Matching the digital mode to the path: decode thresholds, time sync, CAT and audio setup
- Replies 2
- Views 55
- Created
- Last Reply
Top Posters In This Topic
-
Admin_Terry 1 post
-
Mod_Dan 1 post
A better way to browse. Learn more.
A full-screen app on your home screen with push notifications, badges and more.
Most problems with digital modes come from three things: a mode that doesn't suit the path, a PC clock that has drifted, or an audio chain that is overdriven. This thread is for comparing notes on all three.
Thresholds. Published decode thresholds are quoted as SNR in a 2500 Hz reference noise bandwidth, not in the bandwidth the signal actually occupies. That is why a signal 20 dB below the noise on the meter can still decode. Approximate figures: FT8 is about -21 dB, FT4 about -17 dB, and WSPR about -31 dB. WSPR is so sensitive because a 2-minute transmission is spread over roughly 6 Hz. Q65 has several submodes and period lengths, from 15 seconds to 5 minutes. The longer periods are meant for marginal paths such as EME, ionoscatter and weak tropo. MSK144 is designed for meteor scatter, where the useful signal is a burst of a fraction of a second to a couple of seconds, so a slow mode is no use there. Keyboard modes need much more signal. PSK31 and RTTY are readable on a modest path, but they don't approach FT8 at the noise floor. Olivia and other MFSK modes trade speed for robustness, which is why they are used for text on poor HF paths.
Time. FT8, FT4, WSPR, Q65 and MSK144 all run on fixed UTC T/R periods. If the PC clock is off by more than about a second, decodes drop sharply. Watch the DT column in WSJT-X. If most stations show a consistent offset, your clock is probably wrong, not theirs. Use NTP or a GPS source, and check that the sync service really is running and not just installed.
Audio and control. Set the rig to a data mode (USB-D, DATA-U or similar), turn off speech processing and any DSP noise reduction, and keep the receive filter wide enough for the whole passband. On transmit, set the power with the WSJT-X output slider and the rig's power control. The ALC should show little or no movement. A clean 50 W is better than an overdriven 100 W, because splatter spoils the decode for everyone nearby. CAT through Hamlib or the rig's native interface handles frequency and PTT. If it is flaky, check baud rate, stop bits and whether a second program is holding the COM port. Sound card interfaces range from the rig's built-in USB codec to a separate isolated interface. Ground loops and RF on the cables show up as hum, raised waterfall noise or random PTT drops, and ferrites plus a separate power feed are the usual cure.
Grid squares. Maidenhead locators like FN31 are two letters (field), then two digits (square), and in six-character form two more letters (subsquare). A four-character square is 2° of longitude by 1° of latitude, which is why FT8 exchanges are usually just four characters. Many award and VHF contest rules count in grids, so it helps to know your own six-character locator.
Which modes do you reach for on which bands, and what has given you the most trouble: time, audio levels or CAT?
Link to comment
https://www.hamradiobase.com/forums/topic/11070-matching-the-digital-mode-to-the-path-decode-thresholds-time-sync-cat-and-audio-setup/Share on other sites