EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Rebuilt5240 on February 13, 2023, 03:38:40 am
-
Hi all,
I'm a student with a CS background but I want to setup a monitoring system in a remote area that doesn't have any power or reception. I investigated the potential to use LoRa as the method to transmit the monitoring data but I'm afraid it won't be able to send it the required distance (2km) since there are a lot of trees in the way and I'm not able to reach a high enough area. Instead, I'm thinking of buying a satellite service that I know reaches the area and setting up the modem along with a Raspberry Pi 3B to run on solar. Ideally the system would run 24/7 indefinitely on solar alone.
Below are some of the conclusions I've reached but I wanted someone to verify that I'm going through the right path before I buy anything :) .
I've researched the modem that the company uses and apparently it is the HT2000W which consumes a peak of 60W and 40W on idle. I think the the Raspberry Pi 3B uses a peak of 3.7W and 1.4W on idle. I will be connecting sensors to the Pi so this will probably consume more power, so just to play it safe I would say the Pi will consume 10W peak and 5W on idle.
I've heard that a good rule of thumb is to have a battery capacity of at least 2-3 times your daily energy consumption.
In this case, since it averages on 45W
45W x 24 hours = 1080Wh
1080Wh x 2 = 2160Wh
I would need a battery with a 2160Wh capacity.
For the solar panel, I would need at least 45W output from the panel. This would mean that I should get a 50-60W+ solar panel since some of the power will be lost at the solar panel -> battery -> devices. I've heard that at least 10% of the power is lost when stored in batteries (I guess this would depend on the type of battery?). I'm thinking I should get a better (100W+) solar panel considering sun availability + weather conditions might not be the best. Edit: I've read that I might need a charge controller. That way I can stop the panel from charging the battery at a certain level and maximize its lifespan.
This would all need to be waterproof and I would need an inverter (DC to AC) to connect my 2 devices.
That is what I have for now. All feedback and advice is welcome and appreciated!!! Please go easy on me :-X
Edit2: Also, are there any well known (newbie) kits that would meet the requirements for this type of application?
-
For the solar panel, I would need at least 45W output from the panel. This would mean that I should get a 50-60W+ solar panel since some of the power will be lost at the solar panel -> battery -> devices. I've heard that at least 10% of the power is lost when stored in batteries (I guess this would depend on the type of battery?). I'm thinking I should get a better (100W+) solar panel considering sun availability + weather conditions might not be the best.
Your panel needs to be a lot bigger unless the sun shines 24/7 where you are. If you want it to function through a week of rainy days, you probably need 20X that much solar panel capacity--three full-sized 1m x 1.5m panels. There is a tradeoff between battery capacity and solar panel capacity. If your battery size was nearly infinite and had no losses, you would just need to take 24h divided by your local average daily insolation and multiply by 45. If your battery size can only get you through one night (the minimum) then your solar panel has to be large enough to produce your 1080Wh with whatever your local minimum daily insolation allows.
-
According to this site https://www.gaisma.com (https://www.gaisma.com)
The local minimum daily insolation is 4.19 kWh/m²/day while the average is 5.48 kWh/m²/day.
Am I following what you said correctly?
Using the minimum insolation of 4.19 kWh/m²/day
24h / (4.19 kWh/m²/day) * 45 = 258 (W?)
-
60W is an enormous amount of power for this. And that's to transmit 2km as the crow flies? As a practical matter I think you need to looking at a much lower power setup.
What I read is that LoRa has a range of 3 miles in the city, and 10 miles line of site. Another alternative would be walkie talkies. Can you describe what's available at the nearest reception point - power, internet access?
It just seems to me that you have to use far less power on transmission, and no power at all unless there's something to transmit. What kind of monitoring do you need to do?
-
According to this site https://www.gaisma.com (https://www.gaisma.com)
The local minimum daily insolation is 4.19 kWh/m²/day while the average is 5.48 kWh/m²/day.
Am I following what you said correctly?
Using the minimum insolation of 4.19 kWh/m²/day
24h / (4.19 kWh/m²/day) * 45 = 258 (W?)
You have the right idea, but 'minimum insolation' as shown there is just average daily insolation by month, the minimum being the lowest month. You have to figure that if you have an overcast, rainy day you will get maybe 20% of the average. There isn't any combination that is guaranteed to always work, it is more like as you add capacities the probability your battery running out goes down.
Does the remote operation have to be in real time? In a low-power situation, could you power down the satellite link and just store the information locally until you get power back?
-
Can you describe what's available at the nearest reception point - power, internet access?
That's the nearest point where I get signal to my cellular network. I was planning on having it reach at least this point and use a Telit LE910C4-NF with a SIM to connect to the cellular network and use it to upload the data. But I don't have access to power at this point which would mean a separate power system.
It just seems to me that you have to use far less power on transmission, and no power at all unless there's something to transmit.
Yes, you're correct. Ideally, if I'm going to use the modem setup I would only want it turned on when I'm trying to transmit data (Edit: Thanks, I hadn't considered that as an option).
What kind of monitoring do you need to do?
I'm monitoring the pH and temperature of a body of water. I want to transmit it at most 4 times a day. Ideally, I wanted to see captures of the wildlife (snakes/rats/frogs) whenever a camera detects them in front of it. I want to know right away if an animal is detected and also if the water reaches a certain pH or temperature. But when I was considering LoRa, I saw that the bandwidth is very limited so if I do end up using LoRa (in this case I'm considering the SX1262 LoRa HAT for the Raspberry Pi), I'm opting to only send the text of the animal detected without the picture.
-
According to this site https://www.gaisma.com (https://www.gaisma.com)
The local minimum daily insolation is 4.19 kWh/m²/day while the average is 5.48 kWh/m²/day.
Am I following what you said correctly?
Using the minimum insolation of 4.19 kWh/m²/day
24h / (4.19 kWh/m²/day) * 45 = 258 (W?)
You have the right idea, but 'minimum insolation' as shown there is just average daily insolation by month, the minimum being the lowest month. You have to figure that if you have an overcast, rainy day you will get maybe 20% of the average. There isn't any combination that is guaranteed to always work, it is more like as you add capacities the probability your battery running out goes down.
Does the remote operation have to be in real time? In a low-power situation, could you power down the satellite link and just store the information locally until you get power back?
I could have it powered down and only turn it on when transmission is needed. In that case, it would look something like this right?: I would have the Pi running 24/7 and the modem would be off until the Pi sends a high signal. It would remain on until the signal goes to low again (when it finishes sending the data). The high/low would be programmed into the code used in the Pi.
Could you recommend me or point me towards the component that would be used to control the satellite link based on these highs/lows? Is there a battery that has the option to disable/enable the power outlet ports this way?
Edit: I'm just writing out how I think it would work but please let me know if I'm wrong :)
-
I'm the last one to claim expertise on anything RF, but it just seems that if you could find or build a cellular antenna - a dish or yagi that's high gain and directional - that might get you to the nearest cellular tower from your remote site even though your phone won't connect from there.
-
A friend of mine lives on a large lake, he has a floating dock that is about 1/4 mile from the house. It is quite hilly, so he has a fairly high point to mount his antennas. He got some Yagi antennas and set up a Wi-Fi link between the dock and house. There is no power (by law) at the dock for safety reasons, so the dock end runs off solar + batteries. He has a camera on the dock, so he can watch for unauthorized activity. Part of the issue is as lake level rises and falls, the dock moves, and finding the right average aiming point for the house antenna is tricky. If you have two fixed antennas this could work better. I can ask him for brands of this gear, if it sounds like what you are looking for. The Wi-Fi transceivers use WAY less than 60 W, probably just a couple W, and he sends live video to the house.
Jon
-
I'm the last one to claim expertise on anything RF, but it just seems that if you could find or build a cellular antenna - a dish or yagi that's high gain and directional - that might get you to the nearest cellular tower from your remote site even though your phone won't connect from there.
Yup, my friend at the lake has no cell service from ground level. He put up a tower and gets wireless internet service through a local wireless internet provider. He has a Yagi antenna up on a tower pointed at their tower. That provider gives him wi-fi cell access from near his house (as well as other internet service.) (At least, I think this is how his connectivity works.)
There are commercial cell repeaters that are made for locations behind hills and inside shielded buildings. Some of these are totally passive, just little indoor antennas and big rigid coax to a Yagi on the roof.
Jon
-
Yeah, it seems that any time you can solve your problem with an antenna, it's almost like a free lunch. An antenna doesn't have any ongoing requirements for a battery or solar panels. It just extends the range like magic, which is basically what RF is in the first place.
I also wonder if something as power hungry as a Raspberry Pi is needed. Maybe you need something like that for the pictures, but if not, something less capable would be a lot more friendly to battery life.
-
What bitrates do you require? Your LoRAWAN link budget depends upon the spreading factor and if you can tolerate lower bitrates you can achieve longer distances.
If you don't think you can achieve a point to point connection, do you have the option of placing an intermediate node along the path? What about a third party LoRAWAN networks like Helium or The Things Network? It's worthwhile checking to see if there are gateways close by that you could leverage.
Satellite connectivity can be quite expensive and is still subject to conditions like rain fade depending on your starting SNR... so if you go that route don't assume it's going to be 99.999% reliable either.
-
Why do you need a DC-AC inverter? The Pi runs on DC so a simple high efficiency buck regulator is all you need for that.
The transmitting part I know less about, but as others have mentioned, you should be able to power it up only when needed. I have (shorter range) wireless devices that run from batteries for years, or in the case of my weather station there's a CR123 that powers it at night and a small solar panel in the day, the original battery is around 8 years old.
Do you even need the RPi at all? If you're just monitoring something this seems like a task for a small microcontroller which can draw microamps in standby and then wake up periodically to measure and transmit a burst of data. A 433MHz radio module with a directional antenna could potentially go miles line of sight without needing to be enormously powerful.
-
Why do you need a DC-AC inverter? The Pi runs on DC so a simple high efficiency buck regulator is all you need for that.
The DC-AC inverter was mainly for the satellite link modem
Do you even need the RPi at all? If you're just monitoring something this seems like a task for a small microcontroller which can draw microamps in standby and then wake up periodically to measure and transmit a burst of data.
Before the RPi I was considering using an ESP32-CAM but the area gets pitch black at night so I wanted to include a camera with night vision. I decided that it could be easier to use the MakerFocus Raspberry Pi Camera Night Vision. With the ESP32-CAM I would have to remove the IR filter lens (which will sacrifice quality) and include separate IR lights that would activate at night with a photoresistor.
A 433MHz radio module with a directional antenna could potentially go miles line of sight without needing to be enormously powerful.
I had read in this site (https://www.thethingsnetwork.org/docs/lorawan/frequencies-by-country/ (https://www.thethingsnetwork.org/docs/lorawan/frequencies-by-country/) and https://randomnerdtutorials.com/esp32-lora-rfm95-transceiver-arduino-ide/ (https://randomnerdtutorials.com/esp32-lora-rfm95-transceiver-arduino-ide/)) that 433 MHz would be a frequency used in Asia and that in the United States one should use 915 MHz. Should I still use 433MHz to get the farther distance? Do you know of any trustworthy components that have long reach that you could share?
I can ask him for brands of this gear, if it sounds like what you are looking for. The Wi-Fi transceivers use WAY less than 60 W, probably just a couple W, and he sends live video to the house.
Jon
I would appreciate it if you could ask him, thank you.
What bitrates do you require? Your LoRAWAN link budget depends upon the spreading factor and if you can tolerate lower bitrates you can achieve longer distances.
If you don't think you can achieve a point to point connection, do you have the option of placing an intermediate node along the path? What about a third party LoRAWAN networks like Helium or The Things Network? It's worthwhile checking to see if there are gateways close by that you could leverage.
Satellite connectivity can be quite expensive and is still subject to conditions like rain fade depending on your starting SNR... so if you go that route don't assume it's going to be 99.999% reliable either.
I looked at https://explorer.helium.com/ (https://explorer.helium.com/) and it seems there aren't any hexes covering the area so I assume there is no coverage. For The Things Network it looks like it's only towards the West coast. I could set up nodes along the path but what comes to mind is that'll I need separate power system for each (I guess they would be low power if I use something like the ESP32). For the bitrate I'm unsure but I will only be sending text if LoRa is used.
For all recommending antennas, I'll have to do a bit more research on that. I have no background with antennas but from what all have mentioned it looks to be a method I should highly consider. I'll be researching Yagi antennas jmelson and Peabody have mentioned. Thanks!
-
There's also the Moxon antenna.
Is it possible to identify the location of the nearest cell tower and which carrier runs it? Also, is there anybody living near your site?
-
I haven't looked at the regulations but I do know there is a lot of 433MHz wireless stuff sold/used in the USA. I've had several wireless thermometers and weather stations that use it, my current weather station is IIRC 915MHz. You'll want to check that of course to make sure you're on the level.
-
There are plenty of good tutorials and calculators available for stand alone power systems. Most of the information above is good but a bit scattered across messages.
1. Minimise your load (most efficient comms system) and then calculate Wh required daily.
2. Convert this to Ah at the battery voltage. Look at minimum recommended discharge for the battery type. E.g. often 50% for SLA. Calculate new required Ah. Allow 10% extra for losses in the battery.
3. Consider local weather conditions and how mission critical this is. Reliability usually allows 3-4 days of effectively no solar input, so increase your battery size according.
4. Look at Peak Sun Hours for the worst time of year in your area. Calculate size of solar panel required to recharge your battery in these conditions.
5. Find a suitable solar regulator based on currents calculated. Find suitable cables, fuses, breakers etc for each section of the system based on currents calculated.
6. Find suitable mounting systems, weatherproof boxes etc to assemble the whole thing.
7. Find someone competent to wire it all up.
8. Or go back to step 5 and buy something off the shelf that meets the criteria.
-
I haven't looked at the regulations but I do know there is a lot of 433MHz wireless stuff sold/used in the USA. I've had several wireless thermometers and weather stations that use it, my current weather station is IIRC 915MHz. You'll want to check that of course to make sure you're on the level.
Here in the US it falls under FCC Part 15.231. It's far more limited than using LoRA on 900 MHz.
https://www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-C/subject-group-ECFR2f2e5828339709e/section-15.231 (https://www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-C/subject-group-ECFR2f2e5828339709e/section-15.231)
-
Could you actually pull some cable? Would it be a problem with going over streets or lakes or places where wires are not allowed?
Maybe not even all the way to you, but to some high point or somewhere you have reception, so that you could use phone data/sms or something
rs-485 works up to 1200 meters or so, and there's ethernet over vdsl/adsl (2 pair wires) converters that work up to 2000+ meters (but they're kinda expensive) ... on the upside you also have more bandwidth maybe something like a couple mbps at 1000+ meters with the ethernet/vdsl converters, but could be more at shorter lengths.
A 305 meter spool of ethernet cable is relatively cheap, and you can just strip the exterior to get 4 pairs of 2 wires and make yourself 1200 meters of twisted wires.
example of ethernet-vdsl2 extenders :
up to 1000 meters : https://www.amazon.com/StarTech-com-110VDSLEXT-Ethernet-Extender-Single/dp/B002CLKFTG/ (https://www.amazon.com/StarTech-com-110VDSLEXT-Ethernet-Extender-Single/dp/B002CLKFTG/)
up to 1500 meters (but I think it's same manufacturer and product as startech above) : https://www.amazon.com/Ethernet-Extender-Kit-Tupavco-TEX-200/dp/B09L3DHMS3/ (https://www.amazon.com/Ethernet-Extender-Kit-Tupavco-TEX-200/dp/B09L3DHMS3/)
up to 2100 meters (7000ft), but 5mbps down / 2mbps up at that range : https://www.amazon.com/Tupavco-Ethernet-Extender-Kit-Repeater-VDSL/dp/B01BOD8C9W (https://www.amazon.com/Tupavco-Ethernet-Extender-Kit-Repeater-VDSL/dp/B01BOD8C9W)
-
There's also the Moxon antenna.
Is it possible to identify the location of the nearest cell tower and which carrier runs it?
There is a T-Mobile cell tower approximately 5.70 km NE from the location (I was able to obtain the exact coordinates). That is the closest one available from T-Mobile. At the moment, I don't know if there is a closer one from another company.
One question, would there be anything stopping me from connecting a large antenna (such as Yagi or similar antenna pointed towards the tower) that covers cellular frequencies to a device similar to this one (https://makeradvisor.com/lilygo-t-sim7000g-esp32/) with an SMA to u.fl adapter?
Also, is there anybody living near your site?
There is someone but they don't want to run any cables on the field and I need the system to run independently.
-
I'll defer to someone who knows about this stuff to answer about the antenna, but that's the kind of thing I was thinking of. So I hope the answer is that it should work.
On the neighbor issue, you don't need a cable, but if they have internet, you could just put some kind of radio receiver in their house, connected to their wifi router. It would receive your data from the site and retransmit it. You could pay them so much a month to allow that. But it would require that they have internet, and that they keep the power on all the time, and of course that they are closer to your site than the cell tower is.
I don't know what the terrain is like between your site and the cell tower, but unless it's awful I would think the 3.6 miles would not be at all out of bounds with a good antenna.
-
What sort of sensor are you using? I skimmed the above and didn't see it mentioned.
I'm assuming it's some sort of environmental sensor or something?
I typically do this with equipment from Campbells Scientific for environmental monitoring. Put one of their dataloggers with a spreadspectrum radio at your monitoring site running off solar. Put a cellular modem and spreadspectrum radio at the telemetered site, follow their online guides and off you go.
You may despise their programming language, but it's effective.
I'd use an 18AHr lead acid battery with a quality 30W panel at each end if I was doing the above although you could definitely get away with less. I'm aware of sites running a campbells logger and spread spectrum radio using a 10W panel and 10AHr battery without any issues.
-
Hi all,
I'm a student with a CS background but I want to setup a monitoring system in a remote area that doesn't have any power or reception. I investigated the potential to use LoRa as the method to transmit the monitoring data but I'm afraid it won't be able to send it the required distance (2km) since there are a lot of trees in the way and I'm not able to reach a high enough area. Instead, I'm thinking of buying a satellite service that I know reaches the area and setting up the modem along with a Raspberry Pi 3B to run on solar. Ideally the system would run 24/7 indefinitely on solar alone.
Below are some of the conclusions I've reached but I wanted someone to verify that I'm going through the right path before I buy anything :) .
I've researched the modem that the company uses and apparently it is the HT2000W which consumes a peak of 60W and 40W on idle. I think the the Raspberry Pi 3B uses a peak of 3.7W and 1.4W on idle. I will be connecting sensors to the Pi so this will probably consume more power, so just to play it safe I would say the Pi will consume 10W peak and 5W on idle.
I've heard that a good rule of thumb is to have a battery capacity of at least 2-3 times your daily energy consumption.
In this case, since it averages on 45W
45W x 24 hours = 1080Wh
1080Wh x 2 = 2160Wh
I would need a battery with a 2160Wh capacity.
For the solar panel, I would need at least 45W output from the panel. This would mean that I should get a 50-60W+ solar panel since some of the power will be lost at the solar panel -> battery -> devices. I've heard that at least 10% of the power is lost when stored in batteries (I guess this would depend on the type of battery?). I'm thinking I should get a better (100W+) solar panel considering sun availability + weather conditions might not be the best. Edit: I've read that I might need a charge controller. That way I can stop the panel from charging the battery at a certain level and maximize its lifespan.
This would all need to be waterproof and I would need an inverter (DC to AC) to connect my 2 devices.
That is what I have for now. All feedback and advice is welcome and appreciated!!! Please go easy on me :-X
Edit2: Also, are there any well known (newbie) kits that would meet the requirements for this type of application?
Hi,
What exactly are you monitoring?
Also, how often do you need to actually SEE the data? Is it once every minute, once every hour, once every day, once every week, once a month, etc. I ask because they normally use those special cameras to capture wildlife events out in the, well, wild. That includes in the forests.
If you can take data now and then but dont have to actually see it for a week, you can store the data on EEPROM or thumb drive then gather it once a week for uploading to your computer for analysis.
I happen to be in this same boat right now dealing with some telemetry to transmit data so i can analyze it minute by minute.
Solar panels can be a little tricky. You're a little in the mud before you even start because the panels are rated with an insolation level of 1000 watts per square meter while many areas dont get that all year round and may never get that. It depends on your exact location. If your insoluation level is 650 w/m^2 then you are already down by 65 percent. Factor in the solar panel efficiency of 20 percent (monocrystalline) and you are down to 13 percent. So if you had a 100 watt panel you could expect to get 13 watts, on a bright and sunny day with no obstructions that create shadows on the panel surface. If you have your panel in the forest you would have to make sure the surrounding trees dont shadow the panel for too long during the day.
Of course you have cloudy days. One cloudy day you can deal with, but several in a row could be a big problem. You may not have enough power to transmit anything.
You could use a control circuit to turn off the entire rig when you dont have to transmit, i think this has been mentioned before. The only thing then is to be able to create a circuit that can turn the rig on and off as needed, and that consumes power also, so you need to design a low power circuit for that. You might get away with a microcontroller with a sleep function (for very low power) and a low quiescent power 5 volt regulator.
Of course you should probably use a 12 volt Li-ion battery or similar. Lead acid has high self discharge.
Good luck with the project.