Tx/Rx 6 mile telemetry using APM+IMU

Wireless & IoT
1

Hello,

First, here is what we are trying to find out…

What is the maximum usable distance for a pair of Xbee Pro 900MHz (50 mW) using high gain antennas in practice (ie, in the real world)? The distance given is 6 miles with high gain antennas (from Digi).

Second, what we equipment we are currently using and what we are trying to accomplish…

We are currently working on getting telemetry back to the ground station.

We are using the Ardupilot Mega w/ the IMU Shield and a EM-406 GPS.

The telemetry down link needs to cover a distance of 6 miles. We are building a tracking antenna (high gain) to reach our 6 mile goal. If the Xbee Pro’s are not enough to get this distance, what are our options?

Using the Friis transmission equation, we calculate -75 db for the received signal (using a 5 dB Tx antenna and a 14 dB Rx antenna). Digi states that -100 dB is enough for the Xbee’s to work. Is -75 db enough for the Xbee’s to work well?

Thanks.

2

-75dBm received signal strength (inclusive of antenna gain each end and path loss at 6mi line of sight), is excellent.

Good rule of thumb in line of sight conditions: allow 10dB of margin. The -100dBm receiver sensitivity spec is for a given data rate (if it’s alterable) and a desired error rate - usually as “packet error rate”, where a packet is x number of bytes and includes some overhead bytes and may include some overhead for error detection/correction codes. Often, receiver sensitivity is spec’d at 1% PER (packet error rate) for a nominal packet size, which for this kind of radio would be 50-100 bytes per packet. Smaller, the PER improves.

For a device high in the sky, if it is receiving from the ground (bi-directional link), it has the greatest challenge with interference - due to it’s elevation.

15dBi gain antenna at 900MHz - are you sure that’s correct? That’s high, for the size of the typical small yagii. And high gain can make the antenna hard to keep pointed correctly.

The 6 mile line of sight path loss is 111dB. This assumes Fresnel zone clearance, which applies to terrestrial links, not links to airborne things.

With that 15dBi gain (questionable), and a -90dBm desired signal strength, there’s 17dB of margin. You might want to go to less antenna gain on the ground so you can use a less difficult to aim and re-aim antenna. At 15dBi, it probably has a beamwidth of less than 20 degrees. I’d go for 45 degrees if you can.

If the link is terrestrial point to point, you will need elevated antennas for 6 miles, and that elevation depends on minimizing the fresnel zone clearance losses. It’s an ellipse between two antennas, where part of that ellipse can be occluded by terrain, earth’s curvature, buildings, trees, etc. Too much blockage and the path loss increases a lot over simple line of sight. There are on-line link budget calculators and downloadable spreadsheets that include fresnel zone clearance for your chosen antenna heights (each end).

The XBee Pro XSC for 900MHz says it produces 17dBm of transmitter power (50mW). Is this what you’re using? To meet FCC rules, they are frequency hopping. And that helps mitigate interference.

3

Thanks for the reply.

stevech:
-75dBm received signal strength (inclusive of antenna gain each end and path loss at 6mi line of sight), is excellent.

Good to know. I was referencing signal strength (in dB) for wireless networks at it seemed like -75dB would be okay and do the job.

stevech:
Good rule of thumb in line of sight conditions: allow 10dB of margin. The -100dBm receiver sensitivity spec is for a given data rate (if it’s alterable) and a desired error rate - usually as “packet error rate”, where a packet is x number of bytes and includes some overhead bytes and may include some overhead for error detection/correction codes. Often, receiver sensitivity is spec’d at 1% PER (packet error rate) for a nominal packet size, which for this kind of radio would be 50-100 bytes per packet. Smaller, the PER improves.

I’ll keep the 10 dB margin in future calculations. I threw together a spread sheet so I could just plug and chug numbers. After some reading I found that doing a link budget was a good idea as well as the Friis equation which seem to have agreeable results. You can see the spread sheet attached if you want.

stevech:
15dBi gain antenna at 900MHz - are you sure that’s correct? That’s high, for the size of the typical small yagii. And high gain can make the antenna hard to keep pointed correctly.

The 6 mile line of sight path loss is 111dB. This assumes Fresnel zone clearance, which applies to terrestrial links, not links to airborne things.

With that 15dBi gain (questionable), and a -90dBm desired signal strength, there’s 17dB of margin. You might want to go to less antenna gain on the ground so you can use a less difficult to aim and re-aim antenna. At 15dBi, it probably has a beamwidth of less than 20 degrees. I’d go for 45 degrees if you can.

I lied, it is actually a 14dBi antenna. Both antennas we are looking at are from l-com.com, [14dBi Yagi and [5dBi rubber ducky. The beam width on the Yagi is 30 degrees…so not quite 45 but we thought this would still be enough since the plane would always be quite some distance away from the antenna. You say small, but this Yagi is…I would say huge. We are prototyping a smaller tracking antenna and once that tracks properly we will be moving up to the larger Yagi, unless we can find an antenna solution that is smaller or a different Tx/Rx pair.

I calculated the same dB loss for 6 miles, hopefully that is a sign that my math is right. :slight_smile:

stevech:
If the link is terrestrial point to point, you will need elevated antennas for 6 miles, and that elevation depends on minimizing the fresnel zone clearance losses. It’s an ellipse between two antennas, where part of that ellipse can be occluded by terrain, earth’s curvature, buildings, trees, etc. Too much blockage and the path loss increases a lot over simple line of sight. There are on-line link budget calculators and downloadable spreadsheets that include fresnel zone clearance for your chosen antenna heights (each end).

The XBee PRO XSC for 900MHz says it produces 17dBm of transmitter power (50mW). Is this what you’re using? To meet FCC rules, they are frequency hopping. And that helps mitigate interference.

We recognize the curvature of the earth will play a role, but haven’t factored that in yet. I was going to make another spread sheet to determine how high the antenna would have to be mounted after we determine how high the plane will be flying at. Just to clarify, the ground station has the tracking Yagi while the plane has the duck.

I think the transmitter output of the XSC’s is 100mW. We would like to use these; however, I do not believe they will work with the Ardupilot Mega and it’s telemetry port as it is currently setup. We were going to use the XBee-PRO 900’s which are 50mW and 10% PER at -100 dBm (close to what you said). Again, we could probably benefit from the XSC’s or even the Xtend’s but I can’t find any documentation that the Ardupilot Mega supports these.](http://www.l-com.com/item.aspx?id=22181)](http://www.l-com.com/item.aspx?id=21959)

4

These Digi modems all have ordinary serial ports with “Hayes modem-like” AT commands. Simple to program for.

dBm is an engineering unit of power, relative to one milliwatt (0.001 watt). 0dBm = 1mW.

dB is decibels, a log scale. By itself, it is dimensionless.

So if we have a signal with 0dBm power, then increase that by some antenna gain such as 3dB, we have twice the power, or 3dBm.

10dBm = 10mW. 20dBm = 100mW. And so on.

Another popular unit is dBmV (dB millivolts), where 0dBmV = 1mV. Most often used in cable TV systems and the like, rarely in RF.

Antenna gain is commonly measured as the gain in dBi or dBd, usually the former. dBi is decibels relative to an ideal isotropic (sphere). A perfect isotropic antenna radiates power equally in all directions. Antennas have gain because the are directional, of course. So a 14dBi antenna has 14dB of gain relative to the ideal antenna. The term dBd is gain relative to a dipole antenna. As I recall, 0dBd = 2dBi, that is, a dipole is somewhat directional. But again, dBi is more common. A decent antenna will have available its gain plot charts. A horizontal plot is “looking down” on the antenna and the plot shows a bulge in the direction of radiation, with side lobes, back lobe, etc. A vertical plot is the same, but looking “at” the antenna. Some antennas are omnidirectional on the horizontal but directional on the vertical - like a doughnut pattern. Others like yagii are directional mainly on the horizontal. Some are directional on both, often called sector antennas.

Hope this helps.

There are a LOT of on-line link budget and fresnel zone calculators as web pages or downloadable spreadsheets.

5

We have done quite a lot of development with Zigbee in the past and have never managed to RELIABLY acheive the ranges sometimes claimed by RFIC or Module manufacturers. However, I have used the Jennic range of modules (100mW output FCC approved) to get the best range of any tried so far.

That said we have not done any dev with the new Microchip chipsets… yet.

We have also not used any Gain antennas in the application as using these would invalidate any approvals already given to the module in question.

6

In any RF system, it’s all about antenna gain, for some frequency. Of course, the loss per distance unit is frequency dependent and increases about 6dB per decade of frequency.