Ubuntu Manpages

svxlink.conf

Configuration file for the SvxLink server

svxlink is a general purpose voice service system for ham radio use. This man-page describe the SvxLink server configuration file format.

SvxLink look for configuration files in a number of places. First it try to find a user specific configuration file. SvxLink will look for a user specific configuration file in: $HOME/.svxlink/svxlink.conf. If no user specific configuration file can be found, SvxLink will look for the system wide configuration file /etc/svxlink/svxlink.conf. The --config command line option may also be used to specify an arbitrary configuration file.

The configuration file is in the famous INI-file format. A generic example of how such a file might look like is shown below.


[SECTION1]
VALUE1=1
VALUE2="TWO "
VAULE3="Multi "
"line"


[SECTION2]
VALUE1=2

This is a simple format that contain name=value pairs that belong to a section. In written text, a specific configuration variable can be referred to as SECTION1/VALUE2 meaning "configuration variable VALUE2 in section SECTION1".

The same variable name can exist in two different sections. For example VALUE1 in section SECTION1 have the value 1 and VALUE1 in section SECTION2 have the value 2. Values containing spaces at the beginning or end of the line must be surrounded by citation characters (see SECTION1/VALUE2). Likewise with a multi line value (see SECTION1/VALUE3).

Here is the description of all configuration variables that SvxLink understands. The configuration variables are described section for section.

The GLOBAL section contains application global configuration data.

Specify where the SvxLink modules can be found. The default is /usr/lib/svxlink
Specify a comma separated list of logic cores that should be created. The logic core is the thing that ties the transciever and the voice services (modules) together. It contains the rules for how the radio interface should be handled. The specified name of a logic core must have a corresponding section specified in the config file. This is where the behaviour of the logic core is specified.
Specify the path to a directory that contain additional configuration files. If a relative path is specified, the path will be relative to the directory where the main configuration file is at. All files in the specified directory will be read as additional configuration. Filenames starting with a dot are ignored.
This variable specifies the format of the timestamp that is written in front of each row in the log file. The format string is in the same format as specified in the strftime(3) manual page. The default is "%c" which is described as: "the preferred date and time representation for the current locale". The environment variables LC_TIME, LC_ALL and LANG will affect how this time format will look. For example, setting LC_TIME="sv_SE.UTF8" will give you swedish timestamp representation. Other examples of format specifiers are:
  • %d - The day of the month as a decimal number (range 01 to 31)
  • %b - The abbreviated month name according to the current locale
  • %Y - The year as a decimal number including the century
  • %H - The hour as a decimal number using a 24-hour clock (range 00 to 23)
  • %M - The minute as a decimal number (range 00 to 59)
  • %S - The second as a decimal number (range 00 to 61)
  • %f - Fractional seconds in millisecond resolution (000-999)

The last one (%f) is a SvxLink specific formatting specifier.

Example: TIMESTAMP_FORMAT="%d %b %Y %H:%M:%S.%f" would give a timestamp looking something like: "29 Nov 2005 22:31:59.875".

This configuration variable determines the sampling rate used for audio input/output. SvxLink always work with a sampling rate of 8kHz internally but there still are som benefits from using a higher sampling rate. On some sound cards the filters look pretty bad at 8kHz and the amplitude response will not be uniform which among other things can cause problems for the software DTMF decoder.

Some sound cards also sound very bad at 8kHz due to insufficient anti-alias filtering or resampling effects. These, often cheeper, sound cards sound OK at 48kHz.

The downside of choosing a higher sampling rate is that it puts a little bit more load on the CPU so if you have a very slow machine (<300MHz), it might not have the computational power to handle it.

Supported sampling rates are: 8000, 16000 and 48000.

Enter the section name that contains information required for transferring positioning data to location servers. Setting this item makes the system visible on the EchoLink link status page and the APRS network.
Enter here a comma separated list of section names that contains the configuration information for linking logics together (see Logic Linking).

A logic core is what define how SvxLink should behave on the RF channel. The SvxLink server can handle more than one logic core and so can be connected to more than one transceiver. The configuration variables below are common to all logic types. Configuration variables that are specific to a certain logic core type are described below in a section of its own.

The type of logic core this is. The documentation for the specific logic core type you want to use describe what to write here.
Specify the configuration section name of the receiver to use. All configuration for the receiver is done in the specified configuration section.
Specify the configuration section name of the transmitter to use. All configuration for the transmitter is done in the specified configuration section.
Specify a comma separated list of configuration sections for the modules to load. This tells SvxLink which modules to actually load on startup.
Specify the callsign that should be announced on the radio interface.
The number of minutes between short identifications. The purpose of the short identification is to just announce that the station is on the air. Typically just the callsign is transmitted. For a repeater a good value is ten minutes and for a simplex node one time every 60 minutes is probably enough. The LONG_IDENT_INTERVAL must be an even multiple of the SHORT_IDENT_INTERVAL so if LONG_IDENT_INTERVAL is 60 then the legal values for SHORT_IDENT_INTERVAL are: 1, 2, 3, 4, 5, 6, 10, 12, 15, 20, 30, 60. If unset or set to 0, disable short identifications.
The number of minutes between long identifications. The purpose of the long identification is to transmit some more information about the station status (new voice mails etc). The time of day is also transmitted. A good value here is 60 minutes. If unset or set to 0, disable long identifications.
This feature controls when identification is done. By default, identification is done every time the SHORT_IDENT_INTERVAL expires. If this feature is enabled, identification will be done only if there has been a recent transmission. This feature is good for nodes using an RF link to provide echolink to a repeater. Often, in this situation, it is not desirable for the link to identify unless legally necessary. Note that SHORT_IDENT_INTERVAL still have to be set for this feature to work. That config variable will then be interpreted as the minimum number of seconds between identifications. The LONG_IDENT_INTERVAL will not be affected by this parameter.
Specify a time, in milliseconds, after squelch close after which entered DTMF digits will be executed as a command without the need to send the # character. This really only is of use when using a radio that it is difficult to send DTMF digits from, like the Yaesu VX-2 handheld. The down side of enabling this option is that the DTMF detection some times false trigger on voice. This can cause interresting situations when all of a sudden a module get activated in the middle of a QSO.
Point out the TCL event handler script to use. The TCL event handler script is responsible for playing the correct audio clips when an event occurr. The default location is /usr/share/svxlink/events.tcl.
Set the default language to use for announcements. It should be set to an ISO code (e.g. sv_SE for Swedish). If not set, it defaults to en_US which is US English.
The number of milliseconds to wait after the squelch has been closed before a roger beep is played. The beep can be disabled by specifying a value of -1 or commenting out this line. Often it is best to use the SQL_HANGTIME receiver configuration variable to specify a delay instead of specifying a delay here. This configuration variable should then be set to 0.
If set, will report the specified CTCSS frequency upon manual identification (* pressed). It is possible to specify fractions using "." as decimal comma. Disable this feature by commenting out (#) this configuration variable.
This configuration variable controls if a CTCSS tone should be transmitted. Use a comma separated list (no spaces!) to specify when to transmit a CTCSS tone. These are the possible values: SQL_OPEN, LOGIC, MODULE, ANNOUNCEMENT or ALWAYS. Commenting out this configuration variable will disable CTCSS transmit. The tone frequency and level is configured in the transmitter configuration section.
  • SQL_OPEN will transmit CTCSS tone when the squelch is open. This is only useful on a repeater. On a simplex node it doesn't make much sense.
  • LOGIC will transmit CTCSS tone when there is incoming traffic from another logic core.
  • MODULE will transmit CTCSS tone when there is incoming traffic from a module.
  • ANNOUNCEMENT will transmit CTCSS tone when an announcement is being played. Repeater idle sounds and roger beeps will not have tone sent with them though.
  • ALWAYS will always transmit a CTCSS tone as soon as the transmitter is turned on.
Point out a section that contains the macros that should be used by this logic core. See the section description for macros below for more information.
The gain (dB) to use for audio effects and announcements when there is no other traffic. This gain is normally set to 0dB which means no gain or attenuation.
The gain (dB) to use for audio effects and announcements when there is other traffic. This gain is normally set to something like -12dB so that announcements and audio effects are attenuated when there is other traffic present.
The QSO recorder is used to write all received audio to files on disk. The format for this configuration variable is <command>:<config section>. The specified command is used to activate or deactivate the QSO recorder. If the command for example is set to 8, 81 will activate the recorder and 80 will deactivate it. The command may also be left out. It will then not be possible to control the QSO recorder using DTMF commands. Even if the command is left out the colon must always be specified. The config section point out a section in the configuration file that holds configuration for the QSO recorder. Have a look at the QSO Recorder Section documentation for more information.

Example: QSO_RECORDER=8:QsoRecorder

Define two comma separated values here to map the Sel5 tone call to your macro area. E.g. if you have defined: SEL5_MACRO_RANGE=03400,03499 then all incoming Sel5 tone sequences from 03400 to 03499 are mapped to the macros section (refer to Macros Section, next chapter). Other sequences but the one defined under OPEN_ON_SEL5 are ignored so it can be used to call other stations via the repeater without a repeater reaction.
Define a DTMF command that is used to switch the node between online and offline mode. When in the off-state, the transmitter will not be turned on by any event. If a module is active when the node is brought offline, it will be deactivated and no module activation will be allowed in offline mode. No other commands than the online command will be accepted in the offline state.

If the command for example is set to 998877 then 9988771 will set the node online and 9988770 will set it offline. If a module is active or if the ACTIVATE_MODULE_ON_LONG_CMD is used, the command must be prefixed with a star to work as expected. The star means "force core command".

The Simplex Logic section contains configuration data for a simplex logic core. The name of the section, which in the example configuration file is SimplexLogic, must have a corresponding list item in the GLOBAL/LOGICS config variable for this logic core to be activated. The name "SimplexLogic" is not magic. It could be called what ever you like but it must match the namespace name in the SimplexLogic.tcl script. The configuration variables below are those that are specific for a simplex logic core.

The type for a simplex logic core is always Simplex.
Set to 1 to mute the receiver when the transmitter is transmitting (default) or set it to 0 to make the RX active during transmissions. One might want to set this to 0 if the link is operating on a split frequency. Then the link can accept commands even when it's transmitting. The normal setting is 1, to mute the RX when transmitting.
Set to 1 to mute the transmitter when the squelch is open (default) or set it to 0 to make the TX active during squelch open. One might want to set this to 0 if the link is operating on a split frequency or if it's connected to some full duplex device. The normal setting is 1, to mute the TX when the squelch is open.

A Repeater Logic section contains configuration data for a repeater logic core. The name of the section, which in the example configuration file is RepeaterLogic, must have a corresponding list item in the GLOBAL/LOGICS config variable for this logic core to be activated. The name "RepeaterLogic" is not magic. It could be called what ever you like but it must match the namespace name in the RepeaterLogic.tcl script. The configuration variables below are those that are specific for a repeater logic core.

The type for a repeater logic core is always Repeater.
Set this to 1 if you do NOT want SvxLink to play back the incoming audio. This can be used when the received audio is directly coupled by hardware wiring to the transmitter. What you win by doing this is that there is zero delay on the repeated audio. When the audio is routed through SvxLink there is always an amount of delay. What you loose by doing this is the audio processing done by SvxLink (e.g. filtering, DTMF muting, squelch tail elimination) and the ability to use remote receivers.
The number of seconds the repeater should have been idle before turning the transmitter off.
Use this configuration variable if it should be possible to open the repeater with a 1750Hz tone burst. Specify the number of milliseconds the tone must be asserted before the repeater is opened. Make sure that the time specified is long enough for the squelch to have time to open. Otherwise the repeater will open "too soon" and you will hear an ugly 1750Hz beep as the first thing. A value of 0 will disable 1750 Hz repeater opening.
Use this configuration variable if it should be possible to open the repeater with a CTCSS tone (PL). The syntax of the value is tone_fq:min_length. The tone frequency is specified in whole Hz and the minimum tone length is specified in milliseconds. For examples if a 136.5 Hz tone must be asserted for two seconds for the repeater to open, the value 136:2000 should be specified.
Use this configuration variable if it should be possible to open the repeater with a DTMF digit. Only one digit can be specified. DTMF digits pressed when the repeater is down will be ignored.
Use this configuration variable if you want to open your repeater by using a selective tone call that is often used in commercial radio networks. Example: OPEN_ON_SEL5=03345 opens your repeater only if that sequence has been received. You can use sequence lengths from 4 to 25.
Use this configuration variable if you want to close your repeater by using a selective tone call that is often used in commercial radio networks. Example: CLOSE_ON_SEL5=03345 closes your repeater if that sequence has been received. You can use sequence lengths from 4 to 25.
Use this configuration variable if it should be possible to open the repeater just by keeping the squelch open for a while. The value to set is the minimum number of milliseconds the squelch must be open for the repeater to open.
Activate the repeater on just a squelch opening if there have been no more than the specified number of seconds since the repeater closed.
Determines if OPEN_ON_SQL and OPEN_ON_CTCSS should activate the repeater when the squelch open or close. If set to OPEN, the repeater will activate and start retransmitting audio immediately. No identification will be sent. If set to CLOSE, the repeater will not activate until the squelch close. An identification will be sent in this case.
When the repeater is idle, a sound is played. Specify the interval in milliseconds between playing the idle sound. An interval of 0 disables the idle sound.
Flapping squelch suppression is used to close the repeater down if there is interference on the frequency that open the squelch by short bursts. This configuration variable is used to specify the minimum time, in milliseconds, that a transmission must last to be classified as a real transmission. A good value is in between 500-2000ms.
Flapping squelch suppression is used to close the repeater down if there is interference on the frequency that open the squelch by short bursts. This configuration variable is used to specify the maximum number of consecutive short squelch openings allowed before shutting the repeater down. A good value is in between 5-10.
This configuration variable activate a feature that might help users not aware of the SvxLink command structure. The idea is to activate the specified module when a long enough command has been received. The typical example is an EchoLink user that is used to just typing in the node ID and then the connection should be established right away. Using this configuration variable, specify a minimum length and a module name. If no module is active and at least the specified number of digits has been entered, the given module is activated and the command is sent to it. To be really useful this feature should be used in cooperation with EXEC_CMD_ON_SQL_CLOSE.

For example, if this configuration variable is set to "4:EchoLink" and the user types in 9999, the EchoLink module is first activated and then the command 9999 is sent to it, which will connect to the ECHOTEST server.

Tell repeater users that are not identifying to identify themselvs. The number of seconds to wait for an identification, after the repeater has been activated, is set using this configuration variable. A valid identification is considered to be a transmission longer than the time set by the IDENT_NAG_MIN_TIME configuration variable. We don't know if it's really an identification but it's the best we can do. Setting it to 0 or commenting it out disables the feature.
This is the minimum time, in milliseconds, that a transmission must last to be considered as an identification. This is used as described in the IDENT_NAG_TIMEOUT configuration variable.

The QSO recorder is used to record all received audio to files on disk. All audio from receivers, modules and logic links are recorded. Announcements are not recorded.

Use this configuration variable to specify in which directory to write the audio files. A good place is /var/spool/svxlink/qso_recorder.
If the duration of the recorded content for a file is less then MIN_TIME milliseconds, the file will be deleted when the file is closed. Default: 0 (empty files will be deleted).
Setting this configuration variable will set an upper limit for the file size of a recording. No more than MAX_TIME seconds of content will be recorded to a single file. When the maximum time have been reached, the file is closed and another file is created. Note that it is not the maximum time that the recording has been active that we are setting a limit for but rather how much content that have been recorded to the file. If nothing is recorded, the file can stay open indefinitely. Default: 0 (no limit)
To not get abrupt breaks in recordings it is possible to set a soft break time. Let's say that MAX_TIME is set to 3600 seconds (one hour). If we set SOFT_TIME to 300 seconds (five minutes) the QSO recorder try to close the file on a squelch close somewhere between 55 and 60 minutes. In this way we may avoid getting transmissions split up between files. Default: 0 (no limit)
Specify the maximum total size in megabytes of the files in the recording directory. If the limit is exceeded, the oldest files are deleted. The directory size is checked upon file close so the size may grow temporarily past the limit with at most the size of one recorded file. Only files which have a filename starting with "qsorec_" will be considered for deletion. If using an ENCODING_CMD, make sure that the "qsorec_" prefix is not removed from the target filename unless you really want the MAX_DIRSIZE feature to skip them. Default: 0 (no limit)
If this configuration variable is set to 1, the QSO recorder will be activated by default when SvxLink start. Default: 0 (default inactive)
If a timeout is specified, the activation state of the QSO recorder will return to the value specified in the DEFAULT_ACTIVE configuration variable when the node has been idle for the specified number of seconds. When DEFAULT_ACTIVE is unset or 0, if the QSO recorder is manually activated it will be automatically deactivated after the specified amount of time of inactivity. When DEFAULT_ACTIVE is set to 1, if the QSO recorder is manually deactivated it will be automatically activated after the specified amount of time of inactivity. Default: 0 (no timeout)
Set this configuration variable if you want to close the currently opened file and open a new one after each QSO. The number of seconds the node should be idle before closing the file should be specified. Default: 0 (no QSO timeout)
Specify a command to be executed after a new wav file have been written to disk. This makes it possible to use an external encoder utility to encode the wav file to another format. Even though this configuration variable was added to run an external encoder it could do more complicated things with the file if needed. A couple of examples would be to transfer the file to another computer or to send a notification e-mail. If the command line get too complicated it may be a good idea to write a script instead.

The encoder command will be run under a shell so normal shell operators like redirects and pipes may be used. The shell specified in the SHELL environment variable will be used and if not set, /bin/sh will be used. The "-c" command line option will be added so the complete command will look something like: $SHELL -c "$ENCODER_CMD". A number of %-codes can be included in the command. They have the following meaning:

  • %f - The full filename with full path
  • %d - The directory part (what REC_DIR is set to)
  • %b - The basename, that is, the filename without path and extension
  • %n - The filename without path but with extension

The encoder will be started in the background and it will not be stopped even if SvxLink exits. It will run in the background until it's done. As long as SvxLink is running it is monitoring the encoding processes. If a process run for longer than one hour it will be killed.

Note that SvxLink will never remove the original recording so that have to be done in the encoder command. Here are a couple of examples:


ENCODER_CMD=/usr/bin/oggenc -Q \"%f\" && rm \"%f\"
ENCODER_CMD=/usr/bin/lame --quiet \"%f\" \"%d/%b.mp3\" && rm \"%f\"
ENCODER_CMD=/usr/bin/speexenc \"%f\" \"%d/%b.spx\" 2>/dev/null && rm \"%f\"
ENCODER_CMD=/usr/bin/opusenc \"%f\" \"%d/%b.opus\" 2>/dev/null && rm \"%f\"

A macros section is used to declare macros that can be used by a logic core. The logic core points out the macros section to use by using the MACROS configuration variable. The name of the MACROS section can be chosen arbitrarily as long as it match the MACROS configuration variable in the logic core configuration section. There could for example exist both a [RepeaterLogicMacros] and a [SimplexLogicMacros] section.

A macro is a kind of shortcut that can be used to decrease the amount of key presses that have to be done to connect to common EchoLink stations for example. On the radio side, macros are activated by pressing "D" "macro number" "#". A macros section can look something like the example below. Note that the module name is case sensitive.


[Macros]
1=EchoLink:9999#
2=EchoLink:1234567#
9=Parrot:0123456789#

For example, pressing DTMF sequence "D1#" will activate the EchoLink module and connect to the EchoTest conference node.

A logic linking configuration section is used to specify information for a link between two or more SvxLink logic cores. Such a link can for example be used to connect a local repeater to a remote repeater using a separate link transceiver. The link is activated/deactivated using DTMF commands and/or automatically depending on your configuration. When the link is active, all audio received by one logic will be transmitted by the other logic(s).

The name of the logic linking section can be chosen freely. In the example configuration file, there is a section [LinkToR4]. To use a logic linking section in a logic core it must be pointed out by the LINKS configuration variable in the GLOBAL section. Example: GLOBAL/LINKS=LinkToR4

A comma separated list of logic specifications for the logic cores to connect together. Each logic specification has three parts separated by colons: <logic name>:<command>:<announcement name>. The "logic name" is the name of the logic to include in the link. To manually activate or deactivate the link from the just specified logic, "command" is used. The "announcement name" is used when announcing link related activities like activation or deactivation. Both "command" and "announcement name" may be left empty if no manual control is wanted. An example config line may look like this:

RepeaterLogic_2m:99:SK3GW,RepeaterLogic_70cm:94:SK3GK

It will include two logics in the link, RepeaterLogic_2m and RepeaterLogic_70cm. From the 2m side, the link will be activated when the user send command 991 and deactivated when the user send command 990. Upon activation, an announcement like "activating link to SK3GW" will be played back. From the 70cm side the command will be 941 and 940 respectively. The announcement when activating the link from the 70cm side will be something like "activating link to SK3GK".

The link will be connected automatically during startup of SvxLink if this configuration variable is set to 1. Also, if a link is manually disconnected by a user it will be automatically reconnected after some time of inactivity. The time is specified by setting the TIMEOUT configuration variable. If the TIMEOUT variable is not set, no automatic reactivation will be done.
The number of seconds after which the link will be automatically deactivated if there have been no activity. If 1 have been specified for DEFAULT_ACTIVE, this configuration variable will specify after how many seconds the link will be reactivated after being manually deactivated.
Enter a comma separated list of logics, which should automatically activate the link if there is activity (e.g. squelch open) in it. One possible application for this is for example to make the connection of a microphone/speaker combination (without DTMF encoder) for brief announcements but without having to constantly listen in. Example: AUTOACTIVATE_ON_SQL=MicSpkrLogic

A local receiver section is used to specify the configuration for a receiver connected to the sound card. In the default configuration file there is a Local configuration section called Rx1. The section name could be anything. It should match the RX configuration variable in the logic core where the receiver is to be used. The available configuration variables are described below.

Always "Local" for a local receiver.
Specify the audio device to use. Normally alsa:plughw:0. Have a look at the AUDIO DEVICE SPECIFICATIONS chapter for more information.
Specify the audio channel to use. SvxLink can use the left/right stereo channels as two mono channels. Legal values are 0 or 1.
Specify the type of squelch detector to use. Possible values are: VOX, CTCSS, SERIAL, EVDEV, SIGLEV, PTY or GPIO.

The VOX squelch detector determines if there is a signal present by calculating a mean value of the sound samples. The VOX squelch detector behaviour is adjusted with VOX_FILTER_DEPTH and VOX_THRESH. VOX is actually a bit of a misnomer since it's a "Voice Operated Squelch" and VOX actually means "Voice Operated Transmitter". However, the term VOX is widely understood by hams all over the world so we'll stick with it.

The CTCSS squelch detector checks for the presence of a tone with the specified frequency. The tone frequency is specified using the CTCSS_FQ config variable. The thresholds are specified using the CTCSS_OPEN_THRESH and CTCSS_CLOSE_THRESH config variables. Other config variables that effect the CTCSS squelch is: CTCSS_MODE, CTCSS_SNR_OFFSET, CTCSS_BPF_LOW, CTCSS_BPF_HIGH.

The SERIAL squelch detector use a pin in a serial port to detect if the squelch is open. This squelch detector can be used if the receiver have an external hardware indicator of when the squelch is open. Specify which serial port/pin to use with SERIAL_PORT and SERIAL_PIN.

The EVDEV squelch detector read squelch events from a /dev/input/eventX device. An example where this could be useful is if you have a USB audio device with some buttons on it. Some of these devices generate key press events, much like a keyboard. Specify which /dev/input device node to use using the EVDEV_DEVNAME config variable. Set which events that should open and close the squelch using the EVDEV_OPEN and EVDEV_CLOSE config variables.

The GPIO squelch detector read a pin on the GPIO Port. Depending on the level of the pin, the squelch is switched. A HIGH (3.3V) at the pin set the squelch to open and a LOW (GND) level will set the squelch to closed. Specify which squelch pin to use with the GPIO_SQL_PIN configuration variable.

The SIGLEV squelch detector use signal level measurements to determine if the squelch is open or not. Which signal level detector to use is determined by the setting of the SIGLEV_DET configuration variable. The open and close thresholds are set using the SIGLEV_OPEN_THRESH and SIGLEV_CLOSE_THRESH configuration variables. If using the NOISE signal level detector note the following. The detector is not perfect (it's affected by speech) so you will also want to setup SQL_HANGTIME to prevent it from closing in the middle of a transmission. A value between 100-300ms is probably what you need. If using this squelch type in cooperation with a voter, you'll also probably need to setup SQL_DELAY to get correct signal level measurements. A value of about 40ms seem to be OK. Also, when using the NOISE signal level detector the input audio must be unsquelched since silence will be interpreted as a high signal strength.

The PTY squelch expects a very easy protocol over a pseudo-tty device, created by SvxLink on runtime. An 'O' over this pty device indicate an open squelch, a 'Z' is a closed squelch. Define the slave pty in PTY_PATH (e.g. PTY_PATH=/tmp/sql) and SvxLink will create a link to the specified path from it's pseudotty slave device (/dev/pts/X). This can be used by a script to interface custom devices, modems or other hardware to SvxLink. Look for nhrcx.pl to see an example.

The squelch start delay is of most use when using VOX squelch. For example, if the transceiver makes a noise when the transmitter is turned off, that might trigger the VOX and cause an infinite loop of squelch open/close transmitter on/off. Specify the number of milliseconds that the squelch should be "deaf" after the transmitter has been turned off.
Specify a delay in milliseconds that a squelch open indication will be delayed. This odd feature can be of use when using a fast squelch detector in combination with the signal level detector. A squelch delay will allow the signal level detector to do its work before an indication of squelch open is sent to the logic core. A delay might be needed when using the voter to choose among multiple receivers. A normal value could be somewhere in between 20-100ms.
How long, in milliseconds, the squelch will stay open after the detector has indicated that it is closed. This configuration variable will affect all squelch detector types.
At low signal strengths it can be beneficial to use a longer squelch hangtime so that it is less likely for the squelch to close. This configuration variable is unset by default. A value of 1000 milliseconds may be a good value to start out with. To enable the extended squelch hangtime feature, set up the SQL_EXTENDED_HANGTIME_THRESH variable.
At low signal strengths it can be beneficial to use a longer squelch hangtime so that it is less likely for the squelch to close. This configuration variable is unset by default. When set to a signal level it will activate the extended squelch hangtime feature. When the signal strength during a transmission fall below the set threshold, the extended hangtime will be used. Start out with a value between 10 to 15. The SQL_EXTENDED_HANGTIME variable is used to set how long the extended squelch hangtime should be. Make sure that you have calibrated the signal level detector before turning this feature on. Otherwise it will not work as expected.
Use this configuration variable to set an upper limiti, in seconds, for how long the squelch is allowed to be open. If the timeout value is exceeded the squelch is forced to closed. If the squelch close for real, everthing is back to normal. When it opens the next time a squelch open will be signalled. For example, use this feature to make sure that a faulty receiver cannot block the system indefinitly.
The number of milliseconds to create the mean value over. A small value will make the vox react quicker (<200) and larger values will make it a little bit more sluggish. A small value is often better.
The threshold that the mean value of the samples must exceed for the squlech to be considered open. It's hard to say what is a good value. Something around 1000 is probably a good value. Set it as low as possible without getting the vox to false trigger.
This configuration variable set the CTCSS detection method used. These are the ones to choose from:
  • 0 (Default) Will choose the detection mode that is the default in the software. At the moment this is the "Estimated SNR" mode.
  • 1 (Neighbour bins) This detection mode will use three narrow frequency bands (~8Hz) to do the detection. One band is centered around the tone to be detected and then there are one band above and one below the tone. These bands are used to estimate the noise floor. This is the detector that have been used in SvxLink for a long time. It is however rather slow with its detection time of about 450ms. There is no good reason to use this detector anymore but it is kept in case the new detector does not work for some hardware setup.
  • 2 (Estimated SNR) This is a newer detector implementation which have some improvements. The most notable difference is that it is faster. The mean detection time will be around 200ms. This is the default detection mode if not specified. This detector will use a larger passband to estimate the noise floor which make it more stable. The default config use the whole CTCSS passband but this can be customized using the CTCSS_BPF_LOW and CTCSS_BPF_HIGH config variables.
  • 3 (Estimated SNR+Phase) This detector is a bit experimental. It is even faster and more narrow than the other detection modes. The mean detection time will be something like 150ms. The detection bandwidth is very narrow and very sharp so that no adjacent tones will trigger the detector. The price to pay for these improvements is that is it a bit less sensitive.
If CTCSS (PL,subtone) squelch is used (SQL_DET is set to CTCSS), this config variable sets the frequency of the tone to use. The tone frequency ranges from 67.0 to 254.1 Hz. There actually is nothing that will stop you from setting the frequency to something outside this range but there is no guarantee that it will work.
If CTCSS (PL, subtone) squelch is used (SQL_DET is set to CTCSS), this config variable sets the required tone level to indicate squelch open. The value is some kind of estimated signal to noise dB value. If using CTCSS mode 2 or 3 it is helpful to set up the CTCSS_SNR_OFFSET config variable. This will make the SNR estimation pretty good. Default threshold is 15dB.
If CTCSS (PL, subtone) squelch is used (SQL_DET is set to CTCSS), this config variable sets the required tone level to indicate squelch close. The value is some kind of estimated signal to noise dB value. If using CTCSS mode 2 or 3 it is helpful to set up the CTCSS_SNR_OFFSET config variable. This will make the SNR estimation pretty good. Default threshold is 9dB.
This config variable is used when CTCSS_MODE is set to 0, 2 or 3. It will adjust the estimated SNR value so that it becomes very close to a real SNR value. This value will have to be adjusted if CTCSS_FQ, CTCSS_MODE, CTCSS_BPF_LOW or CTCSS_BPF_HIGH changes. Use the siglevdetcal utility to find out what to set this config variable to. There is no requirement to set this config variable up. The downside is that you will then need to experiment more with the CTCSS_OPEN_THRESH and CTCSS_CLOSE_THRESH config variables to find the correct squelch level.
When CTCSS_MODE is set to 0, 2 or 3, this config variable will set the low cutoff frequency for the passband filter. It normally should not have to be adjusted but could improve the detector if some interference falls within the passband (e.g. mains hum). Note however that the more narrow you make the passband, the less stable the detector will be. You may need to compensate by increasing the open/close thresholds or by setting up SQL_DELAY and SQL_HANGTIME. Default is 60Hz.
When CTCSS_MODE is set to 0, 2 or 3, this config variable will set the high cutoff frequency for the passband filter. It normally should not have to be adjusted but could improve the detector if some interference falls within the passband. Note however that the more narrow you make the passband, the less stable the detector will be. You may need to compensate by increasing the open/close thresholds or by setting up SQL_DELAY and SQL_HANGTIME. Default is 270Hz.
If SQL_DET is set to SERIAL, this config variable determines which serial port should be used for hardware squelch input (COS - Carrier Operated Squelch). Note: If the same serial port is used for the PTT, make sure you specify exactly the same device name. Otherwise the RX and TX will not be able to share the port. Example: SQL_PORT=/dev/ttyS0
If SQL_DET is set to SERIAL, this config variable determines which pin in the serial port that should be used for hardware squelch input (COS - Carrier Operated Squelch). It is possible to use the DCD, CTS, DSR or RI pin. If inverted operation is desired, prefix the pin name with an exclamation mark (!).

Example: SQL_PIN=!CTS

Set the specified serial port pins to a static state. This can be good if using a pin for reference voltage or if a pin have to be in a certain state to not interfere with the operation of some equipment. There are two pins that are possible to use, RTS and DTR. If prefixed with an exclamation mark (!), the pin will be cleared and if not it will be set.

Example: SERIAL_SET_PINS=RTS!DTR will set RTS and clear DTR.

Specify which /dev/input device node to use for the EVDEV squelch detector. To find out which device node and event codes to use, install the evtest utility. Find a candidate device node under /dev/input/ or /dev/input/by-id/ and try the evtest utility on it. Press some keys on the device you want to read events from. If you're in luck, events will be printed on the screen.
Use the evtest utility, as described above, to find out type, code and value for the event you want to use to open the squelch. For example if type is 1, code is 163 and value is 1, set this config variable to 1,163,1.
Use the evtest utility, as described above, to find out type, code and value for the event you want to use to close the squelch. For example if type is 1, code is 163 and value is 0, set this config variable to 1,163,0. If you set the same type,code,value combination for both EVDEV_OPEN and EVDEV_CLOSE, that event will toggle the squelch.
If SQL_DET is set to GPIO this configuration variable is used to choose which GPIO pin to use for squelch input. The most common name is gpio<number>, like gpio4. Some GPIO drivers use more complex names, like gpio33_pe11. If inverted operation is desired, prefix the pin name with an exclamation mark (!).

Example: GPIO_SQL_PIN=!gpio4

Choose which type of signal level detector to use. There are two choices, "NOISE" or "TONE". The signal level detector is only needed when using multiple receivers in a voter configuration or when using the SIGLEV squelch type.

Type NOISE use a bandpass filter in the range of 5 - 5.5kHz (CARD_SAMPLE_RATE >= 16000) or a highpass filter at 3.5kHz (CARD_SAMPLE_RATE = 8000) to estimate the amount of noise present on the signal. If the passband contain a small amount of energy, a strong signal is assumed. If the passband contain more energy, a weaker signal is assumed. The noise detector must be calibrated for the receiver and audio levels you use. This is done using the SIGLEV_SLOPE and SIGLEV_OFFSET configuration variables. See chapter CALIBRATING THE SIGNAL LEVEL DETECTOR below for more information.

Type TONE is not really a signal level detector but rather a transport mechanism for getting signal level measurements from a remote receiver site, linked in via RF, to the main SvxLink site. It is using ten tones, one for each signal level step, in the high audio frequency spektrum (5.5 - 6.4kHz, 100Hz step) to indicate one of ten signal levels. Only the receiving part have been implemented in SvxLink at the moment. On the remote receiver side an Atmel AVR ATmega8 is used to map the signal level voltage to tone frequencies. Use the TONE_SIGLEV_MAP configuration variable to map each tone to a corresponding signal level value in between 0 - 100.

The slope (or gain) of the signal level detector. See chapter CALIBRATING THE SIGNAL LEVEL DETECTOR below for more information.
The offset of the signal level detector. See chapter CALIBRATING THE SIGNAL LEVEL DETECTOR below for more information.
This configuration variable set an upper threshold for the estimated signal level when using the noise signal level detector. If the estimation goes over the given threshold, a signal level of 0 will be reported. This can be used as a workaround when using a receiver with squelched audio output. When the squelch is closed, the receiver audio is silent. The signal level estimator will interpret this as a very strong signal. Setting up the bogus signal level threshold will counteract this behavior but a better solution is to use unsquelched audio if possible.

By default this feature is disabled. If enabling it, start with a value somewhere around 120.

This configuration variable is used to map tones to signal level values when SIGLEV_DET=TONE. It is a comma separated list of ten values in the 0 - 100 range. The first value map to the 5500Hz tone, the second to the 5600Hz tone and so on. The last value map to the 6400Hz tone. What levels the tones should be mapped to depends on the tone sender implementation. The default tone map is 10,20,30...,100.

The Atmel AVR processor used by the author have a reverse mapping so that the first tone (5500Hz) indicate the highest signal strength and the last tone (6400Hz) indicate the lowest signal strength. It is also not linear since it's more important to have fine measurement granularity in the lower signal strength range. This is how the mapping look for the AVR: 100,84,60,50,37,32,28,23,19,8.

This is the squelch open threshold for the SIGLEV squelch detector. If using the NOISE signal level detector, make sure to first calibrate the signal level detector using the SIGLEV_SLOPE and SIGLEV_OFFSET configuration variables. The signal level detector should normally be calibrated so that full signal strength is 100 and no signal is 0. Depending on your background noise level a good value for this configuration variable is between 5 and 20.
This is the squelch close threshold for the SIGLEV squelch detector. If using the NOISE signal level detector, make sure to first calibrate the signal level detector using the SIGLEV_SLOPE and SIGLEV_OFFSET configuration variables. The signal level detector should normally be calibrated so that full signal strength is 100 and no signal is 0. Depending on your background noise level a good value for this configuration variable is between 1 and 10.
Apply a deemphasis filter on received audio. The deemphasis filter is used when taking audio directly from the detector in the receiver, like when using a 9k6 packet radio connector. If not using a deemphasis filter the high frequencies will be amplified resulting in a very bright (tinny) sound.
Squelch tail elimination is used to remove noise from the end of a received transmission. This is of most use when using CTCSS or SIGLEV squelch with unsquelched input audio. A normal value is a couple of hundred milliseconds. Note that the audio will be delayed by the same amount of milliseconds. This does not matter much for a simplex link but for a repeater the delay might be annoying since you risk hearing the end of your own transmission.
The incoming signal will be amplified by the specified number of dB. This can be used as a last measure if the input audio level can't be set high enough on the analogue side. A value of 6dB will double the signal level. Note that this is a digital amplification. Hence it will reduce the dynamic range of the signal so usage should be avoided if possible. It's always better to correct the audio level before sampling it.
This is a help to adjust the incoming audio level. If enabled it will output a message when distorsion occurs. To adjust the audio level, first open the squelch. Then increase the audio level until warning messages are printed. Decrease the audio level until no warning messages are printed. After the adjustment has been done, the peak meter can be disabled. 0=disabled, 1=enabled.
Specify the DTMF decoder type. Set it to INTERNAL to use the internal software DTMF decoder. To use the S54S interface featuring a hardware DTMF decoder, set it to S54S. To control it over a pseudo tty device set it to PTY. Setting it to PTY will install the PTY dtmf decoder. SvxLink creates a symlink linked to a slave pty device on runtime. The name has to be defined with DTMF_PTY. NONE or commenting it out will disable DTMF detection.
Mute the audio during the time when a DTMF digit is being received. Note that the audio will be delayed 75ms to give the DTMF detector time to do its work. This does not matter much on a simplex link but on a repeater it could be annoying since you will hear the last 75 milliseconds of your own transmission. To counteract the added delay one can set up the SQL_TAIL_ELIM configuration variable to at least 75 milliseconds. Legal values for DTMF_MUTING are 0=disabled, 1=enabled.
This configuration variable can be used if the DTMF decoder is too quick to indicate digit idle. That does not matter at high signal strengths but for weaker signals and mobile flutter it's not good at all. Each DTMF digit will be detected multiple times. Using this configuration variable, the time (ms) a tone must be missing to be indicated as off can be extended. Setting this value too high will cause the decoder to be a bit sluggish and it might consider two digits as one. The hang time only affect consecutive digits of the same value (e.g. 1 1). If a detected digit differs from the previously detected digit (e.g 1 2), the hang time is immediately canceled and the detected digit is considered as a new one. A good default value is 50-100ms.
When using an external hardware DTMF decoder this config variable is used to specify a serial port (e.g. /dev/ttyS0).
When using the PTY DTMF "decoder" this configuration variable will set the path to the PTY slave softlink that the external interface script use to communicate to SvxLink. Over this symlink a very easy communication protocol is used to tell SvxLink received DTMF digits: 0-9, A-F, *, #. "E" is the same as "*" and "F" is the same as "#". Sending a digit tell SvxLink when it starts. To tell SvxLink that the digit has ended, send a space character.

The PTY DTMF "decoder" can be used by an external script to interface custom devices, modems or other hardware to SvxLink. Look for nhrcx.pl to see an example.

Example: /tmp/rx1_dtmf.

DTMF use two tones to encode digits 0-9, A-F. These two tones should normally have the sample amplitude. The difference in amplitude is called twist. Forward twist is when the higher frequency tone is lower in amplitude than the lower frequency tone. According to the standards, 8dB forward twist should be allowed. Some transmitters do not correctly modulate the DTMF tones to get zero twist. The most common situation is that the forward twist is too large. Increasing this configuration variable above 8dB might allow DTMF from these transmitters to be detected. When doing this, the DTMF detector will be more sensitive to noise and might cause more false triggers.
DTMF use two tones to encode digits 0-9, A-F. These two tones should normally have the sample amplitude. The difference in amplitude is called twist. Reverse twist is when the lower frequency tone is lower in amplitude than the higher frequency tone. According to the standards, 4dB reverse twist should be allowed. The most common reason for getting reverse twist is a bad de-emphasis filter or that none at all is used, like when taking audio directly from the FM discriminator. Have a look at the DEEMPHASIS configuration variable before starting to modify this configuration variable.
1750_MUTING
Mute the audio during a call tone of 1750Hz is received. Note that the audio will be delayed 75ms to give the tone detector time to do its work. This does not matter much on a simplex link but on a repeater it could be annoying since you will hear the last 75 milliseconds of your own transmission. To counteract the added delay one can set up the SQL_TAIL_ELIM configuration variable to at least 75 milliseconds. Legal values for 1750_MUTING are 0=disabled, 1=enabled.
Define here your selective tone call system. You have the choice of the following types: ZVEI1, ZVEI2, ZVEI3, PZVEI, PDZVEI, DZVEI, CCITT, EEA, CCIR1, CCIR2, NATEL, EURO, VDEW, AUTO-A, MODAT, PCCIR and EIA. Only one system can be used at the same time. Please take into consideration that some Sel5 standards are using the same or similar tones so it may have some unwanted effects if you define ZVEI1 for SvxLink and a (e.g.) ZVEI3 sequence is received.
At the moment only SEL5_DEC_TYPE=INTERNAL is valid. Maybe we have support for some external tone detectors later. To disable SEL5 tone decoding, specify NONE or just comment the configuration variable out.

Receiver type "Voter" is a "receiver" that combines multiple receivers and selects one of them to take audio from when the squelch opens. Which receiver to use is selected directly after squelch open. It is possible to set up a voting delay which will make the voter wait a while before choosing which RX to use. This will give all receivers some time to report their signal strengths. After the initial choice have been made a periodic check is done to see if any of the other receivers receive a stronger signal. In the default configuration file there is a voter section called Voter.

Always "Voter" for a voter.
Specify a comma separated list of receivers that the voter should use. Example: RECEIVERS=Rx1,Rx2,Rx3
Specify the delay in milliseconds that the voter will wait after the first sqeulch open detection until the decision of which receiver to use is made. This time must be set sufficiently high to allow all receivers to calculate and report the signal level. Incoming audio and DTMF digits will be buffered for all receivers during the delay time so nothing will be lost, but of course the audio will be delayed the specified amount of time. This is most noticeble when using a repeater logic. Use the BUFFER_LENGTH configuration variable to adjust the buffer length. The default voting delay is 0.
Use this configuration variable to adjust the length of the voting delay buffer. If not specified, the buffer length will be the same as the voting delay. When using the voter with a repeater logic, try to keep this variable at 0 to reduce the latency. Only increase it if you feel audio is lost in the beginning of transmissions.
This is the interval time in milliseconds with which the voter will check if another receiver is receiving a stronger signal. If that is the case, a receiver switch will be initiated. Default is 1000 milliseconds.
The hysteresis setting will prevent the voter from switching back and forth between two receivers that are equal in signal strength. For a switch to occur, the other receivers signal strength must exceed the current receivers signal strength by the percent specified in this configuration variable. So if the hysteresis is set to 50% and the received signal strength on the current receiver is 40, a signal strength of 40*1.5=60 is required on another receiver to initiate a switch. At squelch open, if the received signal strength plus hysteresis is larger than 100, the voting delay will be skipped. The default hysteresis is 50 percent.
When a receiver switch is initiated by the voter, it will wait the number of milliseconds specified in this configuration variable before actually performing the switch. The switch will only occur if the other receivers signal strength is still higher. Default is 500 milliseconds.
The voter will wait the number of milliseconds specified in this config variable after a squelch close before voting in another receiver. There are two reasons for using this delay. The first is to prevent the voter from going into idle state immediately when the squelch close for a fluttery signal. If it goes to idle, the procedure with voting delay may cause longer dropouts than necessary. The second reason to use this config variable is if different receivers have different hang times (explicitly or implicitly). If both a slow and fast receiver is receiving the same signal and the faster is currently chosen, when the PTT is released the slower receiver will be voted before closing. This will cause a double squelch tail and double roger beep. Default is 500 milliseconds.

A networked receiver section is used to specify the configuration for a receiver connected through a TCP/IP network. In the default configuration file there is a networked receiver configuration section called NetRx. The section name could be anything. It should match the RX configuration variable in the logic core where the receiver is to be used. The available configuration variables are described below. How to use a networked receiver is further described in the remotetrx(1) manual page.

Always "Net" for a networked receiver.
The hostname or IP address of the remote receiver host.
This is the authentication key (password) to use to connect to the RemoteTrx server. The same key have to be specified in the RemoteTrx configuration. If no key is specified in the RemoteTrx config, the login will be unauthenticated. A good authentication key should be 20 characters long. If the same RemoteTrx is used for both RX and TX, the same key must be specified in the RX as well as in the TX configuration section. The key will never be transmitted over the network. A HMAC-SHA1 challenge-response procedure will be used for authentication.
The audio codec to use when transferring audio from this remote receiver. Available codecs are: RAW (512kbps), S16 (256kbps), GSM (13.2kbps), SPEEX (8-25kbps), OPUS (8-64kbps). These are raw bitrate values. There will be some overhead added to this so the real bitrates on the wire are a little bit higher. The OPUS codec is the most modern one and it also have the best quality for a given bitrate.
Speex encoder setting. Each Speex frame contains 20ms audio. If using a low bitrate configuration, the network overhead will be quite noticeable if sending each frame in its own packet. One way to lower the overhead is to send multiple frames in each network packet. The drawback with doing this is that you get more delay. If setting this option to something like 4 (default), the delay will be about 4x20=80ms.
Speex encoder setting. Specify the encoder quality using a number between 0-10. Lower values give poorer quality and lower bitrates.
Speex encoder setting. Specify the bitrate to use. Speex will snap to the nearest lower possible bitrate. Possible values range from 2150 to 24600 bps. You should probably not specify quality at the same time as bitrate. Not sure though...
Speex encoder setting. The complexity setting (0-10) tells the encoder how much CPU time it should spend on doing a good job. The difference in SNR between the lowest and highest value is about 2dB. Set it as high as possible without overloading the CPU on the encoding computer (check CPU usage using command "top").
Speex encoder setting. Enable (1) or disable (0) variable bitrate encoding. If enabled, the encoder will try to keep a constant quality by increasing the bitrate when needed.
Speex encoder setting. The quality (0-10) to use in variable bitrate mode.
Speex encoder setting. The average bitrate encoding will try to keep a target bitrate by continuously adjusting the quality. This configuration variable specify the target bitrate and enable ABR. It also need to have VBR enabled so don't force it to off.
Speex decoder setting. Enable (1) or disable (0) the perceptual enhancer in the decoder. Perceptual enhancement is a part of the decoder which, when turned on, attempts to reduce the perception of the noise/distortion produced by the encoding/decoding process. In most cases, perceptual enhancement brings the sound further from the original objectively (e.g. considering only SNR), but in the end it still sounds better (subjective improvement).
Opus encoder setting. Specify how large, in milliseconds, each audio packet should be. Default: 20ms.
Opus encoder setting. The complexity setting (0-10) tells the encoder how much CPU time it should spend on doing a good job. Set it as high as possible without overloading the CPU on the encoding computer (check CPU usage using command "top"). Default: 10.
Opus encoder setting. This is the bitrate that the encoder will encode for. Rates from about 8000 to 64000 bits per second are meaningful but the codec can handle from like 2500 to 512000 bps. Default: 20000bps.
Opus encoder setting. Enable (1) or disable (0) variable bitrate encoding. If enabled, the encoder will try to keep a constant quality by increasing the bitrate when needed and decrease it when the quality can be assured with a lower bitrate. The target average bitrate is the one set by OPUS_ENC_BITRATE. Default: 1.

A local transmitter section is used to specify the configuration for a local transmitter. In the default configuration file there is a configuration section called Tx1. The section name could be anything. It should match the TX configuration variable in the logic core where the transmitter is to be used. The available configuration variables are described below.

Always "Local" for a local transmitter.
Specify the audio device to use. Normally alsa:plughw:0. Have a look at the AUDIO DEVICE SPECIFICATIONS chapter for more information.
Specify the audio channel to use. SvxLink can use the left/right stereo channels as two mono channels. Legal values are 0 or 1.
Use this configuration variable to specify which type of hardware to use to control the PTT. Specify "SerialPin" for using a pin in the serial port, "GPIO" to use a pin in a GPIO port or "PTY" if you want to use an external interface script via a pseudo tty port.

Set PTT_TYPE to "Dummy" or "NONE" to not use any PTT hardware at all. It is an error to not specify PTT_TYPE.

Use PTT_PIN to specify the pin to use for "SerialPin" or "GPIO".

Specify the serial port that the PTT is connected to. E.g. /dev/ttyS0 for COM1.
If PTT_TYPE is set to "SerialPin", specify the pin(s) in the serial port that the PTT is connected to. It is possible to specify one or two serial port pins. Some interface boards require that you specify two pins since one pin does not provide enough drive power to the circuit. A "!" in front of the pin name indicates inverted operation. Some of the possible values are RTS, DTRRTS, !DTR!RTS or even DTR!RTS.

If GPIO was specified in PTT_TYPE, set the PTT_PIN config variable to the pin name of the GPIO-pin to use. The most common name is gpio<number>, like gpio3. Some GPIO drivers use more complex pin names like gpio33_pe11. Have a look at the USING GPIO section for information on how to set up the operating system. Normally, the pin will be active high but if the pin name is prefixed with an exclamation mark it will be active low instead.

If PTT_TYPE is set to "PTY" this configuration variable will set the path for the PTY slave softlink that is used by the external script to communicate to SvxLink.

SvxLink sends a 'T' to start transmitting and a 'R' to turn the transmitter off. This can be used by an external script to interface custom devices, modems or other hardware to SvxLink. Look for nhrcx.pl to see an example.

Set the specified serial port pins to a static state. This can be good if using a pin for reference voltage or if a pin have to be in a certain state to not interfere with the operation of some equipment. There are two pins that are possible to use, RTS and DTR. If prefixed with an exclamation mark (!), the pin will be cleared and if not it will be set. This configuration variable can only be used when PTT_TYPE is set to "SerialPin".

Example: SERIAL_SET_PINS=RTS!DTR will set RTS and clear DTR.

Use this configuration variable to set a PTT hangtime. This can be good to have on a transmitter in combination with using a tone squelch. When the transmitter is ordered to stop transmitting, the tone is immediately turned off, causing the squelch to close on the other side. Since the transmitter keeps transmitting for a while, no squelch tail will be heard.

Another use is on a remote receiver link transmitter where you don't want the transmitter to turn on and off between transmissions or if the squelch close and open quickly due to for example mobile flutter.

This is a feature that will prevent the transmitter from getting stuck transmitting. Specify the number of seconds before the transmitter is turned off. Note that this is a low level security mechanism that is meant to only kick in if there is a software bug in SvxLink. Just so that the transmitter will not transmit indefinately. It is not meant to be used to keep people from talking too long.
The number of milliseconds (0-1000) to wait after the transmitter has been turned on until audio is starting to be transmitted. This can be used to compensate for slow TX reaction or remote stations with slow reacting squelches.
The frequency in Hz of the CTCSS tone to transmit. It is possible to specify fractions using "." as decimal comma (e.g. 136.5). For the tone to be transmitted the CTCSS_LEVEL variable must also be setup and also the TX_CTCSS variable in the logic core configuration section.
The level in percent (0-100) of the CTCSS tone to transmit. What level to set is hard to say. The FM modulation swing of the tone should be in between 500-800 Hz. That is a bit hard to measure if you don't have the right equipment. A normal FM station have a maximum swing of 5kHz so if you manage to calibrate everything so that you get maximum swing when the sound card audio is at peak level, the tone level should be in between 10-16%. However, most often the audio settings are configured a bit higher than max since the audio seldom reaches maximum level. Then the level of the CTCSS tone should be reduced. The default in the configuration file is 9%. For the tone to be transmitted the CTCSS_FQ variable must also be setup and also the TX_CTCSS variable in the logic core configuration section.

Note: The level of the tone affects the level of the rest of the audio in SvxLink. This is to avoid distorision when the two audio streams are mixed together. For example, if a tone level of 9% is setup the rest of the audio will be attenuated by 9%. This is true even if the CTCSS_FQ and TX_CTCSS configuration variables are not set so comment this configuration variable out if CTCSS on TX is not used.

[EXPERIMENTAL] Enable this feature if you are modulating the FM modulator directly, like through a 9k6 packet radio input. If no preemphasis filter is applied to the audio, it will sound very dark when received. If you modulate the transmitter through the microphone input the radio will apply a preemphasis filter so this feature should be disabled. 0=disabled, 1=enabled.
The length, in milliseconds, of DTMF digits transmitted on this transmitter. 100ms is the default.
The spacing, in milliseconds, between DTMF digits transmitted on this transmitter. 50ms is the default.
The amplitude, in dB, of DTMF digits transmitted on this transmitter. Zero dB is the maximum amplitude. -18dB is the default.
It is possible to transmit one of ten tones along with the normal transmission to indicate a signal strength value to the receiver. This is of most use when using a local transmitter as a link transmitter in a RemoteTrx. It is not implemented, and probably not useful, in SvxLink Server. Another requirement is that SvxLink has been compiled in 16kHz mode. Otherwise this feature is disabled.

The TONE_SIGLEV_MAP configuration variable is used to map tones to signal level values. It is a comma separated list of exactly ten values in the 0 - 100 range. The first value map to the 5500Hz tone, the second to the 5600Hz tone and so on. The last value map to the 6400Hz tone. What levels the tones should be mapped to depends on the tone receiver implementation. Typically, if using a SvxLink application as a receiver, the TONE_SIGLEV_MAP should be the same in the RX configuration for that node.

It is possible to transmit one of ten tones along with the normal transmission to indicate a signal strength value to the receiver. This is of most use when using a local transmitter as a link transmitter in a RemoteTrx. It is not implemented, and probably not useful, in SvxLink Server. Another requirement is that SvxLink has been compiled in 16kHz mode. Otherwise this feature is disabled.

The TONE_SIGLEV_LEVEL configuration variable is used to set the tone level. It is a value in the 1-100 range which indicate the output level in percent of the maximum possible level. The default is 10.

A networked transmitter section is used to specify the configuration for a transmitter connected through a TCP/IP network. In the default configuration file there is a networked transceiver configuration section called NetTx. The section name could be anything. It should match the TX configuration variable in the logic core where the transmitter is to be used. The available configuration variables are described below. How to use a networked transmitter is further described in the remotetrx(1) manual page.

Always "Net" for a networked transmitter.
The hostname or IP address of the remote transmitter host.
This is the authentication key (password) to use to connect to the RemoteTrx server. The same key have to be specified in the RemoteTrx configuration. If no key is specified in the RemoteTrx config, the login will be unauthenticated. A good authentication key should be 20 characters long. If the same RemoteTrx is used for both RX and TX, the same key must be specified in the RX as well as in the TX configuration section. The key will never be transmitted over the network. A HMAC-SHA1 challenge-response procedure will be used for authentication.
The audio codec to use when transferring audio to this remote transmitter. Available codecs are: RAW (512kbps), S16 (256kbps), GSM (13.2kbps), SPEEX (8-25kbps), OPUS (8-64kbps). These are raw bitrate values. There will be some overhead added to this so the real bitrates on the wire are a little bit higher. The OPUS codec is the most modern one and it also have the best quality for a given bitrate. There also is a NULL codec that will just throw away samples which can be used in special situations when the audio is sent through another audio path.
Speex encoder setting. Each Speex frame contains 20ms audio. If using a low bitrate configuration, the network overhead will be quite noticeable if sending each frame in its own packet. One way to lower the overhead is to send multiple frames in each network packet. The drawback with doing this is that you get more delay. If setting this option to something like 4 (default), the delay will be about 4x20=80ms.
Speex encoder setting. Specify the encoder quality using a number between 0-10. Lower values give poorer quality and lower bitrates.
Speex encoder setting. Specify the bitrate to use. Speex will snap to the nearest lower possible bitrate. Possible values range from 2150 to 24600 bps. You should probably not specify quality at the same time as bitrate. Not sure though...
Speex encoder setting. The complexity setting (0-10) tells the encoder how much CPU time it should spend on doing a good job. The difference in SNR between the lowest and highest value is about 2dB. Set it as high as possible without overloading the CPU on the encoding computer (check CPU usage using command "top").
Speex encoder setting. Enable (1) or disable (0) variable bitrate encoding. If enabled, the encoder will try to keep a constant quality by increasing the bitrate when needed.
Speex encoder setting. The quality (0-10) to use in variable bitrate mode.
Speex encoder setting. The average bitrate encoding will try to keep a target bitrate by continuously adjusting the quality. This configuration variable specify the target bitrate and enable ABR. It also need to have VBR enabled so don't force it to off.
Speex decoder setting. Enable (1) or disable (0) the perceptual enhancer in the decoder. Perceptual enhancement is a part of the decoder which, when turned on, attempts to reduce the perception of the noise/distortion produced by the encoding/decoding process. In most cases, perceptual enhancement brings the sound further from the original objectively (e.g. considering only SNR), but in the end it still sounds better (subjective improvement).
Opus encoder setting. Specify how large, in milliseconds, each audio packet should be. Default: 20ms.
Opus encoder setting. The complexity setting (0-10) tells the encoder how much CPU time it should spend on doing a good job. Set it as high as possible without overloading the CPU on the encoding computer (check CPU usage using command "top"). Default: 10.
Opus encoder setting. This is the bitrate that the encoder will encode for. Rates from about 8000 to 64000 bits per second are meaningful but the codec can handle from like 2500 to 512000 bps. Default: 20000bps.
Opus encoder setting. Enable (1) or disable (0) variable bitrate encoding. If enabled, the encoder will try to keep a constant quality by increasing the bitrate when needed and decrease it when the quality can be assured with a lower bitrate. The target average bitrate is the one set by OPUS_ENC_BITRATE. Default: 1.

A multi transmitter section is used if one wants to transmit on multiple transmitters simulaneously. The name of the section can be anything. Just point it out from another transmitter specification like the TX variable in a Logic core configuration.

Always "Multi" for a multi transmitter section.
A comma separated list of transmitters.

A module section contain the configuration for a specific module. It have some general configuration variables and some module specific configuration variables. The general configuration variables are listed below.

The name of the module. This name must match the namespace used in the TCL event handling script. If not set, NAME will be set to the section name.
The base name of the plugin. For example if this configuration variable is set to Foo, the core will look for a plugin called ModuleFoo.so. If not set, PLUGIN_NAME will be set to the same value as NAME.
Specify the module identification number. This is the number used to access the module from the radio interface.
Specify the timeout time, in seconds, after which a module will be automatically deactivated if there has been no activity.

Module specific configuration variables are described in the man page for that module. The documentation for the Parrot module can for example be found in the ModuleParrot.conf(5) manual page.

Enter a space separated list of EchoLink status servers that should be used to send node status beacons. Your node information can be found on http://www.echolink.org/links.jsp. The format is host:port. Host - hostname or IP address, port - UDP port. Don't change the default unless you know what you are doing. If you don't want to update the EchoLink status server, comment out this configuration variable.

Example:
STATUS_SERVER_LIST=aprs.echolink.org:5199

This configuration variable specifies connection parameters for connecting to an APRS server network using the TCP protocol. In this case, the positioning information is forwarded to the worldwide APRS network. Have a look at http://aprs.fi/.

To choose a suitable APRS server from the so called tier 2 network, have a look at http://www.aprs2.net/. Either choose a specific server or one of the regional addresses. The regional addresses bundle all APRS servers within a region so that a random tier 2 server is chosen within the region. There are five regions defined: noam.aprs2.net - North America, euro.aprs2.net - Europe, asia.aprs2.net - Asia, soam.aprs2.net - South America and Africa, aunz.aprs2.net - Austrailia and New Zeeland. The format is a space separated list of host:port entries. Host - hostname or IP address, port - TCP port.

Example:
APRS_SERVER_LIST=euro.aprs2.net:14580

The longitude of the station position, entered as "degrees.arcminutes.arcseconds"

Example:
LON_POSITION=09.02.20E

The latitude of the station position, entered as "degrees.arcminutes.arcseconds"

Example:
LAT_POSITION=51.02.22N

Enter your callsign for the APRS network with a prefix that indicates the type of station, (ER- for repeaters, EL- for links).

Examples:
CALLSIGN=EL-DL1ABC # callsign for a link
CALLSIGN=ER-DB0ABC # callsign for a repeater

The tx-frequency of the link/repeater in MHz. For repeaters, information about the RX/TX shift in the COMMENT configuration variable may be useful.

Example:
FREQUENCY=430.050 # tx-frequency is 430.050 MHz

The power of your transmitter in watts.

Example:
TX_POWER=10 # tx output is 10 watts

The gain of your antenna in dBd.

Example:
ANTENNA_GAIN=5 # antenna gain is 5 dBd

The height of the link-/repeater antenna in meters or feet above the terrain, not sealevel.

Example:
ANTENNA_HEIGHT=10m # 10 meters above the ground
ANTENNA_HEIGHT=90 # 90 feet

Main beam direction of the antenna in degrees. If an omni direction antenna is used, specify -1 as the direction.

Example:
ANTENNA_DIR=-1 # an omni directional antenna is used
ANTENNA_DIR=128 # main beam direction is 128 degrees

The PATH variable controls the way of forwarding your beacon inside the APRS network if it is gated by a local APRS digipeater. In some cases it has to be changed according to local requirements. Please contact your local APRS sysop for further information. Changes should be made only according to the NEWn-N paradigm. Leave this variable untouched if you are unsure of its setting. No spaces or control characters are allowed. PATH has no influence on the propagation on non-RF networks.

Examples:
PATH=WIDE1-1
PATH=WIDE1-1,WIDE2-2

The interval, in minutes, with which beacons will be sent to the APRS network. A good value is 10 minutes. If your beacon is gated via RF, please increase the interval a bit to keep the APRS traffic on RF produced by the APRS RF gate as low as possible. Intervals shorter than 10 minutes will be changed to 10.

Example:
BEACON_INTERVAL=30 # APRS-beacons will be sent every 30 minutes.

The CTCSS subaudible tone that is to be used for operation over your link or repeater. If you don't use tone control set it to 0.

Examples:
TONE=136 # we are using a CTCSS-tone of 136.5 Hz
TONE=0 # we don't use CTCSS subaudible or call tones
TONE=1750 # the link/repeater use a tone burst of 1750 Hz

Defines the interval in minutes in that a aprs statistic is sent into the aprs network. Range: 5-60, default is 10 minutes
Specify a short comment here, maybe a link to your website or information that could be interesting for others. The length should not exceed 255 characters and may not have control characters like "Carriage Return" (\r) or "Line Feed" (\n) inside. Make your comment as short as you can to give users with a small display (TH-D7) the chance to display the full comment text.

Example:
COMMENT=[svx] Running SvxLink by SM0SVX

The AUDIO_DEV configuration variables specify which audio device to use for a receiver or transmitter. SvxLink support a number of different audio input and output devices. The format of the configuration variable is "type:dev_spec". There are three different types of audio devices supported, "alsa", "oss" and "udp".

The "alsa" type will use the specified Alsa device. Example: "alsa:plughw:0". Describing the format of Alsa device names is outside the scope for this document.

The "oss" type will use the specified OSS audio device. Example "oss:/dev/dsp". OSS is the old sound system used by Linux. Alsa should be used when possible.

The "udp" type is not really an audio device but instead will read and write audio from/to a UDP socket. This can be used to interface SvxLink to all sorts of audio sources/sinks capable of streaming raw audio through UDP. One example usage is to interface SvxLink with GNU Radio. Example: "udp:127.0.0.1:10000". Note however that the only supported format is raw 16 bit signed samples, two interleved channels. Sampling frequency can be chosen using the CARD_SAMPLE_RATE config variable as usual.

GPIO (General Purpose IO) is used to access hardware pins that are made available for example on an embedded system. Before starting to use a pin in SvxLink some setup need to be done in the operating system. The example below will set up pin 31 as an output for PTT use and pin 30 is used as input for squelch. Make sure that the commands are run as user root.


Enable the PTT pin for GPIO:
echo 31 > /sys/class/gpio/export


Set the direction to output for the pin:
echo out > /sys/class/gpio/gpio31/direction


Make sure that the svxlink user can write to the GPIO pin:
chown svxlink /sys/class/gpio31/value


Enable the squelch pin for GPIO:
echo 30 > /sys/class/gpio/export


Set the direction to input for the pin:
echo in > /sys/class/gpio/gpio30/direction


Make sure that the svxlink user can read from the GPIO pin:
chown svxlink /sys/class/gpio30/value

These steps have to be performed whenever the system is rebooted. Putting the commands in a script is recommended.

The signal level detector is used when using multiple receivers or when using the SIGLEV squelch. The signal level is used by a voter to choose the receiver with the highest signal strength. The choice is made directly after squelch open. For the voter to make a correct choice, the signal level detector must be calibrated on each receiver.

To use the noise signal level detector, first set SIGLEV_DET=NOISE. There are two configuration variables that is used to calibrate the detector. They are SIGLEV_SLOPE and SIGLEV_OFFSET in a local receiver section. The slope is the gain of the detector and the offset is used to adjust the detector so that when there is no input signal, the detector will return 0. The goal is to adjust the detector so that when no signal is received, a value of 0 is produced and when full signal strength is received, a value of 100 is produced. It will never be exakt but that does not matter.

The calibration is normally done by using the siglevdetcal(1) application. To be able to do a correct calibration, it must be possible to open the squelch so that only noise is received. The antenna cable should be disconnected or a dummy load should be used. WARNING: Before starting the siglevdetcal application, pull the PTT cable since the PTT might get triggered during the calibration procedure.

The siglevdetcal utility will also measure the CTCSS tone SNR offset so that the CTCSS_SNR_OFFSET config variable can be set up to a proper value.

If the siglevdetcal application cannot be used for some reason, the manual procedure below might be used. This procedure will only work for a receiver with unsquelched audio. Note: To calibrate a remote receiver it must be connected to the SvxLink server. Otherwise the squelch will not open.

1
Connect a dummy load or disconnect the antenna from the transceiver. If you disconnect the antenna, make sure to also disconnect the PTT.
2
Set SIGLEV_SLOPE=1 and SIGLEV_OFFSET=0 and restart SvxLink.
3
Open the squelch so that there is only noise coming into SvxLink.
4
Use a second transceiver to make a short, unmodulated transmission. Release the PTT when the "Squech OPEN" message is printed. Repeat this for about five times.
5
Calculate the mean diff (open level - close level) and the mean lower (squelch close) value. Make sure to use at least four significant digits in your calculations.
6
SIGLEV_SLOPE = 100 / (mean diff)
7
SIGLEV_OFFSET = - (mean lower) * SIGLEV_SLOPE
8
After changing SIGLEV_SLOPE and SIGLEV_OFFSET, restart SvxLink and check to see that the squelch open value is now around 100 and the squelch close value is around 0.

/etc/svxlink/svxlink.conf (or deprecated /etc/svxlink.conf)
The system wide configuration file.
~/.svxlink/svxlink.conf
Per user configuration file.
/etc/svxlink/svxlink.d/*
Additional configuration files. Typically one configuration file per module.

Tobias Blomberg (SM0SVX) <sm0svx at users dot sourceforge dot net>

svxlink(1), remotetrx(1), siglevdetcal(1)