hour.
>
> Now, in a real RTTY contest, 20 an hour is NOT something to write home
> about. It's dullsville, like maybe I should be doing something more fun
> like mopping the kitchen floor or raking leaves. But it is a real
> improvement for FT8 rate.
>
> I DO NOT PLAN to do ANY FT8 in RTTY Roundup. Again, 20 an hour is a bad
> rate if you are trying to enter that competitively.
>
> Tim N3QE
> Hardcore RTTY contester
> _______________________________________________
> CQ-Contest mailing list
> CQ-Contest@contesting.com
> http://lists.contesting.com/mailman/listinfo/cq-contest
>
>
_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest
|