The third public test of FT8 DXpedition Mode on May 5 is being called a success. The goal of the exercise was to simulate a rare DXpedition pileup on FT8 by having many stations (“Hounds”) calling and trying to work a designated pseudo DXpedition station (“Fox”).
“A number participants and would-be participants reported that propagation was spotty, at best,” said Joe Taylor, K1JT, on behalf of the WSJT development group, which is sponsoring the tests. “Nevertheless, at AA7A, G4WJS, K1JT, and K9AN, we copied 405 unique call signs of stations acting as Hounds, and at least one Fox was worked by 305 of them. Many Hounds worked two or three of the Foxes. We’re sorry if we missed you, or if propagation was unfavorable for you.”
Taylor said a shift from the announced operating frequency to 14.115 MHz was necessary before the test got under way to avoid RTTY contest activity. “Soon after 1400 UTC, they also learned that W1/KH7Z had some unexpected setup problems and was unable to continue,” he added. As a result, consequently W7/KH7Z took over as the Fox at 1425 UTC and continued operating until 1600. K1JT was the Fox between 1600 and 1700 UTC, and G4WJS stepped in for a short bonus run between 1710 and 1750 UTC, Taylor recounted.
In simple terms, FT8 DXpedition Mode permits a DXpedition station (Fox) to work several stations at a time, utilizing different “slots” for each contact. The downside is that the greater the number of parallel slots, the less power for each slot. Taylor said the penalty is 14 dB for five slots.
Taylor said the setup at W7/KH7Z is comparable to what will be used during the KH1/KH7Z Baker Island DXpedition in June. “We operated in ways expected to provide the most thorough tests of the DXpedition Mode software,” Taylor said. “The low-power, simple-antenna setup at K1JT may be indicative of what can be done from KH1/KH7Z when band conditions are poor.”
Call Sign Grid PEP Per Slot Antenna
W7/KH7Z DM43 500 W 20 W three elements
K1JT FN20 100 W 4 – 9 W dipole
G4WJS IO91 400 W 16 W four elements
The WSJT development group plotted (see chart) the equivalent hourly contact rate for each 5-minute interval at W7/KH7Z. The measured rates range from 8 to 33 per 5-minute interval, or between 96 and 396 contacts per hour. “Several factors acted to suppress the rate at various times,” Taylor said. “The dip at t = 30 minutes was caused by a restart following a program crash. We have identified the reason, and this bug will be easily fixed.”
Taylor said that W7/KH7Z started to run out of available Hounds after about 1 hour of operation, and by the 90-minute mark, the operator had worked 83% of those copied at least once. “This test run has shown that peak QSO rates with FT8 DXpedition Mode can approach 400 per hour, and that sustained rates well above 200 per hour will be readily achieved in good conditions,” Taylor concluded.
He said the test also helped the team to identify “a few relatively minor software bugs” that need to be fixed. For example, one bug prevented Hounds using compound call signs such as S5/N1YU, VE3GFW/W6, or W1/V47KA from working the Fox. “Unknowingly, the Foxes wasted significant time trying to work these stations, thereby suppressing their QSO rates,” Taylor said.
Problems aside, Taylor said the development team considers the third test run “very successful,” and that the team hopes to release the general availability WSJT-XVersion 1.9.0 before the end of May.
“Hearty thanks to everyone who participated in the third public test of FT8 DXpedition Mode!”