Meshtastic Nodes

• It's been a while since the last March update! Glad to see the Meshtastic firmware has moved on a fair bit and along with that I passed my UK Ham Radio Foundation exam back on 2024-05-20 and now operate with the callsign, M7VOK. With that comes a few changes to the nodes:

  • The node "eth0" changes it's long_name from "m.ethzero.uk/#router" to simply "m.ethzero.uk" as Meshtastic have dropped the "router" role

  • The node "eth1" changes it's long_name from "m.ethzero.uk/#dipole" to my Ham Radio callsign, M7VOK

  • Max Hops had somehow changed to "7", this has been pulled back to the recommended default of "3"

  • Updated nodes to v2.5.11.8e2a3e5 Beta

• The software that drives the live node list is now available as a public alpha on GitHub

• Good news! I'm preparing the source code + documentation for the Live Node List I use on this website for publishing on GitHub. Sadly I can't give an ETA as when writing the documentation I realised one of the requirements would limit the type of web hosting, so I'll be re-writting that part for a broader audience.

  Thanks to all that have found the Live Node information useful and have expressed an interest in this setup.

• Updated #router & #dipole nodes from firmware 2.3.1 alpha to v2.3.2.63df972 Alpha

• Updated #router & #dipole nodes from firmware 2.3.0 alpha to v2.3.1.4fa7f5a Alpha

• Wiped the #router & #dipole NodeDBs as there was a fair few nodes that where unlikely to ever made contact again as I believe this was due to a Meshtastic user broadbasting from a airplane. This will result in a cleaner Live Neighbouring Nodes List

• Updated #router & #dipole nodes from firmware 2.2.24 beta to 2.3.0 alpha

• Found out that I need to temporarily disable the live node data gathering program that interact with the serial port while I'm updating the firmware as, for some odd reason, the firmware flasher doesn't appear to call an exclusive lock on the serial port, causing the following somewhat oblique error message: A fatal error occurred: Invalid head of packet (0x11): Possible serial noise or corruption.

• Fired-up a very old but perfectly capable Raspberry Pi 2 Model B and installed Raspberry Pi OS Lite (without the Desktop) preconfiguring all of the networking & SSH stuff using the Raspberry Pi Imager.

• Next step was to install the excellently documented Meshtastic Python CLI for managing the nodes & esptool for flashing their firmwares.

• After a few hours of testing both tools succesfully, I notice in the Pyhton CLI documention the ability to dump a list of neighbouring nodes, current and historic, listed in descending "LastHeard" time order but the use of the --nodes flag. I quickly realised this would be fairly trivial that this plain text, mash it with some simply HTML and periodically upload it to this website for a near real-time view.

  The result can be viewed in the neighbour nodes links in the node list above.

  Once I've tidied up some of the code I'll glady upload the script I run every minute on the Raspberry Pi (via crond) and associated HTML files.

• Quite the rollcoaster today! Started off bumping the loft node firmwares from 2.2.23 beta to 2.3.0 alpha, then down to 2.2.24 beta and finally fully reverted back to 2.2.23 beta 🤦‍♀️. The primary reason was assist TechMind's Matt trace a fault where his node (at by virtue of the mesh, all Aylesbury nodes!) was completely uncontactable. While Matt mentioned there was some evidence people reporting packets where not being recieved with the 2.2.24 beta firmware ultimately he traced the issue to his antenna.

• With all the climbing up and down the loft ladder today, I'm going to investigate using an Raspberry Pi to power, firmware flash, and manage the two loft nodes.

• #router node firmly pointing towards Whitchurch Router & BigCW nodes. This has resulted in a ~+3 SNR bump from -0.5 to +2.5. This is due in no small part in noticing that the previously invisible Whitchurch repeater apparently reverted back to being Whitchurch Router meaning I could definitively determine via traceroute if I'm really making direct contact with M0DQW Base in Aylesbury or not.

• Took a drive out Milton Keynes way today to confirm the existance of a few nodes and to get an in-the-field feel for the insurmountable challenge of trying to contact MK from Bucks. Put simply, you'd have to have a node approximate mid-way along the A421 where the terrain peaks. Unsurprisingly I spotted am unknown radio tower out at Little Horwood

• First stop was Wolerton for the self-reported YNVC/Yanovich node. Unfortunately I couldn't get a hit even after pinging out serveral message on LONG_FAST in the nearby Tesco's carpark.

• Newport Pagnell ended up having a trove of nodes far more than on any Meshtastic map! While there is only one self-reported node in that area again, I parked up on a side street to broadbasting a test message on LONG_FAST. A whole bunch of nodes starting pouring in, including an acutall reply from 2BS/2e0npl who said, "You close to my location". That turned out to be very true as his node was only 38 meters away! I felt quite embarrased at that point to be somewhat randomly that close to someone's node so I posted a reply and make a discret exit 😊. The list of Newport Pagnell nodes spotted are:

  • UKF2/M0UKF Francis (Mobile Phone)
  • mez1/M0MEZ Eliot
  • MEZ2/M0MEZ Base Station
  • TAI2/G0TAI - Base Station
  • 809c/Meshtastic 809c
  • TAI1/G0TAI - client device
  • 2M/2e0npl Mobile
  • UKF1/M0UKF Home
  • 2BS/2e0npl Base Station

• #router node is back on the Yagi pointing to Aylesbury. Very glad to say the SNR to the nearest node is up from -1.0 (omni) to 2.0 (yagi). On a pratical point this means that DMs are consistently responding with a person-with-a-tick as opposed to the vague cloud-with-a-tick which just means the message was received by the mesh.

• CORRECTION: Apologies to all, but on re-reading the LongFast channel chatter I've subsequently realised that some of the original information is factually wrong or omitted. The key correction is seeker101 (skr1) position that was apparent apparently configured on the node as a fixed position in Bedford when in reality it was in Dunstanble Downs, which to be fair at 30 km is still pretty darned impressive!

  Conversation excerpts

  

  ...

  

  ...

  

  ...

  
• ORIGINAL:

• Exceptional success! The yagi didn't manage to pick up the MK nodes as indended but did manage to pick up a node 40 km across in Bedford, quite remarkable!

• The Bedford node was seeker101 (skr1) and with this node being able to contact DB2875-01 (DB01) has a lengthy chat, albeit my message had to be repeated about 5 times before they made it through

• Additionally a couple of other nodes appears around the Aylesbury area, M7SYN (M7SY) [qrz.com] and The Buzzard A (TBuA), the latter being temporarily located on Coombe Hill, Aylesbury with a elevation of 257m

• Finally got around to configuring SSL and subdomains on the this website meaning a more secure and slight shorting of the node names e.g from ethzero.uk/m/#router to m.ethzero.uk/#router

• New 12.5 dBi Paradar Yagi arrived today. Swapped this out from the equally new Paradar 8.5 dBi LoRa omnidirection long range antenna. Only had an hour to tweak the antenna position but did manage to achive a minor improvement for transmitting to towards Aylesbury. Further testing required.

• I'd pointed the antenna towards Newton Longville where there is one active node, Newton Longville 682c (G7PK) but alas the terrain between has proven a challege and yeilded no contact for 4 hours

• I've now re-pointed the antenna towards Wolerton in MK where there *may* be one or nodes at the expense of a temporary degraded link with Aylesbury. Unfortunately because map.mpowered247.com is self-reporting and not automatically updated this map it's actually that useful when trying to do qualative diagnostics. This is another long-shot but at least the terrain profile is more favourable albeit at a greater distance.

• 😵 RIP mobile node :(. Added a protoboard backpack to the device. Unfortunately some of the expose legs on the PCB must have touched some of the alternate power leads attached to the battery pack and shorted out. Replacements on order.

• Installation of loft antenna

• Update mode of loft node to ROUTER_CLIENT; node renamed from #station to #router

• First Direct Message over Meshtastic with Matt in Aylesbury

• Matt reports a *direct* traceroute with to my #router node with his newely installed 2.2m Paradar 12 dBi antenna

• Matt advice with antenna suspension

Finally decided on the format of the long_names for my nodes, a simple URI scheme that's also a valid URL.

Got a contact from BigCW @ 20:04 today in the Whitchurch area. Sent a bunch of messages detailing my contact with TechMind's Matt hailing from node M0DQW_Base2. I don't think any of these got through though.

Speaking of Matt, we switched over using Discord today (as the previous YT thread was ~20 comment deep!). The main result is with his help I've ordered a Paradar 868Mhz LoRa antenna, weatherproof for harsh outdoor environments, for HNT, LoRaWAN and FLARM – 8.5dBi.

The real hope is that not only will I not only be able to receive mesh data from the "Aylesbury" mesh group but be able to contact them. As a bonus, it'd be great if I can also hook up with western edge of the Milton Keynes nodes. If both meshes are contactable I'll reconfigure my "station" node running in CLIENT mode to run as a ROUTER_CLIENT.

The antenna is due to arrive 2024-02-28 so hopefully live shorty thereafter.

Finally, I played around with Meshtastic's Telemetry module that when you hook up a supported environmental sensor will transmit that information along with the device's battery and signal data and presented in the device list.

Shortly after moving my station node to the loft, I received my first LongFast message most likely via the "Whitchurch Router" between Buckingham & Aylesbury.

The message was from none other than M0DQW_Base2 (TMB2) aka TechMinds!
We've been having a back-for-forth in the YouTube comments section of his video A New Live Global Meshtastic Node Map - Meshmap.net, his latest comment it partically fasinating:

Ahhh yes! That's my test node that's in the loft on a short amazon lora antenna. I have been playing with Yagis, pointing north, but not from that node. So you received my node from my loft! ;-) Unfortunately I did not receive you though.. There is a router at Whitchurch, so maybe it went through that, but if you could reach that router/repeater then I would of heard you as its strong here. Middle of next week I have the 2.2m Paradar 11.5dBi antenna going above the roof on my house, so maybe we might be able to make contact then. I will be changing my node to a permanent node, most likely just called M0DQW_Base, so if you see it, try and send a message! :-) I just gotta figure out which one of my modules to use! lol. Cheers for the info!

From my Buckingham-based station node I can currently see 6 other nodes (including my own "mobile" node). I've attempted contact with all and while the client reports a "Delivered" status, I don't believe they've been actually recieved. I think may to down to the excellent setup my the other nodes, and my poor, far from optimal setup on my end.

Meshtastic Screenshot of the Android app's node list as off 2024-02-25 13:04.02
[Nodes]