Hi,
I’m new to lora and I’m having a hard time deciding what module and board combination to go with. Stock theft is a major issue in our area. I would like to make use of lora to track (or at least just calculate distance of node to gateway) as many cattle as feasibly possible. For that reason my combination needs to reach distances of ±8kms, needs to be as cheap as possible, needs to have a low power consumption and needs to be as small as possible. I am aware of the existing tracker products (as i do make use of them) but they have their down sides.
It is not feasible to have more than 2 tracker collars per 100 cattle
It is often hit and miss with the GSM units due to poor signal
The monthly sim fees
I am rather fond of the pro mini but i believe there are smaller and cheaper combinations out there. I read earlier discussions on a similar topic and here is what I found. ESP32+LoRa, 32u4, DRF1276, RN2483, most of which were not available locally but aside for that i found them all to be a bit pricey.
The best combination that I came across was the RA-08H-KIT (also not available locally!), I’m not sure if anyone has had any experience with them?
The following MCU’s are available locally,
Seeeduino SAMD21 Cortex M0 -7$
Beetle esp32 - 5.5$
D1 mini wemos - 5.4$
attiny85 board - 5$ (even if it is compatible I’m sure it will be a headache)
The following loRA modules are available locally in the 868 range,
sx1278
RFM98W
RA-08H
All LoRa modules use the same Semtech RF chips and settings so the performance of all those modules will be the same.
The only feasible way to do that is with GPS trackers and GPS trackers are not a good fit for TTN applications.
There are high end Gateways that can do location by accurate timestamping of receipt of packets but you would need 3 of them widely spaced for that to work.
There are no magic methods of ‘calculating’ how far a node is from a gateway.
8km range is possible but you would only get that range in a flat area free of obstructions such as buildings, trees, hills etc.
Hi Stuart,
Thanks for your reply. I actually came across your site a week ago, i was fascinated by the $50SAT project that you were involved in. What an achievement!
Why are GPS trackers not a good fit for TTN?
Based on your reply I’ve decided to go with the Seeeduino and RA-08H modules. As I am still learning I will start with only two complete combinations and then work my way up from there once I am more knowledgeable in the lora field.
The question that I have is what other components other than common resistors, capacitors, voltage regulators will I need to get started with a simple bi-direction communication prototype. Im quite far out of town so purchasing components that I need as I go along is not an option.
What I understand from the diagram above is that I will need
C32 1x10uF cap
C35 1x0.1uF cap
110 OHM res
Im not sure what NC is at C33 and C34 at the antenna end?
Im also not sure what ESD should be used at D3 for electrostatic prevention?
Thanks
Yep, it was fun, 10 year launch aniversary coming soon.
Because mostly people want to keep a timely track of where stuff is. You have already said that your likely to be operating at the limit of reception distance, which given the fair use limits of TTN limits you to around 20 location packets a day from each node.
However given the requirement for an 8km range from nodes that may be close to the ground I would suggest you carry out some basic range checks before designing nodes etc.
If you need to cover 8km and a range check at your location suggests you cannot then there is not much you can do to ‘improve’ the range apart from putting gateway antennas at higher altitudes or adding more and closer gateways.
I dont know that module, but be sure to check the development environment is Ok and that there is supportecd code for TTN.
Just to clarify, we do LoRaWAN on TTN here. As for deciding on device hardware, getting to grips with all the moving parts makes this totally irrelevant in the early stage, working with something that is know to “mostly just work” whilst you figure out the code, range & gateway placement will serve you well. Getting in to the details of which capacitor where or pricing isn’t really a thing until you get down to designing a PCB - and even then the vast majority of the work is done in the reference design which, given it’s a radio system, you leave bits out at your peril.
It would be useful if you could expand on your mechatronics skill set - it seems you know about components, how are you on coding in C (preferably) or Python (tends to run on power hungry devices)? There are various implementations using C, some more friendly than others. At present the full-fat C spaghetti that is the reference code of LoRaMac-node runs on the wallet friendly all in one STM32WL chipset - the Seeed LoRa-E5 SIP needs almost no supporting components and costs around $£€10.
As well as very expensive geolocation gateways, there is also the LR1110 chipset that uses normal gateways but can do geolocation server side.
The Ra-08H has an MCU on it already but the LoRaWAN stack is very old - if you use a Seeeduino I’d drive a RFM95, so you may as well get an Adafruit Feather M0 with RFM95. The SX1278 is a reference to the 433MHz chipset which is used on the RFM98W which consequently is 433MHz. Gateways at this frequency are few & far between and I don’t believe there is a channel plan in SA for that either.
For emphasis, we do LoRaWAN on TTN on this TTI funded forum only, not point 2 point comms.
The whole import duties / items going astray thing seems to seriously hold back small businesses in SA - is there no collective attempt to get the government to sort this out. I’m not talking about cheap Chinese electronics here, but being able to confidently receive hardware from overseas suppliers without financial penalty.
On the other hand, as you’ve listed all the cheapest components available locally, what is the actual financial loss that you are trying to prevent? With some smarts on the collars and perhaps some local relays (as discussed on the recent elephant poaching thread), a seemingly expensive collar may well be a better investment.
OK if I’m understanding this correctly a gateway is just simply an antenna. According to The Things Network there are 127 gateways in SA however I can only see 7 on the map. The closest gateway to my farm, according to the map, is 70kms away barricaded by mountains. So to clarify there is no loRaWAN in my area, I didn’t think there was to start off with (not even in CPT for some reason). My apologies I originally thought this forum was for all things lora related I now understand that my questions need to be network related. My plan all along was to use several lora slave nodes that communicate with one lora master node that is on my farm wifi network.
Will I be compromising the range by doing it this way even if there is still line of sight between the nodes?
Thanks those definitely fall in the budget!
Like I mentioned before and if I’m not mistaken the elephant thread had the same issue of there being poor cell service. Then there is the issue of the price, if one collar costs $350 once off and $20 monthly sim fee then it doesn’t make sense to have more than 2 collar per 50 heads and even then they can still steal 48 cattle without any alarms going off.
I don’t need GPS on all my nodes, even if I can still just read the node every now and then then I know that cow is still alive or still on the farm. My final reason for doing this is because i think it will be a fun projects to do . My family is very involved with the anti-poaching of rhinos on several reserves and if I can get something like this to work on my farm then why not in those areas where there is no signal at all.
Thanks
A TTN\LoRaWAN Gateway is a great deal more than just an antenna.
A Gateway is a multi channel LoRa tranciever, with an Internet connection that allows the individual nodes to join TTN and forward packets\information to application servers on the Internet.
There is almost no correlation between my statement and yours. I am talking about using better hardware with better software to create a better solution, not about spending more money with mobile operators or the cell service coverage. Chucking a GSM+GPS solution together is easy so becomes overly attractive but lacks any actual use case design.
I’d hang around to ask if these are diamond encrusted or just the gold plated ones, but if you could excuse me, I’ve an urgent appointment with my workbench, I’ve spotted a huge opportunity, I’ll ring you from Necker Island next week.
@LoRaTracker Understood thank you. No its not a TTN/loRaWan application for now. Is there perhaps a link to an old thread for getting started with TTN/loRaWan I am keen to learn more about the network side of things for future projects that aren’t within my farm network range.
. My family is very involved with the anti-poaching of rhinos on several reserves and if I can get something like this to work on my farm then why not in those areas where there is no signal at all.