1500 WATTS of LoRa Troposcatter on 440MHz

1500 WATTS of LoRa Troposcatter on 440MHz


We have had a lot of fun with LoRa at 1 watt over the last two years in the 440MHz / 70 cm band with HamShield: LoRa Edition. Over the last 2 years, we have done plenty of drive tests, with strong signals through 6 miles of forest, and all sorts of seemingly impossible feats. We even got to try out our 8 watt amplifier for more impressive results. 

But what if we took things a little bit further? How about we plug HamShield: LoRa Edition into our new 80 Watt DMR amplifier? Then, lets plug it into a 13 dBi Diamond Yagi antenna and point it straight up? Did you know that would be 1500 watts EIRP? You couldn't possibly do NVIS-like transmissions at 440MHz with cheap consumer hardware..right guys?


Radio waves bounce off a lot of strange things, including differences in temperature, moisture and more. Ducting through the troposphere can cause line of sight signals to travel hundreds and sometimes thousands of miles. It has been used since the 1950's as an alternative to satellite communication. Check out this short Wikipedia article regarding the subject.

Math Time

Lets do some quick math here first...How the heck did we get 1500 watts from 80?!

First, we will pop over to one of the Inductive Twig calculator and calculate our EIRP. Will need to know what 80 watts is in dBm (milliwatts on the decibel scale) so we can add it to the gain of the antenna, which is measured in dB. We will use the Watt to dBm converter.

Now that we have converted 80 Watts (Technically, 80,000 milliwatts) to dBm, we need to add the gain of the antenna to figure out the EIRP or Effective Isotropic Radiated Power. While we are not creating new power, we are just focusing it. This is like using a magnifying glass with sunlight, starting a small fire--which we will soon do to the sky. Okay, not really.

49 dBm + 13 dBi = 62 dBm

dB, or decibels are neat, huh? When combining the gain of the antenna (13 dBi) with the power we input (49 dBm), we get a number in dBm again -- effectively the total wattage output of the antenna's focal point if it just were a simple isotropic (aka, perfectly omni directional) antenna.

Lets convert 62 dBm into watts. Since dB (decibels) are logarithmic, we can probably assume that 62 dB is a stupidly large number. Indeed, the Inductive Twig dBm to Watts calculator concurs:

Well, not quite a "jigawatt", but that's still 1,584 watts! Wow! Now, while we can probably expect that at the focal point, there are a few key things to keep in mind:

  1. Only a small beam from the antenna will be this power level, with radiation dropping off dramatically away from its focal point.
  2. There are losses in our coaxial cable and other connections, which could take away a dB or two. Remember that even just 1 dB at these power levels can be pretty dang dramatic (Power doubles every 3 dB), which is why we need to use quality connections and cable.
  3. We have to assume the antenna manufacturer didn't over embellish its advertised gain, but Diamond does have a good reputation.
  4. This is actually quite normal power in the RF world, which is why you see high gain directional antennas for long distance point-to-point links
  5. You can expect dipole-like performance from any directional antenna (exception of a dish). Since a yagi focuses a dipole, testing has shown that they have great dipole omni directional performance. But since dipoles generally suck with gain, they only create localized coverage areas. This is mitigated by a thing called front-to-back ratio, but the back side of a typical Yagi is only 15-20 dB lower.
  6. People use even more power to bounce signals off the moon, a mode called EME (Earth Moon Earth)

Why do we want to even bother having such high wattage? Well, our goal is to (hopefully) have a small faction of our energy get reflected back off meteors, temperature inversions, planes, dust, etc.

Test Setup

For our demo, I had to get creative. I didn't exactly have a vertical Yagi antenna mount, so I used the next best thing: a landscaping shovel that I shoved into the ground and bolted to the antenna. Thats pretty easy here in Cape May, New Jersey as my yard is just grass on sand and I haven't seen a real rock in months. NOTE THIS ANTENNA IS NOT EVEN A FOOT OFF THE GROUND. Dipole radiation should be almost non existent at a mile or so.

I found a random chair, insulated the HamShield, and connected our 80 Watt DMR amplifier to a beefy 12VDC power supply. I used our SMA to SMA adapter to connect HamShield: LoRa Edition to the amplifier, powered the Arduino Uno off a lighter jack USB adapter, and then ran the amplifier coax to the antenna.

For the receiver, I connected a HamShield: LoRa Edition + Arduino Uno to an android-based radio receiver my wife has in her car. I used a cheap magmount dual band antenna I found at a Seattle-based online megaseller. I downloaded a serial terminal program from the Google Play Store. She got really mad I didn't take my car, but my stock Honda navigation system couldn't even dream of running such cool applications like a serial terminal! As I drove along, beacons would come through with time and signal strength. 

I would have got a picture of the car, but the second I arrived, I got a lecture about taking her car for several hours without asking, was told to immediately remove the giant antenna from her roof, and to "take all my wires inside".


Usually, these drives have a sense of excitement and then eventually sadness. You get reception in some challenging environments, then a bit of packet loss, with the reception gradually trialing off in decibels until it stops. Path loss is extremely mathematical. 

However, it didn't stop this time.

After leaving the dipole's estimated range of 2 miles, the packets continued to come in. However, in an unusual turn of events, they bottomed out with signal strength. -115 to -130 dBm was common, peaking to -140 dBm -- no matter how many miles away I was. However, they came in during bursts. There appeared to be brief windows of less than a minute -- no matter where i was -- that packets would come in. Sometimes they were very strong that made no sense given my path loss.

With a distance on flat ground of 20km / 12 miles, I was at the most southern tip of New Jersey and could not go further south. Regrets of not driving north...

The preliminary results show that the minuscule reflection of 432MHz combined with LoRa's insanely low receiver sensitivities, even in forecasted non-tropospheric ducting conditions, can provide reliable digital communications with low amounts of power. Let that sink in. In 2020, we can now bounce high speed digital UHF signals off the upper atmosphere like our grandfathers did with morse code, short wave radio, and the ionosphere.

Above are the 12 test locations i selected during my drive test. All locations were able to receive packets, in bursts, at random times. The blue dot is the transmitter location. The scale of the image is 20 miles x 20 miles (32km x 32km).

I later spoke with several folks who have done similar tests with even lower power. I am being told that it wouldn't be entirely unusual for this signal to be received for another 50-100 miles. One person suggested to point the antenna at a 45 degree angle, in order to improve the "skip" performance.

Another interesting factor is that a car's vertical magmount antenna is probably not the most appropriate for this type of communication, but it works anyway. I suspect that a horizontally polarized antenna -- or even better -- another vertically pointed Yagi could result in even greater range. 

This preliminary test was quite exciting. So, it goes without saying, more testing soon!



  • Eamon

    ERP is radiated power times isotropic gain, so a dipole has a bit more ERP than actual radiated power.

  • Casey

    @William You are technically correct. Although, I was worried about going on a tangent about explaining dBd and dBi. You are correct that a dipole antenna is not perfect and has slightly more gain than the non existent isotropic antenna. I edited the text a little bit to reflect that. All antenna gains are mentioned in dBi.

    We have an amplifier for 8 watts, however, it doesn’t have a switch back to receive. Our 80 watt amplifier is a fast switching DMR-friendly bi-directional amplifier, so it will work fine. We have a limited number of these while we gauge market demand, as they are expensive.

  • William

    Thank you for saying “focusing” the signal, not actually changing it.. 13 dBi … followed by … “gain over a ‘dipole’”… dBi is isotropic gain. dBd is gain over a reference dipole usually about 2.1 less than dBi I don’t know if the 62 Dbm should include a ‘i’ somewhere to note it’s isotropic ERP or if the 62 should lose 2.1… Perhaps it’s close enough where it is.

    Either way this is way promising. If we do have an off the shelf 8 W unit for Ham data – it could be a game changer. Even 1 W has great promise.

Leave a comment

Please note, comments must be approved before they are published