Radios, sites and stations

Rigs: one or several radios

A rig is one radio you own and how it is wired up: its model (profile), the serial port and sound card it uses, an optional modem, its bands, and its own ports. Several can run at once, for example an HF rig with Mercury and a VHF/UHF rig.

hamstack rig add hf   --profile ic7300mk2 --modem mercury --bands hf    # detects its USB devices if plugged in
hamstack rig add vhf  --profile ftdx10 --bands vhf
hamstack rigs                          # what you have, and whether each radio is plugged in
hamstack rig identify vhf              # unplug / replug to find (or re-find) its USB devices
hamstack rig disable vhf               # keep it in the config but do not run it
hamstack rig set hf modem none         # also: name, profile, bands, cat, audio
hamstack pat on|off                    # Winlink client, shared by all rigs

setup winlink creates the first rig for you. Systemd units are templates (hamstack-rigctld@hf, hamstack-mercury@hf) so adding a rig never needs new unit files, and they hold no device paths: each runs hamstack run, which finds the radio’s serial port and sound card when the service starts, using the USB identity (vendor, product, serial number) saved with the rig. A radio on another USB port, or moved to another site, keeps working.

Service names are rigctld@hf, mercury@hf and pat. Commands take a service (mercury@hf), a kind (mercury: every one), a rig (hf: everything of that rig) or all.

Each rig gets its own ports: rig control counts up from 4532, and a rig’s Mercury uses a block of ten from 8300 (its web UI, broadcast and message history are separate too). Ports are kept when a rig is removed so the others never move under Pat’s config. Pat has one HF modem setting, so with several Mercury rigs Pat’s is wired to the first; doctor checks that it points at a Mercury that exists. Configs from before rigs existed are migrated automatically.

Choosing a radio

The guided setup (and the Rigs tab) shows a short list first: the dummy radio (testing without hardware), a Digirig, eight common rigs (IC-7300, IC-7300 MK2, IC-705, IC-7610, FT-991A, FT-710, FT-DX10) and FLRig (control through flrig). Anything else is one search away: s, then type part of a name (kx3, ic 7100, yaesu). The list is every radio the installed hamlib supports (it offers to install hamlib first), matched loosely (spaces and dashes ignored, letters in order), paged. A radio without a profile gets one made and saved on the spot, asking the serial speed (it shows the range hamlib says the model supports). Profiles other than the dummy, Digirig, IC-7300 MK2 and FT-DX10 were written from hamlib’s list and typical settings, not tested with the radio: check the CAT rate on the radio matches.

With a radio that is not plugged in, setup finishes and says which services will wait; they show no radio (not a failure) and start as soon as it is plugged in and you run hamstack start. A Digirig with Mercury switches that rig to Direwolf (it has no CAT control to key Mercury with) and says so.

Sites: home, the EOC, the club

A site is a place you operate from. It records which of your rigs are there, which station (callsign) is active, and the grid square, so moving between places is one command instead of a dozen.

hamstack site add home                  # save what is set up now (rigs, station, grid) as "home"
hamstack site add eoc --rigs vhf,dig --station eoc --grid FN31pr
hamstack site use eoc                   # its rigs on, the others off, its station and grid; services restarted
hamstack sites                          # your places, and how many of each one's radios are plugged in now
hamstack site here                      # which site do the plugged-in radios match?  (--switch to go there)
hamstack pat sync                       # wire Pat's config to the rigs and grid you are at now

The current site is a live view: enabling or adding a rig, or switching station, is recorded in it, and config set grid changes the current site’s grid (or the default when you have no sites). site use restarts things only if the stack was running, and stops the rigs that are not part of the new site.

hamstack suggests a site but never switches by itself: doctor and sites say when the radios plugged in belong to a different site. Radios are recognised by the USB identity saved with each rig, so a rig with no saved device (the dummy radio) takes no part in the matching.

Moving to a new site changes the grid and the radios, but Pat’s own config is never rewritten behind your back. doctor warns when Pat’s locator, HF modem or packet TNC no longer match, and hamstack pat sync fixes exactly those settings (the locator, the rig connections, the VARA HF address, the AGWPE address), keeping everything else, passwords included, and leaving a .hamstack-bak copy.

Stations (several callsigns)

A station is an operating identity: a callsign with its own Pat config, mailbox and web UI port, for example a personal call, a club call and an emergency-communications call. Stations are yours to define; none are built in, and they are kept in ~/.config/hamstack/.

hamstack station add club --callsign N0CLUB --name "Club"   # starter Pat config created if missing
hamstack station import-pat                                 # adopt existing ~/.config/pat/*.json as stations
hamstack stations                                           # list
hamstack use club                                           # switch: regenerates config, restarts mercury + pat

Radios (profiles) and stations are independent: any station works with any radio. hamstack never stores your Winlink password; it lives in the station’s Pat config. Only one station is active (and can use the radio) at a time. Pat logs go to ~/.local/state/hamstack/.