Laird Sentrius RS1xx Series

This is correct: Adaptive Data Rate | The Things Stack for LoRaWAN

It’s good to know, but part of the specification of the LoRaWAN is about leaving it to the server to find the best rate.

From your perspective, there’s no dial to move to increase power to increase range - not without potentially acquiring matching silver bracelets & a nice young person in blue to supervise you.

2 Likes

Thats part of the info you need, you then need to know what the legal power limit is for for your part of the World.

America ?

1 Like

As I said in the first post, US 915 region. I am in Canada - Ontario, not far from Toronto.

Oh man. The Spec is tedious reading, as specs are. Still, the concepts and intent is clear but making my brain hurt!

Ok. I get it now, unlike my previous experience with “simple” RF systems where the Tx power is directly proportional to range, this system is far more nuanced. Indeed, as you said, it tends to use full Tx power all the time. Using the spectrum spread and bit-rates to extend range. Although, under certain special circumstances the system may determine the Tx power should be reduced that is rare. And certainly no likely to affect me.

So now I want to figure out how long it takes to “fix” a signal loss problem. And I’m doing some practical tests starting at a closer range. Today, I put the Laird sensor in my car, and parked in the far end of the plaza across the street. About 500m.

In the office, I was getting:
“rssi”: -42,
“channel_rssi”: -42,
“snr”: 13.5,
“frequency_offset”: “-1929”,
“bandwidth”: 125000,
“spreading_factor”: 7,
“coding_rate”: “4/5”

And I have more reading to do before I understand all of these numbers’ meanings.

500 m away and I get:
“rssi”: -96,
“channel_rssi”: -96,
“snr”: 5,
“frequency_offset”: “1505”,
“bandwidth”: 125000,
“spreading_factor”: 7,
“coding_rate”: “4/5”

I see there is two sets of numbers in the event details. I think the gateway is reporting something slightly different. Same RSSIs, but they mention a “frequency offset” and “channel index”? like:

      "rssi": -96,
      "channel_rssi": -96,
      "snr": 5,
      "frequency_offset": "1505",
      "channel_index": 4,

sigh. more reading…

Not sure what you are now trying to achieve? Those numbers look about and what might be expected depending on RF & Built environment around you etc…

What problem are you trying to solve now? Looks like you have confirmed the broad outline I gave you so unless you are trying to ‘break’ something…? :wink:

1 Like

how far can I go?

That is a bit of a how long is a piece of string question - with the technology the record is >800km for general terrestial - using a HAB! Best I have personally see (but only when deliberately looking/evaluating rather some random connections which may have been much better but never logged/reviewed) ground to ground using a reasonable mast for GW and node not sat on/in the ground is aroun 70-75km. Generally for the technology we advise think 300-<3km in dense urban - depending on built environment, urban canyons etc. 2.5-10km suburban, depending, and 5->25km in reasonably flat rural, with 30k+ a real ‘possible’ and often seen…mostly it depends on the placement of the GW antenna (generally higher = better) Per earlier

Best way to find out is war driving! :slight_smile: and will then also depend on if other GWs in the area provide coverage :wink:

Just focus on the RSSI & SNR ignore the rest, note as you drive away and signal gets lost or if you move very slowly such that the LNS is able to ‘track’ the declining signal quality and instruct the node to compensate via ADR you should see gradual increase in SF etc. (Note ADR isnt recommended for devices that move frequently or rapidly e.g. trackers). It typically takes many messages for the system to adapt ADR values and alternatvely identify a lost link and try to re-initiate…so be patient, it can take days depending on update rate…

Best way to begin this process is look at the topology (topographic/terrain maps or your local equivalent of the UK’s OS Maps may help) start as far away as you reasonably think you can get true LOS to the GW and see if you get connection and bring it inwards from there until you do… the alternative is to sign up for one of the RF Signal prediction packages…

Tell me where node/home and GW are and I can have a quick peep at likely hood of connection, when I get a few moments. If you are concerned about privacy dither you home location by say 50 to a few hundred meters - wont make much difference in context of 7km unless you run into local terrain masking problems…house in a valley when in reality it on a hill visible from GW or say its on a hill when in reality you are in a valley! etc.

2 Likes

So what is the legal power limit at that frequency, in that part of Canada ?

1 Like

Almost immediately if you know LoRaWAN at level Theta-Omega but you are at Zed-Zed-Nine-Plural-Zed-Alpha.

There is no real concept of signal loss as there is no handshake, LoRaWAN largely works on a broadcast & hope basis. But as outlined above, the LNS will manage the device transmission parameters to get the best results and adjust if there are changes. And the device with an appropriately compliant / up-to-date LoRaWAN MAC will also check it is being heard and make changes to regain connectivity if it thinks it isn’t being heard.

So unless you are operating at Delta-Tau-Chi level, best let the well formed, well tested spec do it’s job.

But not for a device that is moving. That’s an outlier for which other configurations are suitable. Unless you are driving, then it’s moving sufficiently fast that other configurations are suitable.

And then there is the whole other world of mapping, which is like a sub-culture in it’s own right. Very very simple rule of thumb, take a rolled up newspaper, stand where the device is, look through the tube, if you can see the gateway antenna, you are probably good to go. If you can’t see it but you are pointing the tube up or down at 30° or more, ie the device is significantly higher or the gateway antenna is significantly higher, then you may be in luck. To understand that summary further, search for fresnel zone.

2 Likes

I have two, related goals. I need to write software to run an offline (no internet) system for commercial use.

And I have a “cousin” bee keeper on a tight budget who wants to keep watch on his 4 hives over winter. I am planning to gift him whatever hardware he needs.

My system will be installed in factories, usually the power/drive rooms. Lots of metal, grounded metal, high voltages, high flux fields, etc.

So I am trying to learn the range ropes in both scenarios. In my cousins case, what antenna(s) will be needed to make it work. Or is there a handy public node near his hives? or can I create one?

My 1.2 Km test failed - but I only had 15 minutes to spare last night. From what I read (somewhere) the end-device will request replies after "x’ uploads. And after “y” failed replies will start its own fall back logic. I have to look up X and Y for my region or devices. But I’m pretty sure 15 mins wasnt enough.

When I got back to my office with the devices, the first uploads had these RSSIs and settings:

Laird:
“rssi”: -76,
“channel_rssi”: -76,
“snr”: 12.8,
“frequency_offset”: “2604”,
“bandwidth”: 125000,
“spreading_factor”: 10,
“coding_rate”: “4/5”

Dragino:
“rssi”: -78,
“channel_rssi”: -78,
“snr”: 14,
“frequency_offset”: “-3889”,
“bandwidth”: 125000,
“spreading_factor”: 7,
“coding_rate”: “4/5”

Nice to see the Laird had at least taken some action.

I read someplace that there are ADR settings in the cloud server, and possibly in the end-devices. I have to go dig out what I have and their settings. Maybe the Dragino has something disabled I need.

I also read something about being able to “force” specific settings from the downlink. That might be a handy way to set up new end devices: program in the office, set the radio for longest distance, then deploy. After deployment the ADR can figure out the best settings. But this way, it will report data right away - or not. Which well alert me I have a range problem at the location where the sensor is going. And I can “do something”. Maybe a larger antenna on the end device? Or upgrade the gateway antenna. I have not even looked at antenna options for either yet.

‘Upgrading’ the gateway antenna might be illegal, part of the reason I asked you what the legal power limit was in your part of the World (no answer).

1 Like

You are trying too hard and over thinking things - trying to assimilate the whole thing without having any waypoints of success.

There is no substitute for the core knowledge - well, there is, but it’s a long slow boring process with bloodshed & tears. The clever ones read the basics, get the basics working and then complicate things.

So read: https://www.thethingsnetwork.org/docs/lorawan/ and watch the “Everything you need to know in 60 minutes” video in small chunks on YouTube where the first comment has the links for each section - there’s a whole section on free space path losses & link budget that will fill in some of your knowledge regarding range.

Plus the getting started section in the main docs: https://www.thethingsindustries.com/docs/getting-started/

But for now, just get the basics done - a gateway, a device, working for a period without diving in to make their & your own lives difficult.

PS, antennas as explained above, gateways don’t have settings relevant to your thinking, if you are forcing things, you need to know the basics so you know why you need to do it. As to your cousin’s hive, there is indeed at least three public gateways nearby but if I got that wrong based on the location you provided, you can always host a gateway which will be public if it’s on TTN. It’s not a node - a node is a device. And there are no uploads, uplinks and were they joins ie what kind of settings are you using for all this proactive range testing?

2 Likes

@LoRaTracker I have not gotten around to that, but I really don’t care. I have no intention to violate the limits - whatever they are for my area.

I will use only approved antennas, and the system will only do what is allowed. That should be enough.

@Jeff-UK

The hives are near here:
43.991657292621156, -79.62338820108134

Cousin lives on the south edge of the large city North-East: Bradford East Gwilimbury.

Looks like mostly trees between. No large buildings. Might be a radio tower. There is a very busy 4-lane highway (the 400).

Just over 10 Km line of sight distance.

WRONG!

There is no concept of approved antennas. Just levels of quality and its impact on the signal. If you put a “larger” antenna on, you HAVE to dial back the gateway power accordingly. So you need to know what the cable & antenna you use will do to the radiated output.

The people answering &/or asking for details here have experience & background knowledge that takes years to acquire. Feel free to ignore them, but please do not dismiss them - you may think you don’t care, but the federal authorities certainly do - and they have police cars and courts and fines and stuff to keep you entertained whilst you have that Road to Damascus moment and start to care …

1 Like

LoS to where?

There are tools to check actual LoS but it needs the other end …

1 Like

The closest gateway is 3133303729007200 some 14.7Km away

Then trilliant-multitech-conduit-ip67 some 21Km

Then uoft-nrl-gateway some 41.1Km

Then matt-ttig some 42km

The dark circle is 15Km radius
image

Unless there are gateways in the area that the location is private, I don’t see you have much chance of using a community gateway unless you add one.

@descartes Thanks for all the input, I really appreciate it.

I have to catch up your posts.

First, that one you linked to the intro and mentioned the video. I just finished the video - excellent stuff. My only complaint about videos is they take too long to communicate knowledge. I actually find I can gleam data faster from reference books and standards docs. And of course posting like these.

Most important thing in that video is the slide on expected range. They say to expect 500m range for indoor GWs. My test GW is sitting on a desk in my office. Similarly, my first steps was to get that gateway and 2 COTS sensors working for about a week. Only after that did I start to experiment with different scenarios. First I locked the Dragino into a 600V 3-phase power cabinet, just a little more RSSI. Then I put both sensors in my car in the parking lot out front, again not much change. Then I drove home for the weekend 7Km, that did not work. Then I parked my car across the street. That was just over 500m, with only a few trees in the way that dropped the RSSI as listed in earlier posts. My next test at 1.2Km failed. And I think that info graphic explains why: indoor GW.

Second, I am not ignoring anybody or anything. Nor am I dismissing anything. On the contrary, I have to absorb it all in so I can build a model of this technology in my mind. Then I will understand it enough to figure out complex things. Its all baby steps for now. By “approved” antenna, I am thinking of a COTS device GW that comes with a mast antenna, or is designed for mast mount. If I am forced to replace a GW or end-device antenna, I will certainly read the tech specs and adjust the radio settings to get best performance and stay within limits. Not like when I was a kid and we made RF boosters for our CB radios…

Third, I posted the lat/long of the bee hives earlier. Clearly, the only way my cousin will be able to get a signal form his hives is if he puts up a tall mast at home. And even then, there is a ridge line between him and the sensors. I have to look up the elevations to see how tall his mast would have to be. There is actually a great site for TV antenna tower calculations that is handy for that.

Lastly, nobody is ignored. All input is greatly appreciated.

I would like to force at least one end device to the lowest data rate for a test from home this weekend. But I fear its a moot point with my indoor GW.

@Johan_Scheepers Thanks!
What map is that?

I was just looking at https://www.thethingsnetwork.org/country/canada/

Which just shows one in Markham. And another in Vaughn. Noting further north.

7Km for a indoor gateway is ambitious, no matter the SF. Unless you have LOST through a window and some luck.