So - about 6 messages/hour of SF10-12 (if I can implement proper mixing to avoid banning due to regulations which prohibit 11/12 only) are good under FUP, giving that they will be used like two weeks on / two weeks off. (We have 60 of these sensor boxes in total, FYI)
In that case, I will investigate in other gateways and look into those listed by Jeff. Does the GW itself affect range or is that antenna only? I do want very high sensitivity / range, so if there are certain GWs or combos known for that, I am open to suggestions. The current combo has a tested range of 15~20 km “as the crow flies”.
I will be happy to discard the MultiTech
Another bit of a side question but hopefully useful to be asked on a forum: are there any documents that describe the back-end of a gateway setup? The document that I got from my external source really expects me to know everything inside-out. The only thing I understood is that there is connection to a database with indefinite data retention, and that’s pretty much all I could grasp.
I am looking for something that also might explain the role of NodeRED or other servers, since there are running subscriptions of which I can’t tell the use.
All the COTS indoor units I initially list (apart from the TTIG) ship with an adequate ant - trick is get the GW (are at least the Ant) as high up as possible, give clear line of sight where you can. with extenal ant connection units adding 1, 2 or 3m of feeder cable will help placement and flexibility in placement - use an ultra low loss cable for longer feeds - 1.5-2m of RG58 is probably a limit, thereafter go ultra low loss type. As noted whilst the chipsets have evolved a bit overall RF performance isnt hugely different (later ones drive lower power consumption or add extra SF’s, # of channels or modulations - the latter not yet applicable for TTN deployments, hence I still to proven systems unless power an issue (a couple of W/hr less on a solar panel install can help ;-)!)
Moving to higher gain antennas unless there is a specific application or deployment need is typically a zero sum game - beyond 3.15dbi you have to start dialling back the power of the GW Tx, you start to induce potential coverage notches and directionality and even beam tilt which can hinder perfromance. Save your money for better cable/connectors and feed higher! There is a reason the emergency services stick with smaller, lower gain, ants that are closer to isotropic radiation patterns, and I assume your students are randomly scattered around the colledge and will be mapping over wide ranges vs being clustered in one direction or in a distance band from the college?
As a good discipline I would encourage your students to be mindful of SF impact on airtime and hence impact under FUP and on that precious resource RF spectrum from the very start, tolerance of short term abuse for a few hours run time whilst mapping coverage to the benefit of the community is one thing running for days/week(s) on end is another story. Also if you are permissive at this stage they will learn bad habits & tend to retain and abuse later. Having your own instance of TTS will help avoid FUP but you still have to be mndful of the spectrum impact - a shared resource - and legally enforceable limits. Depending on clutter and land topology and type of node build/ant 5km should be easily achieveable with lower SF’s (7,8, poss 9) helping keep you under the limits…
Nick, you feeling well? Do you need to lie down? This isnt like you!
Also if worried about air time for the GPS location it may be a useful lesson to teach your students some code compression or data reduction methods or simply parse the data be selective of what you send - perhaps send Lat & Long only and drop altitude if projecting/collating onto a 2D map? Perhaps send a full fix every so often with lat & long deltas on a the follow on tx’s (may need a scheme to cover missed messages then mind, if delta vs last tx vs delta vs last full fix etc.)
But all this getting away from Thread topic of getting the Multitech back online (or replacing/augmenting)!
In the context of this Forum there is nothing GW specific as the purpose of a GW on TTN is purely as a transparent message passing machine - acting as a media convertor taking in RF messages and converting them to IP messages sent over the Internet to the Network Server (NS) which is where all the magic happens. It is connections from the TTN NS that you need to concern yourself with - that is where the various Integrations - e.g. webhooks come into play, this can be as simple or as complex as you want it to be and the Things documentation is linked bottom right from every console page. As mentioned earlier Multitech’s can either be packet forwarding engines as used on TTN or can run full NS implementations, and from that early ones also supported eg Node-Red etc. hence the compexity and confusing nature I mentioned earlier if all you want to do is route messages into TTN, and from that the ‘heavy’ documentation for the GW.
You could do worse than set up each student with a Cayenne account, add their device (GPS tracker? T&H Monitor or whatever) to TTN, add webhook to Cayenne dashboard and register the device EUI for the student then they can watch the data come in - historic records of minute, hour, day, week, month simple data or 24 record of GPS tracks (not retained long term) etc. Add TTN Mapper integration (webhook) to GPS devices and you will help build a longer term view of coverage provided in the area by your and others GW’s etc… Its perhaps 5mins max per device per integration for a novice following basic steps and instant satisfaction for your students as they quickly see ‘their’ nodes doing stuff (Note: another use for you if teaching F2F is put a LoRaWAN CO2 monitor in the classroom and watch level change with ventilation/class size/time and help protect the students from potential Covid infections - if CO2 spikes high open more doors/windows! ;-))
I can’t thank you enough for your replies, once again!
The current external antenna has been installed great - on top of the ventilation system on top of the school, and its range has been tested before to >15 kilometer so no worries about that!
Correct
It’s a great physics lesson anyway, but I’ll most definitely kick things off with this
They will only be sending GPS data if an external pushbutton is pressed, as of the current software revision. May change it to once every 24h or so, but I’ll have that covered!
Jokes on you: that part is all figured out and working already if you wish you can snatch a quick look at the (Dutch) project website https://www.meetjeleefomgeving.nl/ and click around - the data compression you mentioned is also quite adequately implemented with CayenneLPP.
@kersing sent me a message and we will likely be working together to get it up and running.
If everything goes to plan, I will be adding more gateways on other schools too, later; so I might still look out for easier GWs in that case.
One more question: adding indefinite data retention could/should be done how? Through something like those webhooks? Or do I need to start elsewhere?
Jokes on you I’m afraid, LPP is verbose boarding on Argggggghhhhhhhhh.
But it is a starter for ten and an extra credit student could figure out how to make it more compact - which is fully documented all over this forum but still requires some knowledge to implement.
Now that is my domain:
Webhook is the most robust as it just needs a web server.
Data Storage is a backup, only holds 24 hours on TTS CE (aka TTN).
To be honest, I have never had so much fun on forums . True though, I have only been looking at sending the values themselves and thought “I definitely could not send these values in less bytes” but never cared about the full package.
The original idea is that the boxes will be used to measure the front yard, back yard or local playground for example for a couple days each, which does not require that many GPS fixes. But that’s something that depends on the usecase.
That is an excellent idea. I have already put in well over 100 hours so far so I stopped going the extra mile
Those are welcome words. If I understand correctly, I can use the PHP file to set as an integration on the TTN application? Will talk to IT department to see if they can allow running it on the home network since we have got plenty of storage on our network
Bonus points for getting a student to run it on a simple Raspberry Pi set up ! You dont need much (if I understand correctly)…and if IT get shirty and wont allow on school network connect the Pi to a WiFi/MiFi 3G/4G dongle with a few £$€/mo cellular data plan and go independent!
Yes, PHP in the repro above, doesn’t even need configuring as it writes to text files.
IT just need to provide a web server end point that you can play with - which they may not be so keen on as it is technically a hole in the firewall but it is only an http request - hardly an ssh login.
If not, run MQTT on a machine in the office - code provided in the repro.
@Jeff-UK does gateways but not integrations, so don’t put it on a dongle, you could burn through all your credit for the duration of the project and then it will stop catching data and the IP address has to be visible to the internet at large - use MQTT on a machine with a big “Do not touch” sign, no screen saver & the console visible so you can see the incoming data. But still bonus points for any student that can put that together on a Pi - should take about an hour.
Completely different perspective: why use your own gateway? I assume that all this takes place in Utrecht. There are more than 80 gateways active there. They just pass on the data from the nodes to TTN. I myself use a raspberry with a RAK 2245 gateway, which is cheap and works fine.
Hi andreas,
I live about 20 kilometers from Utrecht so the gateways in Utrecht itself do not cover our area; but since it is still relatively close, I gave the Utrecht community a shot!
Also, as called out many times on the forum, if you have your own GW and are developing a node there can be advantage in being able to see the GW traffic - locally or in TTN console - to assist troubleshooting and diagnosis…and beside anther GW helps extend coverage and densify the network to the benefit of all! Go for it!
I definitely agree with that! The last few weeks I have been debugging the project using a Kickstarter gateway - its range in the area is quite terrible (I also did not bother to install it properly), but it helps immensely to verify (non-)operation.
I just sent out our ‘real’ GW to the very helpful fellow kersing and let’s hope that we’ll be able to make it function properly again - this time with our own credentials and on TTN v3
Thanks to the great and awesome @kersing the gateway has been reset and is now fully functioning on TTN V3 plus we have the credentials to maintain and update the GW whenever necessary
Next week the gateway will be set up at school and we will finally be able to test the range of our sensor boxes properly. All the help has been amazing, thank you very much!
Don’t be a stranger - I for one would be happy to suggest stretch tasks and resources for students - just ask, at worst I’ll just giggle and say no time, at best, I may already have the materials to hand.
You may also want to hook up with @pe1mew who is also university based.
There is a reason the emergency services stick with smaller, lower gain, ants that are closer to isotropic radiation patterns, and I assume your students are randomly scattered around the colledge and will be mapping over wide ranges vs being clustered in one direction or in a distance band from the college?
. True though, I have only been looking at sending the values themselves and thought “I definitely could not send these values in less bytes” but never cared about the full package.
