Did you have to do any config on your Heltec V3 to get MeshDash working? 4403 isn’t open on a freshly flashed T-Deck (confirmed by nmap), and I’ve tried a few different versions of firmware. Maybe communicating via 4403 is just flat out not supported by the T-Deck - will have to keep tinkering.
I think the web maps are only showing positions allowed to be pushed to MQTT. I suspect MeshDash’s maps are already giving you the data directly from the mesh instead of via MQTT, but you can also access your node via the https://client.meshtastic.org/ interface and open the map there too.
I haven’t explored this in depth but I think the Heltec v3 can only handle one connection at a time whether that be bluetooth or wireless.
MeshDash works for me on those configurations but https://client.meshtastic.org does not if the meshdash connection is live. The Heltec also does not seem to want to go back to bluetooth mode unless I reflash it.
Can ESP32’s WiFi and Bluetooth be used at the same time?
No, the WiFi LoRa 32 don’t have an external PSRAM, if your application need use Bluetooth and Wi-Fi at the same time, you can look into this product: Heltec Wireless Bridge.
Interesting. I found the same thing on the T-Deck now that I’ve got the latest Meshtastic UI running on it. Can’t find anything in the docs regarding any incompatibility between the web interface and the on-device MUI, but I may have missed something. Perhaps a non-MUI firmware will sort out my MeshDash issue as well.
I woke up early this morning to find it in a boot loop state. Will re-flash it with the latest non-MUI firmware when I get a chance, and start from scratch yet again. Hopefully that gets me back on the mesh reliably. At the very least if I can start eliminating some variables that will probably help me figure out why I’ve dropped off in the last few weeks.
It turns out that port 4403 is actually open and that MeshDash works astonishingly well if you use the correct IP address for the Meshtastic device… (I will not be taking any follow up questions on that issue…)
I did also find a few references in GitHub issues in the Meshtastic repo that the MUI runs as a client and that devices can’t support more than one client connection so that possibly would not have helped (despite me using the wrong IP all along). I have gone back to the base Meshtastic firmware for the time being in case I was hitting a MUI/T-Deck/MUI+T-Deck specific bug or even something like an out of memory condition which was also causing issues. My node list and map have already begun to repopulate, so we’ll see if I can get two-way communications between some HLB members over the coming days.
I saw that pop up very early this morning, tried to send a direct message at 4:32am, and have received a single position report (about an hour ago).
I can see encrypted packets of other stations relaying my message to you @jdownie - curious to see if you’d received it or not. @zeeclor I’m yet to see ZeeLotan pop up in my node list, but I’m curious to see if MeshDash records a flurry of activity around 4:32am (my super secret encrypted “testing” message to @jdownie).
I was offline overnight. (My wife objected to the flashing of the Heltec in the room next to the bedroom.)
I have hills to the north and west of me but mostly flat ground to the south and east. I am going to reposition the device to the other side of the house. Given this what would you recommend for an antenna?
As an alternative I wondered if I could befriend 3 Cowry Close.
I recognise this is an absolute non-answer… but my recommendation is the best one that fits your budget and requirements.
I’m currently looking at various options myself and can’t offer anything more specific for the time being. The antenna I had on the T-Deck at Nundah a few weeks ago which you saw in person was a cheapie from overseas advertised as 5dBi gain, but I suspect it’s 2-2.5dBi gain at best. I’d argue that it’s completely impractical and limits the portability of the device so it’s probably not a “good” antenna from that perspective alone. As I’m looking at setting up a base station to play with, I’ll be shopping around for something larger/full wave/high gain/and so on. I’m happy to make a specific recommendation if I find something that works well for me, but it may not suit your requirements at all.
In general, aim for elevated positions, short cables, line of sight (no doubt one of the homes on the local islands would be a good match!), and limit the number of connectors. A quick search online suggests that the HELTEC V3 has those fiddly little u.fl connectors on the board (but please correct me with that) and kits often come with a u.fl connector and a pigtail to an SMA connector (again, please correct me if I’m wrong). Something suitable for 915MHz-928MHz with an SMA connector will likely be a good place to start. Be sceptical of any generic no-name unbranded antenna specifications, particularly gain figures. Any antenna that also claims to cover a huge frequency range or gives amazing coverage in every direction should also be viewed with a great deal of scepticism. Given the HELTEC V3’s advertised spec of 21dBm ±1 dBm, any antenna gain up to 8dBi will keep you under 1W, assuming an unrealistic cable/connector loss of zero (paging @techman to check my homework, please ).
If you want to explore outdoor antennas, you’re probably looking at $100+ for an excellent one. That’ll also be a trade off between the losses from having a really long cable between your antenna and the HELTEC V3 and the benefits of having an outdoor antenna (e.g., LoS, elevated, higher gain, etc.)
I’d suggest starting by figuring out how much you really want to spend, whether you want the HELTEC to be inside or outside (weatherproofing, power considerations, etc.), and how the cabling will pan out without having to cover a huge distance or involve a mess of 10 different connectors along the way. Once you’ve narrowed your requirements down to something along the lines of “$50 max, inside but it’ll be near a window, must have an SMA connector” then the antenna shopping is the easy (and fun) bit.
I was going to post a more in-detail article another day as I’m up to my neck in Nixos and Ollama atm, but for now a few words of caution.
900 Mhz is a very EXPENSIVE band for antennas and there are only two types of antennas.
Directional, highest gain equals narrowest beam angle
Omni, lower gain equal radiation everywhere.
Now this is all well and good, but frankly you’re operating in the dark unless you install a decent ‘Radio Propagation" software. The best I used to use is named "radio mobile’ made by a French Gentleman for Windows, but runs perfectly on Linux under WINE.
You need the Australian Space Shuttle Mapping Mission data, (freely downloadable) but not only will RM provide detailed color contoured maps of your location and surrounds, it will also provide an accurate estimate of the Receive Level you can expect given:
EIRP
Soil Type
antenna height
receiver minimum gain level
You can also do a directional analysis between two points of known GPS values.
I used this software every day to design radio links that I sold over 14 years, from a few hundred metres to 30 km, and its vital given your radio hobby, to know it well, understanding all its features.
There is another one named “SPLAT” that’s Linux only, great for P-P designs, but I don’t recall using it for Omni base stations.
I recommend holding off buying any 900Mhz antennas until you all have mastered RM and know the actual facts of your topological maps first. That $300 8bB gain 900 Mhz whip antenna, plus the TV installers fitting fees won’t result in a system that can go thru hills
But RM will simulate a $100,000, 30 metre tower in your backyard and it won’t cost you a cent
No doubt the “big guns” in the area are looking at NEC2 simulations and seriously calculating topography and propagation before investing in their sites, but the Meshtastic Site Planner does a great job of simulating propagation for the low cost Meshtastic units we’re tinkering with. Feeding the basics (lat long, height AGL, changing to 915.125MHz - frequency slot 45) and leaving the rest as defaults will get most the way there for our purposes.
From that point fiddling with the antenna gain value (somewhere between 0-5) will give a reasonable indication of how much of an impact antenna would make, and that might help to add to one’s thoughts about whether it’s worth one’s while to spend $100 on a 5dBi antenna vs. $20 on a 2dBi antenna vs. running with the default one in the kit.
It’d definitely be an interesting academic exercise to go in depth down the simulation rabbit hole to compare to real world results once I’ve figured out how much time and money I want to invest in the Meshtastic side of the homelab!
Hey @Belfry , when we get these new gadgets, do you want to have a special Meshtastic Jitsi meeting to work on this subject specificallty? Of course all will be welcome, but we’d be trying to focus specifically on establishing a reliable communication channel with Meshtastic devices.
I’m thinking that a meeting like this would be a third kind of gathering that we’d publish on our calendar. In addition to our in-person and online get togethers, we could have topic specific Jitsi meetings.
For now i reckon Meshtastic is a subject that we could work on. We didn’t make it to discussing my progress with Tube Archivist, but if @zeeclor wanted to have a session working through getting that up and running, we could book a meeting working that together.
I had a similar but slightly different thought bubble recently.
I think the third type of meeting - a topic specific one - is a great idea.
Regarding Meshtastic, I was wondering whether it’d be worthwhile to get everyone in the same room (e.g., Chermside or Carindale) with their devices and laptops. For those without any Meshtastic equipment it’d be good to give an intro, go through the flashing and setup process, and talk about various hardware options. For those with Meshtastic gear, we can finally eliminate the combination of distance + low signal levels, confirm our configs are working, actually exchange some messages, pull telemetry from each other’s devices, and maybe send some messages over a private channel we set up on the night. It’s possibly an opportunity to set up a non-Meshtastic SDR or passive decoder (powered by the mighty RTL2832, of course) to watch the packets flying around in the background as well.
I wouldn’t use the Meshtastic map for a precise coverage prediction any more than I’d use the Telstra/Optus/Vodafone ones for 4G/5G phone coverage purposes. It does however give a reasonable “vibe” of where the strong signals are. In my case, doing a “quick and dirty” run using my lat long at home, and known elevations does produce something that passes the pub test - “very big coverage to the NE, okay coverage in an arc through E and S, and quite poor coverage to the West” which is exactly what I experience with the other 50MHz+ radio based adventures here.
The coverage maps generated for my location with a bit of tweaking for height AGL (in my case antenna in the yard vs. antenna inside vs. antenna on the second story) and antenna gain (have bought a couple of new antennas to test which I’m hoping arrive next week) show that I’m unlikely to ever get a direct connection to @jdownie, but I may get one occasionally to @zeeclor with a bit of luck. What the Site Planner does show is that a few potential configurations are going to be “better” (without necessarily precisely measuring what “better” is), and that in some cases I’ll have a pretty strong signal up towards Ocean View or down towards Tanah Merah. There are a couple of very strong (comparatively) sites in those areas, and by being able to reliably TX/RX to them directly, I should be able to reduce the hop count of packets and there’s a decent chance that other users have a single hop to those sites themselves. I think this alone will dramatically increase reliability. Site Planner has been a simple way to play with “is it better on this side of the house or on that side of the house” and so on, rather than being used to generate a 100% accurate coverage map.
Like @Belfry I have signal to 30 to 220 degrees at least for a few kilometres. However I am never likely to reach catnip or 96b0. I can “see” parts of Mt Cotton but I am sure I am still on the wrong side of the hill.
Weirdly I have Dave HQ and David at Tanah Merah only one hop away. MQTT is officially turned off now but perhaps this is residual data from when it was on.
After running the radio propagation software and establishing that comms is possible with your target using antenna gain of 11dB at the estimated height of your roof or mast, one would then purchase a 11dB gain Yagi Antenna, mount the receiver at the antenna, aim it at the target and power it remotely. The transmitter output would then be reduced so that when the antenna gain of 11dB is added to it, the EIRP is within legal limits.
Although one can’t exceed the transmit power level, or go outside the legal frequency for the use, there is no limit on antenna gain, which can increase the actual ‘link budget’ significantly.
I picked up a Seeed Studio SenseCAP Indicator yesterday, and have had some brief time to play with it this evening. Coincidently the MediumFast channel was quite active with some other users experimenting elsewhere in Brisbane, so that was a great time for me to be setting up my devices while other nodes were active.
I also bought two decent looking antennas and have been trying them out tonight. I’ll do a proper write up when I get a chance in the following weeks. For this evening, the T-Deck got changed to CLIENT_MUTE role (i.e., it won’t rebroadcast packets), and the SenseCAP Indicator was left as CLIENT. I was reliably able to send/receive messages from either device, both using the Android app, two separate instances of MeshDash, or the official Meshtastic interface. I think the SenseCAP Indicator base station setup with some slightly better antennas attached made all the difference.
Seem to have hit a weird snag now where any received message causes either device to immediately reboot. Perhaps there was a corrupt message somewhere within the ones back and forth, or one that contained a character that couldn’t be decoded properly. Perhaps both devices got fed up with me regularly changing from the Android client to the web client to MeshDash while I was experimenting. Who knows. Will dig a bit deeper on that when I get a chance.
For now, the combination of the base in CLIENT and the handheld set to CLIENT_MUTE role was excellent, the configs for talking to the rest of Brisbane are correct, and a private conversation was possible between the two devices as well. I’ll also briefly add that the SenseCAP Indicator is very slick and a well built product overall.
I’ll resume tinkering later next week and when I get some free time I will write up a bit more about the Indicator, the antennas, and the potential issues I run into. If I get a chance to sort the boot loop issue out, I’ll send a brief PM via Meshtastic to the known users here to see if we can get two-way comms going.
).
