Skip to content

CW decoder and sending

SP5XTF edited this page Nov 3, 2026 · 1 revision

CW – decoder and sending

Wersja polska

CW tab: a telegraphy decoder for the radio's audio and sending text with a computer-driven key.

Band buttons (tuning panel, 160–6 m): the radio in CW-U at the start of the CW segment (e.g. 14.030, 7.025, 3.530 MHz). The program remembers the last CW frequency on each band – after retuning and changing band the button returns to the same spot; right-click – back to the default. Changing band stops transmitting.

Decoder

  • Needs monitoring on (audio from the radio). The CW tone is found automatically (~1 Hz accuracy); clicking the spectrum picks a station manually, Auto returns to automatic.
  • Speed (WPM) and threshold adapt by themselves – also to an uneven fist and fading (QSB).
  • The decoder prints only when there is a narrow CW signal on the tone – with SSB voice or plain noise it stays silent (no “garbage”). When CW appears, the decoder replays the last 3 s, so the start of a transmission is not lost.
  • A disabled decoder does not analyse audio at all; the spectrum in the tab is computed only while the tab is open.
  • APF for the decoder (100 Hz) – a narrow filter on the character-reading path only (station search still sees the whole band). It helps with very weak signals (measured at SNR 0 dB: errors 22.5 → 18.4 %) and changes nothing with strong ones. For the decoder it is always 100 Hz – a narrower filter would turn noise into a “tone” and the decoder would print garbage. APF (headphones) in this tab is the AUDIO tab setting; the filter band is marked on the spectrum.
  • AI – language model: when a character is ambiguous (dot or dash, a gap inside a character or between characters), the decoder weighs several versions and picks the one that best fits typical QSO language (CQ, DE, 5NN, TU, 73…) and your own and logbook call signs. The current word is shown at once (amber), and corrected by the model when it ends.
  • Measured on test recordings: clean signal 18–35 WPM ≈ 0–1 % errors, SNR 3 dB ≈ 3 %, uneven fist 2.8 % with the model (7.8 % without), very weak signal (SNR 0 dB) ≈ 24 %; SSB voice – the decoder stays silent.

Sending – two methods (the “Send via” field):

  • DTR/RTS line (CAT-2 port) – precise timing from the computer; needs PC KEYING = DTR/RTS on the radio and the radio's second port. If the line does not change state, the program stops sending and says why.
  • Radio keyer (CAT) – the text goes into FT-710 keyer memory 5 (50 characters at a time) and the radio's built-in keyer sends it; memory 5, speed and keyer state are restored afterwards. Needs neither a second port nor PC KEYING. Requires the radio keyer (KEYER) on and a keyer type set (not OFF) – the program turns the keyer on itself and, after the first piece of text, checks on the radio that it is really transmitting; if not, it gives the reason.
  • Check radio reads PC KEYING, BK-IN and mode over CAT; Fix settings sets them with one click. With the “Radio keyer” method PC KEYING is shown for information only (ⓘ) – that method does not need it.

Line method details

  • On the radio: CW mode, BK-IN on, PC KEYING = DTR menu (or RTS – matching the choice in the program).
  • The key port is the radio's second port (Standard COM / CAT-2) – the program suggests it (e.g. …C1 for CAT port …C0).
  • Text or macros (pencil – edit; {CALL} = your call), speed 5–40 WPM.
  • Requires Enable TX; the text duration is checked against the transmit limit. STOP, withdrawing TX consent or disconnecting the radio releases the key immediately.

Log (column on the right)

  • Call with a QRZ/HamQTH lookup (name, QTH and locator fill in themselves; for a station away from home only the name), entity and “new station” or “worked: N QSOs”.
  • RST sent and received – empty fields are saved as 599.
  • Log QSO (or ⌘↩) – CW mode, frequency from the radio, saved as in the program's log (entity, zones, distance, upload to services).

← SSTV – images · Home · Propagation →

Clone this wiki locally