i’m currently trying to figure out the best settings for my usecase:
5 to 10 nodes
all nodes should act as repeaters/forwarders
quick integration of new nodes, fast new node discovery
near real time communication/messaging
near real time location tracking of all nodes
some nodes are just used for location tracking and have no display or BT connection
18650 batteries
12 hours runtime
devices are T-Beams V1.1
Lora32 as stationary repeaters on elevated positions
This is my current config:
meshtastic --set position_broadcast_secs 60 # can't go much lower here
meshtastic --set gps_update_interval 60 # whats the unit here? seconds?
meshtastic --set send_owner_interval 1 # whats the unit here? i found an example that stated a value of 4 means a 1 hour interval
meshtastic --set is_low_power false
meshtastic --set screen_on_secs 30
meshtastic --set wait_bluetooth_secs 28800
This sets how often to refresh the shared database with the information of who owns that node.
The interval of 2 is to send that information every other network ping. An interval of 4 is to send it a quarter of the time.
Leaving it at the defaults are probably best.
From the docs:
/** How often should we try to get GPS position (in seconds) when we are in GpsOpMobile mode?
or zero for the default of once every 120 seconds
or a very large value (maxint) to update only once at boot.
*/
My idea was to set a lower interval to speed up node discovery.
I had the issue that after setting up a new node, although all nodes they were connected, when i was sending messages from the command line using the meshtastic CLI tool, these messages where not displayed in the app. The hardware devices showed the message on the built in screen, but the messages from the new node were not shown in the app, messages of other nodes were displayed correctly. After a long time, it just worked. So i assumed it was an issue with unknown owner IDs and i lowered the send_owner_interval to speed up the time new nodes get added to the node DB.
Btw, in src/mesh/MeshService.cpp:63 the comment seems to be wrong: nodeInfoPlugin.sendOurNodeInfo(NODENUM_BROADCAST, requestReplies); // Send our info (don't request replies)
The “dont request replies” is conditional.
In docs/software/power.md:54 there is a typo, “send*owner_interval” should be changed to “send_owner_interval”.
Hmm. I fixed a bug which would cause slow discovery in 1.1.23. discovery now should be rapid (less than 30 secs of a new node booting). Was discovery still slow for you?
I’m running 1.1.23alpha on all devices.
I just tried to reproduce and reinstalled the firmware on the lora V1, but now i don’t get a connection between the t-beams and the lora. I’m setting the channel via “meshtastic --seturl” on all devices (“long range but slower” mode). Not sure whats going on now.
Hmm, this seems to be a connectivity issue between the t-beams and the loras, but only when operating in long range or very long range mode. The loras can exchange messages and the t-beams can exchange messages, but there is no exchange between the loras and the t-beams.
I’m using two loras and three t-beams, channel is set via command line using the --seturl parameter. Firmware is 1.1.23alpha on all nodes. Meshtastic CLI is 1.1.29