Stew Perry Topband Challenge - 2019
Call: K1LT
Operator(s): K1LT
Station: K1LT
Class: Single Op HP
QTH: EM89ps
Operating Time (hrs): 13:44
Radios: SO2R
Summary:
Total: QSOs = 828 Total Score = 4,947
Club: Mad River Radio Club
Comments:
I again had lots of vacation time to consume this year, so I spent
some time integrating the new phased array receiving system with the
rest of my station in time for the 2019 Stew Perry Topband Distance
Challenge. The new system has a 2x2 broadside / end-fire array of
short verticals feeding 4 SDR receivers and an FPGA implementing the
beam-forming algorithm in hardware. The verticals are arranged to
look north or south which fills in a gap in the existing arrays. This
part of the project works.
The hard part is adding that new capability to the systems that
already work without degrading or breaking anything. This part of the
project is not yet complete.
First problem: I have 3 antenna / receiver systems now. The 2
existing systems are phase-locked to VFOs provided by Si570 chips.
These VFOs are very stable and I have tweaked them to within 1 Hz of
each other. Thus when the same signal is audible of both arrays, the
beat note varies slowly. The newest receiver does not yet have a
stable oscillator, so there is an annoying wavering on strong signals.
There were lots of strong signals in this contest.
Second problem: the 2 existing systems run on an Intel-based PC while
the new system runs on dedicated ARM-based hardware. For some reason,
the 2 different architectures show different signal processing
latencies even though they should not. This provides a fast "echo"
effect on strong signals. This problem seemed worse at the beginning
of the contest and was less troublesome by the end.
Third and worst problem: the new 2x2 short vertical array consumed the
2 verticals I had been using for my spotting receiver. To recover the
full-duplex spotting capability, I used a couple of magic-Tee
splitters to provide the signals for the spotting receiver. The
nulling circuit nicely suppresses my transmit signal but during the
contest, the spotting receiver showed severe distortion and noise
from blocking. I haven't figured out what is wrong. Fortunately,
this problem was largely mitigated by the propagation conditions
during the contest.
Fourth problem: one of the hacks I added to my old software for
integration introduced a bug that manifested as a software hang after
a few hours of operation. I diagnosed this problem during the ARRL
160 test but I forgot to fix it. Twice during the night by receivers
stopped responding to the controls. Both times I restarted the
software while CQing by using the receiver in the K3. One can never
have too many receivers as long as the complexity does not cause
premature insanity.
The entire operation reminds me of the movie trope of the great
battleship (or starship) leaving port on a daring mission with
sputtering and broken engine sound effects. This statement is
supposed to evoke a chuckle.
So the immediate goal of all this effort was to be able to work a UA9
or a South American or maybe just Tennessee or Florida without
spending 3 minutes going AGN? AGN? to some caller who can't understand
while I can't hear him as I desperately hunt for the correct receiving
antenna to use.
During my time off I also tried to adjust my sleeping schedule to make
staying up all night easier. The plan was to start at 2200Z and take a
break sometime during the night when the rate got sufficiently low.
I started at 2152Z because I couldn't stand to watch all of those
traces on the waterfall without listening. There were strong
Europeans everywhere 20 minutes before sunset. Many of them could not
hear me but a few could. My first QSO was LA3MHA. I figured I could
search and pounce for a few minutes and then stop for 31 minutes. Ha!
I started CQing and 6 minutes after my first contact I had an 18 point
QSO with UW2M in the log. Not yet sunset. At that point the strategy
changed to "work 'em now while you can".
The DX QSO rate stayed reasonable until the conclusion of European
sunrise plus an hour. The DX QSO rate stayed fairly constant even as
the K/VE QSO rate declined. On a normal good SP TBDC, there is a
burst of DX at local sunset and another burst for Eu sunrise. So
conditions were good enough that the Europeans stayed up all night to
work us.
all K/VE DX
hour QSO QSOs QSOs
21 14 12 2
22 113 89 24
23 100 73 27
00 87 62 25
01 71 52 19
02 77 57 20
03 76 62 14
04 50 35 15
05 53 25 28
06 42 18 24
07 47 29 18
08 21 12 9
09 7 4 3 (KH6 + ZL)
11 19 18 1 (JA)
12 30 29 1 (KL7)
At 0914Z I decided the QSO rate was sufficiently low that I decided to
take a 90 minute nap. For some reason the 90 minute nap required more
than 2 hours. This meant I spent an extra hour after sunrise
operating when I could have been finished. Nevertheless, the last
hour rate was probably as good as the 5am rate.
Here is the annual propagation and participation "trends" table:
raw raw points cooked cooked points cooked-raw
year QSOs score per QSO QSOs score per QSO ratio
2005 491 2033 4.14 483 2439 5.05 1.22
2006 604 2224 3.68 didn't submit log in time
2007 691 3712 5.37 669 4293 6.42 1.19
2008 633 3328 5.26 617 3895 6.31 1.20
2009 761 4006 5.26 737 4871 6.61 1.26
2010 642 2477 3.86 623 2931 4.70 1.22
2011 656 2501 3.81 642 3169 4.94 1.27
2012 679 3214 4.73 667 3986 5.97 1.24
2013 723 3559 4.92 710 4525 6.37 1.30
2014 667 1937 2.90 655 2474 3.77 1.28
2015 620 1934 3.12 614 2440 3.97 1.27
2016 759 3381 4.45 742 4204 5.66 1.27
2017 antenna feedline disaster - quit with only 101 QSOs
2018 785 3587 4.57 764 4238 5.55 1.18
2019 828 4947 5.97 813 5328 6.55 (preliminary)
Note: the raw numbers are before log checking while the cooked numbers
are after log checking. The cooked numbers exclude busted QSOs and
include the per-QSO multiplier for low power and QRP contacts. Note
that the score "boost" for low power and QRP QSOs seems to hover
around 20% while the points per QSO goes up and down with propagation
trends.
DX worked: 9A (3), DL (40), E7, EA (2), EI (2), F (9), G (18), GI, GM
(4), GW (3), HA (8), HB, HR, I (4), JA, KH6, KL, KP2 (3), LA (5), LY
(9), LZ (4), OH (9), OH0, OK (12), OM (6), ON (6), OZ (2), PA (8),
PJ2, S5 (6), SM (12), SP (7), SV, UA (14), UR (12), XE (2), YL (6),
YO, and ZL (2) for a total of 39 entities and 229 DX QSOs. Missed
South America completely but worked more DLs than VEs. Best DX was
ZL4AS (29), followed by ZL3IX (28), JH2FXK (22), and R6YY (19).
The man-made noises all stayed away and there was no QRN except for a
tiny crash or two when the rain started around sunrise. Overall,
experimenting with new equipment didn't make things worse (unlike the
last ARRL 160 test) and conditions were probably best ever. HNY!
Equipment: Elecraft K3S, P3, K3, P3, half power Alpha 8410, homebrew
SO2R stuff, too many computers, 60-foot "tee" vertical over 70
125-foot radials, 2x4 broadside end-fire array of short verticals and
SDR receivers for beam steering to the east, 2x3 broadside end-fire
array of short verticals and SDR receivers for beam steering to the
west, and 2x2 broadside end-fire array of short verticals to the
north or south.
Posted using 3830 Score Submittal Forms at: http://www.3830scores.com/
______________________________________________
3830 mailing list
3830@contesting.com
http://lists.contesting.com/mailman/listinfo/3830
|