APRS over Lora 433MHz

Completed portable tracker. 

With recent advancements in Lora, and the cheap boards on the market due to Meshtastic, I started researching into APRS over Lora. I found that several locations are already using this method in the EU, as well as here in the states. Specifically, I noticed a lot of activity around the Denver area. I reached out to Ham Radio Operators that were sending APRS packets over Lora using the iGates in their area, and they pointed me to an updated firmware by CA2RXU. I did run into some problems getting my device operational, and I am hoping these companion instructions will help others get started with APRS over Lora. 

What is impressive about APRS over Lora, is how well it works, and only 1 watt, while on a UHF frequency. Lora can be in the background noise, and still be heard by the sensitive receivers, making extreme distance possible on minimal power. 

Some minor picket fencing of the signals still occur which is expected, but overall this still appears to work extremely well. iGate setup with a Diamond VHF/UHF antenna 15' AGL, while the mobile used a simple mag mount antenna.

APRS over Lora is starting to be used more in the US. 

These are the firmware download links:
https://github.com/richonguzman/LoRa_APRS_iGate
https://github.com/richonguzman/LoRa_APRS_Tracker

Detailed configuration info:
https://github.com/richonguzman/LoRa_APRS_iGate/wiki/01.-Installation-Guide
https://github.com/richonguzman/LoRa_APRS_Tracker/wiki/01.-Installation-Guide-%23-Guia-de-Instalacion

I got started with Heltec V3 for the hardware. These can be purchased directly from Heltec.org, but shipping times are longer. They are also available from Amazon for faster shipping, but higher cost:

Heltec v3 ($26)
https://www.amazon.com/dp/B07FYWFH4C

5db 433MHz SMA Antenna ($7):
https://www.amazon.com/dp/B07J6GYKTZ

iGate Case Only:
https://www.thingiverse.com/thing:6540986

Tracker Parts Only:
GPS Module (~$9.50 Each):
https://www.amazon.com/dp/B08MZ2CBP7

3000mah Lithium Battery (~$7.25 each):
https://www.amazon.com/dp/B08T6GT7DV

433Mhz Mag Mount Antenna ($15):

https://www.amazon.com/dp/B09FNB8JW1

3D Printable Case:
https://www.thingiverse.com/thing:6554936

This makes for a total of each device:
iGate - ~$30 (Using Ham radio antenna)
Tracker - ~$43 (standard antenna)

iGate Setup:

Once at least the Heltec v3 kit is in hand, its easy to build an iGate because of the available firmware. Download VStudio, and open the extracted zip folder of the firmware.

Make sure you have an antenna hooked up to the board before plugging in any USB cable to power it. Select the correct board type that will be used. Now, the firmware is in 2 parts, the main code, and the Filesystem. If both are not uploaded, the firmware will not start. (Do not modify the firmware files, only set the board type).

Install PlatformIO extension

Once the firmware is loaded, the iGate will start its own Access Point for configuration.

Connect to "N0CALL AP", with 1234567890 as passcode. Access http://192.168.4.1, to set to the settings of iGate initially.

Change the Callsign, and add APRS code for APRS-IS. https://apps.magicbug.co.uk/passcode/
Add a WiFi network for the device to connect to, and then save the configuration. 

Device will reboot and start gating traffic to APRS-IS. Hook up an external antenna if possible to improve the range.

Completed and operational iGate at home. 

Tracker Setup:
The same process is needed for the tracker, but a file will need to be modified for the callsign. 

Not required, but if you want to add a buzzer, the pins need to be selected, and options enabled. These are the settings I am using on my Heltec v3 tracker. Other devices will have different pinout numbers.

GPS:
TX -> GPIO48
GND -> GND
VCC -> 3.3v

Buzzer:
- -> GND
+ -> GPIO33

Solder the wires for the GPS first (3 wires), and install into case. Then add the SMA connector. Careful with the case, you should not need to make the hole bigger unless 3D printer is not calibrated correctly. 

This case fits the larger battery, which should last a very long time.

Detailed view of inside the case. 

Be sure the main code as well as the file system is uploaded to the device.

Ed
KI6PSP
http://KI6PSP.blogspot.com

Effective Diffraction of Radio Waves

Effective Diffraction of Radio Waves

By: Edward Czajka

Radio waves act differently when their frequencies (wavelengths) are changed. UHF frequencies, between 300 MHz and 3 Ghz, have a tendency to only operate by near line of sight, while HF frequencies, between 300 KHz and 30 MHz, tend to bend and hug the surface of the earth or bounce off the atmosphere. VHF is between the aforementioned frequency regions, between 30 MHz and 300 MHz, and it will act like both HF and UHF, as it will sometimes bend around solid objects. This is a graphic demonstrating the different radio bands. 

AM Broadcasts are between 520 KHz - 1710 KHz, Citizen Band (CB) is around 27 MHz,  Family Radio Service (FRS) is around 462 MHz, Cell phones operate around 800 MHz, and WiFi networks operate on 2.4 GHz. My intent is to test the ability for a radio wave to bend around objects, and compare that to other frequencies, to measure performance.

 Experimental design

This experiment will be conducted with two radio sites, one radio station with a long wire antenna for receiving, and the other station will transmit with a handheld radio through a multiple band antenna. I am a licensed Amateur Radio Operator (KI6PSP), and I will be transmitting on Amateur Radio Frequencies to my receiving station (KC6UWM).

This experiment will involve transmitting on frequencies with wavelengths of 2m (144-148 MHz), and 70cm (420-450 MHz). Multiple locations will be used while transmitting, and the positions will be near a mountain, so I can measure how the signal will bend around a solid object. My transmitting station will be a Yeasu VX-7R with a SRH940 Antenna. The receiving station will be a Kenwood TS-2000 utilizing a 6m dipole antenna. Nearly every Amateur radio has what is called an "S Meter", for Signal Strength meter, that will show us how strong the signal being received is. We will use the S Meter to determine the signal strength of the transmitting station, and compare the signal strength values when we change location of the transmitter. I will begin my tests with a clear line of site to the receiving radio station, and establish a baseline to plot my data from. I will transmit on a given frequency from my control point with 2.5 watts, and the receiving station will log the signal strength he receives, then without changing any settings on the radio, I will walk 25 feet away from my control point, into the canyon, and transmit again for 10 seconds. The receiving station will log the result, and I will continue this process until the signal is unreadable due to the frequencies inability to diffract around the mountain. As I move away from my line of sight control point, there will be more mountain mass between the two radio sites, thus we can measure the diffraction of the radio waves around the mountain. Once my signal is unreadable, I will return to my control point, and then change frequencies to test the next frequency band. I will plot my data, and base the loss of signal on the line of sight control. I chose this design because I could easily obtain the required equipment, I could easily perform the experiment, the results can be easily duplicated, and the transmitting station would be easy to operate due to it's simple setup. To reduce threats to internal validity, I will utilize the same equipment with the same settings (Antenna, Transmitter Power, Height above the ground) for each transmission, and perform my tests on the same day concurrently, so environmental variables (changes in temperature) will not affect the results. 

Because I will not be able to correct for Free Space Loss, I will baseline my data on a line of sight data set, and plot the reduction of signal as I move behind the mountain.

The Red Pin is the receiving station, while the Purple Pin is the Transmitting Station. Notice the small mountain near the transmitting station.

Google Map (2011)

 Literature review

A RF diffraction experiment was performed by 2 amateur radio operators in Virginia, and they demonstrated that radio waves can bend around an object like a mountain, even on Ultra High Frequencies (between 300 MHz and 3Ghz) (ARES/RACES of VA, 2007). This is a good example for my experiment because it shows that signals will diffract (bend) around solid objects, such as mountains.

A similar radio wave experiment was conducted by students at Kansas State University, they tested the effective range of radio frequencies with limited power levels (Kansas State University, 2009). In their experiment, they measured the effective amount of signal they could receive at a distance. 

 Dependent, independent, and controlled variables

Independent: Frequency (wavelength), Location relational to a solid object

     Because I am comparing the performance of different frequencies, I must have two independent variables, otherwise this would simply be a demonstration of how one frequency diffracts when I change position, verses comparing the results of several frequencies.

Dependent: Measured signal levels of transmitted signals on other side of object, relative to line of sight control.

Controlled Variable: Height of Transmitting station above the ground, Transmitter power, Antenna used to radiate the signals, Length of transmission.

 Hypothesis

The ability for a radio wave to bend (diffract) around a solid object is inversely proportional to it's frequency (wavelength).

I developed this Hypothesis based on some personal experience and the desire to quantify radio frequency performance. Upon research, I discovered that there are various mathematical equations that Radio Frequency Engineers use to build RF links when objects are blocking line of sight (Afar Communications, 2011). They use a Fresnel Zone to calculates the area of the object that is blocking the path, factoring the wavelength of the radio frequency used, and use the resulting data to plan their links. They will usually plan on having no more than 40% of their Fresnel Zone obstructed, to have reliable communications. 

 Experiment Data

Conclusion

The data collected clearly shows that the different frequencies diffracted around the mountain in the experiment. When you compare the results from 2m and 70cm, they confirm my hypothesis. The amount of diffraction of the 2m signal is clearly more than the 70cm signal. The 70cm signal dropped to 50% of it's signal strength at 50 away from the control point, while 2m reached 125 feet before it yielded the same results, clearly an improvement of diffraction.

Experimental design is essential to a reliable experiment, as lack of design and planning can influence the experiment results. I had to redesign this experiment because of lack of equipment, and technical difficulties. I originally started with a Service Monitor, capable of giving data by the dBm, but it wasn't sensitive enough to receive the signal at the distances we were using. During the redesign process, I was able to figure out, that I didn't need to have the expensive equipment, and I could achieve the same results by using a control point, and keep everything relative to the control point. This would eliminate various discrepancies with the equipment used, and how they operate with different frequencies. Proper planning was still a big factor to producing accurate results. 

I performed this test with two basic amateur radio stations, so it can be easily replicated by another group of people with minimal equipment. This would simply involve using two licensed operators, and two multi band radios, with a solid object to test with.

References:

Google (2011) Terrain Map. Retrieved on July 22, 2011 from:

http://maps.google.com 

ARES/RACES of VA (2007) Knife Edge RF Diffraction. Retrieved on July 20, 2011 from:

http://aresracesofva.org/index.php?option=com_content&view=article&id=71&Itemid=116

Kansas State University (2009) Propagation Comparisons at VHF and UHF frequencies. Retrieved July 20, 2011, from:

http://www.ece.ksu.edu/~wkuhn/pubs/Propagation_Comparisons_at_VHF_and_UHF_frequencies.pdf

Afar Communications (2011) Fresnel Zone Calculator. Retrieved July 20, 2011, from: http://www.afar.net/fresnel-zone-calculator/

Mobile Directional Lightning Detector

I have discovered that lightning causes interference to AM radio receivers from a great distance. This is due to the way the AM recievers operate, Amplitude Modulation, and the spark of lightning causes a wide band pulse of energy around 500Khz. This means that while I am out chasing storms, I can listen to the AM radio around 530Khz, and determine the relative activity of the storms around me.

The interesting part of this, I can use the radio to listen to a storm even though it is not producing Cloud-To-Ground (CG) lightning. The radio can also recieve the Cloud-To-Cloud static discharge as well. The only problem is the radio and antenna of the car is Omni-Directional (I receive the signal from all directions), so I cannot use the radio to determine where the stronger storms are around me.

I started looking online for Lightning Detectors, and what commercial products I found were either expensive, not informative. There are some portable lightning detectors on the market that will estimate the time a storm will arrive, based on the same AM Pulses. Other detectors are dependent on a computer with expensive software, and it uses other similar stations to triangulate the location of a lightning strike.

I started thinking about Doppler Shift Radio Direction Finding, and even asked a local Elmer if I could use the same technology tuned for the 144MHz band on the 500KHz lightning pulses. This was ruled out due to the need to adjust the antenna size and spacing to achieve what I wanted to do, making this a non portable project.

Directional Lightning Detector
below my mobile HAM Radio
APRS and NOAA WX

The Elmer did suggest an alternative technique for radio direction finding, and that was to use the signal strength. This is what most people do when they first go Fox Hunting (Radio Direction Finding), they put a handheld radio against their body, and while rotating, they observe the signal strength meter. Using the body as an attenuator, you find the dip in signal, and the signal is coming from directly behind you.

I came to the conclusion that I could build four Lightning Detectors, and build a four element antenna so I could determine direction of lightning activity. I found some diagrams online for Lightning Detectors, and placed a parts order with Mouser.

I started building the recievers into a box, and used a 5v regulator for each receiver. I had some issues with two of the receivers, and that required some troubleshooting and replacing a few blown components. I also noticed some interactions between the receivers while testing, so I added a diode to the ground of each reciever and that seemed to isolate the recievers from each other.

Built receiver box

I salvaged an IDE connector from an old motherboard and used a computer case wiring harness to enable me to disconnect the front panel lights to work on the unit if necessary. While I was building the units, I added the Red LEDs to each receiver for testing purposes, but I didn't bother to remove them once complete.

Bearing indication of lightning

Testing each receiver is easy, simply take a multimeter on conductivity test, and when you test between the GND and the antenna lead, the circuit should activate the LED for a moment. The lightning simulator from TechLib is also useful, but a similar can also be done with a BBQ sparker or an Aim-N-Flame. A small spark will cause a similar burst of interference as a bolt of lightning. Because of the small size, the sparker test stops working when more than a foot away from the antenna.

Next, I started building a simple antenna array for the recievers. I used a tupperware container, and wrapped cardboard with aluminum foil.  I double stick taped the foil so it would stay on the cardboard. I created a small channel under the bottom square, so I would not damage the RG-174/U Coax. I placed two hard drive megnets inside the case, under the base square of the antenna, so I could Mag Mount the antenna to the top of my car.

Ground wire, and magnets on bottom

Before securing the base of the antenna to the case, I poked a hole in the center of the case, and ran a ground wire. This would allow the antenna to use part of the car as a ground plane. This might not be the most efficient design, but I am not transmitting with this unit, only recieving.

Antenna Elements

Next, I started building the ellements. I took the ground from each coax, and soldered them together in the center. I added the cross to isolate each antenna. This design would give me an eight way direction of the storm, in theory. I also added a ground wite to each of the crosses, and soldered it down onto the base. I checked that I had a good ground with a multimeter. Next, I took a few feet of wire for the element and attatched it to the cross with more stick tape.

Here is the completed antenna. Now if I can just get people to accept it, instead of yelling at me, while driving 70 MPH, that I "have something on the roof". I think a gray can of spray paint should do the trick.

Now if I can just get a few thunderstorms in the area.

Ed KI6PSP http://KI6PSP.blogspot.com

Portable Lightning Detector

This is not directly Ham Radio Related, but Radio theory does apply. Some of my hobbies include Photography and Storm Chasing. I chase storms from a safe distance, and capture lightning as seen in this picture. While chasing storms, I use the AM radio in my car, tuned to the lower part of the band where it is quiet. I can hear the interference generated by the lightning strikes through the AM radio as it is susceptible to the wide band low frequency pulse from the lightning. This method helps to give me an indication of how active the storms are. The Pops I hear on the AM radio are generated by both Cloud-To-Cloud and Cloud-To-Ground strikes, so this can be useful in determining if charges are still building within the storms nearby. One problem with the AM radio, it hears the lightning strikes from a good distance away, but it doesn't give me any direction information.
	With this in mind, I began researching a lightning detector that would react to the same low frequency pulse. I found a few diagrams online that would listen to the AM burst around 300KHz generated by the lighting. I ordered a bunch of parts, and spares, to build a Directional Lightning Detector. After I built the Directional unit, I used the spare parts to make a Portable Lightning Detector. When a pulse is detected, the red LED will illuminate for about a half second. I added a low voltage vibration motor to this unit so I could have it on my side, and be alerted to lightning in the area. This type of vibration motor can be found in old pagers and cell phones. I plan on adding a 75db 3v buzzer to this design, so I can switch between an audible or a silent alert.
Here is the inside of my Portable Lightning Detector. The unit operates from 2 AA batteries. The White switch is for power, and the silver switch is for which alert, audible or vibrate. I was going to use the small speaker in this picture, but I am opting for the buzzer instead.
	I used a 2m/440 Amateur Radio Antenna for this setup, onto the SMA jack I mounted onto the case. Sensitivity can be improved with a longer antenna if needed. This size of antenna should make this unit sensitive for what I need, while keeping the length to a manageable size. Current draw while on standby is about 5mA. When the vibrate motor starts, the draw spikes to about 50mA. Based on 2500ma AA batteries, I estimate a runtime of about 300 hours or 12.5 days of standby time.

Below is the schematic diagram of how this Lightning Detector is currently put together.

Ref:

http://www.techlib.com/electronics/lightning.html

Person & Event Tracker (PET)

I have ran logistics for several horse rides in my area, that consists of having a radio operator at water stops along a 50 mile course, and having them radio times of each horse back to base camp. This information would be collected so I can maintain accountability of each participant of the event. This way, we can see who is still on the trail. This system works well, however, when I have operators that don't want to abandon their location, but need to leave, Base Camp is asked for an ETA of that last rider. In this instance, I looked up some equations, calculated the average speed between previous checkpoints, factored in a hold time, and used it to estimate the ETA to the checkpoint requesting the information. I quickly realized this math could be automated.	Analog Whiteboard for Data
Raw Data Tab, with calculations	My first thought was to host a database on a PC, with a web portal to be used for data entry and display. The main reason for this mentality is i want to be able to enter data quickly via an iPad. When checkpoints call in, they give me a very quick dump of information. Something like "Base this is Water 1 - at 0821, horse number 103 and 107". With data moving quickly, I would like to select the water stop, then type the horse numbers, and the time. The database could put the data entry in the correct place automatically, without me having to scroll and find the field for each entry. I knew this can be done with Java, but I did not have the time to start building something that big from scratch. I began piecing this together as an Excel file, and created the equations to give me the ETA data of each point, based on historical data.
I deployed my creation as Person & Event Tracker 1.0 (PET), and started using it with real data. I was able to let the checkpoints know what time to expect the first riders to arrive, within 5 minutes of the actual time they did arrive. This worked across the whole event, and I was able to have the base camp people ready when the lead horse arrived, proactive instead of reactive. With this information, I was also able to project times of the slow riders, and calculate their arrival time to the finish line. In once instance, I passed along data to the last checkpoint, to inform the riders to speed up or they will arrive 10 minutes past the cutoff time for the event. This also gave me the ability to inform the ride organizers that they needed to go break glow sticks along specific sections of the trail, an hour before sunset. This system was being proven as a great logistical support tool.	Main Display Page on PET 1.0
Main Display Page with Next ETA Field on PET 1.5	The really cool part of this system, was I embedded the Excel file into a webpage that was hosted on my computer. This allowed me to hand the ride organizers an iPad that would display the data I have for them, and automatically update every few minutes. This reduced the chatter between me and the organizers, as they could see the information they needed from their location. While using this system, I occasionally needed information from my computer through my iPad but I was out of range of the wireless access point, with the auto refresh clearing the data on my screen. I made another page that would display the same data, without the auto refresh, allowing me to walk to the end timers, and collect more data, while still able to see main data.

After a few lessons learned, and wanting to make the system easier for other people to start using, I revised the data entry table, to make it easier to use and lowering the risk of messing up the equations. I also added a "Next ETA" field to the main display, so the organizers can quickly see when the riders should arrive at the next checkpoint.

I am willing to share my work to others interested in tracking an event like this. My current version will track up to 200 people, and up to 10 checkpoints of calculations. I can expand this, but I wanted to keep the file size down for now. The current version does load on a 1st Generation iPad via Safari. The webpage is hosted on a Windows computer, running Internet Information Services as the hosting method. This could be used on a real website, to give this data to ride participants and support personnel if the Excel file is regularly uploaded to a public website. This version must be edited on a computer running either Excel or Open Office. Pages on the iPad will delete some of the complex equations because it does not support some of the methods I use for the math.
 

 Summary Tab showing totals for each checkpoint Used to see how many riders still need to go through a checkpoint.

 Event Variables Distances between checkpoints and hold times are entered here. Accurate distances = Accurate ETAs

Raw Data This is the main working tab, for data entry. The equations have been moved to protect them from changes.

If you are interested, just send me an email or post here. My Email address is current on QRZ.COM