Advice on sheep GPS beacon

[haurs]

It seems we got a few different opinions on the CC131 vs Lora among the people here. I haven't used any of them, so I really can't tell who's better than the other, but if we look at for example LoraOne (https://www.kickstarter.com/projects/sodaq/loraone-the-lora-iot-development-board) you seem to get a lot your your buck at maybe $75. Throw in a good gateway and you should be good to go.

LoRaWan might be a great standard - but their business polices are "If you have money and donate $50k a year" you can vote and participate in the big things.... You can get doc's for free but everything else starts at $3k / year. I find these "extortion" scheme "standards" a pain in the backside.

Now I'm not sure what all the talk about costing a lot, but maybe we're talking about different things / services (LoRa vs LoRaWeb)?

[NivagSwerdna]

Receiving GPS is a power hungry occupation, and transmitting position even more so.  I can see that having dots wandering around on a virtual map is quite appealing but it's not the only way to find your sheep.  A radio transmitter that sends a very short pulse every now and again can be subsequently located using radio direction finding techniques.

Just an idea.

[haurs]

Receiving GPS is a power hungry occupation, and transmitting position even more so.  I can see that having dots wandering around on a virtual map is quite appealing but it's not the only way to find your sheep.  A radio transmitter that sends a very short pulse every now and again can be subsequently located using radio direction finding techniques.

Just an idea.

You're probably right. GPS is power hungry, but what other options are there other than carrying a huge antenna in your backpack? I'm not good at radio direction finding techniques, but if you can give some example how well this could work (with good distance) then I'm very much open to considering it.

[Spikee]

Receiving GPS is a power hungry occupation, and transmitting position even more so.

Not really as the gps is only on for 60 seconds or so until lock.Since the amount of data transmitted is limited the rf does not that up that much energy either. Since this is probably open ground, high elevation, the lock should be quick. -> detect that it is a bad place to get gps lock -> shut it off and try again in x time or if significant movement has been detected.

Receiving GPS is a power hungry occupation, and transmitting position even more so.

Not really as the gps is only on for 60 seconds or so until lock.
Since the amount of data transmitted is limited the rf does not that up that much energy either.

How to save energy:
1. Shut off any devices you do not need at the moment
2. Send as little data as possible
3. Use long sending intervals
4. Shut off the mcu (or use one of the sleep modes) when it is doing nothing. Having a low power core also helps

How to get good range:
1. Use the RF frequency best suited for your application
2. Good antenna
3. Good sensitivity
4. Power


Also the CC1310 had and has quite some TI library issues that made the device kinda unusable for a lot of applications.
After complaining for a month or two they finally fixed the major RF issues.  Apparently I was the only one really doing something with the device at that time.
The sensor controller documentation is shit, library documentation is pretty shit (functions that "exist" do not work for example, not clearly documented or shown in errata). Took quite some hours to figure it out.

After spending way to many hours figuring out this shit it works now, biggest issues are fixed and it works pretty good now.
The average current consumption is amazing.

[LaserSteve]

For those who are curious, the attached picture is  the basic classic method on VHF. Range would be around 300 meters with a good YAGI antenna at the receiver.

Here is a chart from a commercial animal tracking outfit, battery size vs GPS logging rate, vs. useful time. This is just position logging, they get close to the animal using VHF tracking and download the GPS data from the logger. From ATS at ATSTRACK.com, typical GPS logging rates.  They must have a phenomenal GPS receiver, in terms of current consumption.


Battery Size:
                10 mAh  100 mAh  200 mAh 
1 second:   41 minutes  7 hours  14 hours 
1 minute:     36 hours  16 days  32 days 
1 Hour :      37 days  379 days  759 days 

ZebraNet serves as an interesting example.

http://www.princeton.edu/~mrm/asplos-x_annot.pdf

At the power level the 1990s technology TREK radio consumes, they could do much, much, better on HF.

I ran the topic thru the academic search engine at work, and came up with this:

http://www.grazinganimalsproject.org.uk/stock_management.html?publication=1;101

If a HF frequency could be allocated, I smell a market...
 
Steve.

[haurs]

For those who are curious, the attached picture is  the basic classic method on VHF. Range would be around 300 meters with a good YAGI antenna at the receiver.

Here is a chart from a commercial animal tracking outfit, battery size vs GPS logging rate, vs. useful time. This is just position logging, they get close to the animal using VHF tracking and download the GPS data from the logger. From ATS at ATSTRACK.com, typical GPS logging rates.  They must have a phenomenal GPS receiver, in terms of current consumption.


Battery Size:
                10 mAh  100 mAh  200 mAh 
1 second:   41 minutes  7 hours  14 hours 
1 minute:     36 hours  16 days  32 days 
1 Hour :      37 days  379 days  759 days 

ZebraNet serves as an interesting example.

http://www.princeton.edu/~mrm/asplos-x_annot.pdf

At the power level the 1990s technology TREK radio consumes, they could do much, much, better on HF.

I ran the topic thru the academic search engine at work, and came up with this:

http://www.grazinganimalsproject.org.uk/stock_management.html?publication=1;101

If a HF frequency could be allocated, I smell a market...
 
Steve.

Thanks for the the interesting read Steve. This is good old traditional tracking

However the range of 300 meters is a bit too low in this project. The sheep is wearing bells as well, so we could hear them from around that distance also. But thats why I'm looking for a long range solution here, and thats also why I've been thinking about the TI or Lora, with the long range of up to 10km.

[haurs]

It seems the preferred choice here is a long range system like TI or LoRa for example. But why don't we discuss the differences with these 2?

What advantages does TI have over LoRa - and vise versa? (for example the kickstarter project LoRaOne: https://www.kickstarter.com/projects/sodaq/loraone-the-lora-iot-development-board)

[Koen]

TI low power wireless chips use proven technology, are well known, are cheap, have an ecosystem of chips/baluns/devices designed by other manufacturers for it and are compatible with other manufacturers offering.


LoRa is expensive, tied to a single manufacturer and heavily marketed with less than useful use cases. OMG so much range* !!

*: 18 bits per second, 100mW, mast high, open field, no radio interference whatsoever, operator dancing on one leg, rainbow over the hills.

On a more serious tone, check out openLora.com for real users experience. It's less glorious.

[botcrusher]

Perhaps you should use something similar to our helium filled bud over here. Getting WSPR data over the ocean! Balloon madness
two or three small pv panels could account for a collar rotating.

Now, sheep don't float 10000km up, but your drone at a much lower level still allows for a large line of sight area.

[LaserSteve]

Friis Calculation for CC1310 at 868 Mhz:

Two dipoles at 2.1 dBi, 10 dBm TX power level, 10 Kilometers range:  -97.0 dBm at the receiver.  Ti claims -110 dBm RX sensitivity. 
Going to "boost mode" nets 4 dBM more at the RX, for -93 dBm

Friis in this case is assuming perfect free space conditions of course. The TI sensitivity figure -110 dBm is a bench test with no thermal noise and perfect RF matching,  and I note there is no noise figure spec for the front end in the data sheet, at least that I can find.

Wiki claims the average height of an Adult Ewe to be 70 cm at the withers.  Running a UHF line of site calculator I get 3 Km line of site for the Ewe in perfect terrain.  With the Human holding the antenna on a stick at 3 meters, I get 10 Km LOS  for the human in perfect terrain.  If the receive antenna is on a portable tower, and was a gain antenna such as a YAGI, or stacked dipoles, you'd do much better. At 6 meters for the RX antenna height, you start to have a 10 kM receive bubble.

I like to use a rule of 50% degradation at UHF/VHF in rough terrain, so  a 1.5 kM range bubble for the sheep if the receiver is near ground.  The limiting factor being the sheep's height.

HF near field propagation starts looking really, really, good at this point. Unless you can build a tower for the receiver, or get the antenna onto a flying platform.  Reducing frequency to 150 Mhz for example, results in a much better -65 dBM at the receiver. 

All of the above assumes  very good VSWR match  at both antennas.

This also makes a strong point for development of a really good antenna pattern  on the sheep's collar.  Boosting Sheep TX power to a modest 23 dBM really would help with the signal to noise ratio at the receiver.   Your operating in a region where antenna gain at the sheep does not buy you much increase in range, but where modest power increases help your bit error rate  immensely.  The modest power increase gives you a chance of some reflected paths aiding in rough terrain, but you do not want to increase TX power to the point you start getting severe multipath propagation.

Friis calculator:

https://www.pasternack.com/t-calculator-friis.aspx

The classical Java range calculator, which uses simplified  equations worked out by two way radio companies in the 1960s...

http://www.hamuniverse.com/lineofsightcalculator.html

There are more sophisticated freeware radio propagation programs out there that use actual terrain data, but the first order approximations here will hold if there are not tall rocks or severe vegetation.

http://www.nautel.com/wp-content/uploads/2013/05/Radio-Coverage-Tool-NAB-2013.pdf

The ever so useful chart from MiniCircuits:

http://www.minicircuits.com/pages/pdfs/dg03-110.pdf

Steve

[Marco]

LoRa is expensive

Judging by the cost of the low cost modules available they can't be that expensive. Of course Semtech being the only supplier (I assume Microchip just packages their chip in a module) would be scary for a real product with long term support, but less relevant here.

On a more serious tone, check out openLora.com for real users experience. It's less glorious.

Okay, I did. What am I supposed to be seeing? I think I see excellent LOS range, lousy range with chip/PCB antennas and 1-2 km range in non LOS urban environments with good antennas. Not much different with TI.

Friis Calculation for CC1310 at 868 Mhz

Shouldn't he be designing for 433 MHz? Europe and all ...

[Koen]

LoRa is expensive

Judging by the cost of the low cost modules available they can't be that expensive. Of course Semtech being the only supplier (I assume Microchip just packages their chip in a module) would be scary for a real product with long term support, but less relevant here.

One Semtech transceiver needing an extra MCU is the same price as one TI transceiver+mcu combination and more than one Silabs transceiver + MCU combination. If he needs ready-made modules, there's plenty of cheap chinese modules. If he wants to build it himself, there's plenty of baluns to help him avoid RF traps.

On a more serious tone, check out openLora.com for real users experience. It's less glorious.

Okay, I did. What am I supposed to be seeing? I think I see excellent LOS range, lousy range with chip/PCB antennas and 1-2 km range in non LOS urban environments with good antennas. Not much different with TI.

Yeah that's my point, it isn't any more "Long Range" than the classic chips.

[Koen]

Friis Calculation for CC1310 at 868 Mhz

Shouldn't he be designing for 433 MHz? Europe and all ...

That would limit him to 10mW. 169 MHz would be best provided he can accommodate an efficient antenna on the sheep.

[Macbeth]

(Sheep farming background as a kid)

I guess now you're no longer a kid you're just an old goat? 

^{...Ok, I'll get me coat...sheepskin at that...}

[tautech]

(Sheep farming background as a kid)

I guess now you're no longer a kid you're just an old goat? 

^{...Ok, I'll get me coat...sheepskin at that...}

At times a grumpy old goat too.
Na, laughter is good and required each day. 

[Apollyon25_]

We had a Lora point to point link running ~12km on 433MHz with small helical antenna on the device and a 1/4 wave on the main base station.

[John Heath]

I have an out of the box idea for this. The biggest problem is cost per sheep. That means the sheep transmitter must be cheap. The one grand receiver can cost a few bucks but not the sheep transmitter. A crystal oscillator and ripple counter that produces a 1 minute pulse every 24 hours at precisely 12 midnight. That should be cheap at cost per sheep with CMOS gates , lithium battery and 1000 MCD LED. This should run for 3 or 4 months on cheap button type batteries.  One minute per day lighting one LED for 1 minute a day is not a lot of power. All that is required at this point is a 100 dollar drone to fly over the land and take a picture from 1000 feet up at midnight. With even the cheapest crystal of 10 PPM the sheep crystals clocks should still be within 1 minute after 3 months for the 12 midnight 1 minute LED turn on. The LEDs can not flash all night long as predators would soon realize lit LED means free lunch.  This way you will have a ball park idea where most of your sheep are to find them.  One LED on your house plus another 100 feet away dead north will allow the drone picture to be superimposed on a map for sheep location within 10 feet or so.

[botcrusher]

I kinda like that.
What about software that uses drone camera to search for white blobs?
Sheep recognition could work...
though, it relies on the drone to actually find them.

[Fungus]

.....This should run for 3 or 4 months on cheap button type batteries.

Question: Does this actually have to run for three or four months?

Can it just sleep for three months and then activate? That way you can use a lot more battery power for the transmitter because it only has to last a week or so. It could make the whole thing a lot cheaper and more effective.

[Pjotr]

You have to do the math: 3 months = 100 days = 2400 hours. With a 5Ah battery you can draw app. 2 mA average. Transmitter power is less of a concern because it is used shortly and infrequent. Although you have to calculate its Ah budget. More of a concern is the GPS receiver. Most draw a lot more to keep position up to date. Although there are ones that only keep the map data and time up to date to do a warm/hot start for accurate tracking. Have seen GPS modules that draw only 0.5 mA average in that state.

Also a link receiver substantial draws current in this case.. But you have very accurate GPS time on board, so a receiver isn't strictly necessary. Just use a defined time slot for each sheep to transmit its position. Although a 2-way link is much more reliable. But even then you can make use of time slots to scan each sheep.