Broadcasting LoRa packets wihout a radio [video]
youtube.comI really liked the explanation of how he used harmonics to get to the frequency he desired.
Also, LoRa is really good when it comes to low power data transmission. To share a recent example, with just ~100mW of transmission power, they managed to fly an RC plane to a distance of 100km: https://www.youtube.com/watch?v=CYJ2UOrlXgM .(Their transmitter from the ground station was configured at a higher power, but the transmitter on the aircraft was transmitting telemetry back at 100mW throughout the distance)
Another related fun project is using raspberry pi gpio to work as an FM transmitter: https://github.com/markondej/fm_transmitter
I looked a bit into the LORA protocol. It does offer reception below the noise floor. LORA doesn't do anything to attempt to make other LORA transmissions seem like noise to a receiver that's attempting to lock to your transmission. This makes it possible for a transmission that's closer to the receiver to essentially jam a transmission that's far from it. I guess since LORA isn't used that much, it's not a problem these days. I think it's being used as part of Amazon's sidewalk network though.
One of the things you can do with Lorawan is use multiple gateways. Interferers are always an issue for a bunch of reasons.
The worst is when someones on your channel using the same modulation as you are. If he's louder he'll blow your packet out of the water. Keeping your packets short and using max power (cheezy grin) and retrying helps.
But even interferers on different channels will degrade your sensitivity and reduce your range.
None of this is unique to lora modulation.
RC stuff (ELRS, Crossfire, etc.) like the parent post mentions mostly avoids this issue by also using FHSS.
I think the different spreading factors can share a frequency at the same time.
Notably, Wez's flight was done with 2.4GHz LoRa (this was one of his main points, to address naysayers who claimed a 2.4GHz link could never achieve >100km) while Charles is using 900MHz - the same setup that achieved 100km (100mW with a lightly directional antenna on one end, ExpressLRS 50Hz RF mode) would have a theoretical range of >500km on 900MHz. LoRa at longer wavelengths can achieve some incredible feats - like microwatts or nanowatts of power to >1 mile as we saw here. :)
Were the naysayers saying 2.4GHz was impossible at all, or were they saying in a normal setting?
The distinction is quite important because I imagine anything bigger than the tiniest of villages will be swamped with interference..
In particular, people used to plain FSK RC protocols like Spektrum DSM, FrSky D8, Futaba FASST, and similar which tend to failsafe around <1km even with 100mW of power believed that a 2.4GHz link could never do 100km regardless of interference or noise issues.
Also notably, LoRa does tremendously well with real-world noise and interference. Of course any receiver is susceptible to frontend overload, but in "normal" situations I never worry about flying 2.4GHz even in crowded cities, and have never had a failsafe even in some ridiculous situations. As someone above mentions, LTE on the 800MHz band with 900MHz control links are much more of a problem - I've had a failsafe for that exact reason at <200m total distance from takeoff.
Damn, that's quite impressive. Thank you for clarifying!
> Were the naysayers saying 2.4GHz was impossible at all, or were they saying in a normal setting?
The naysayers, even if they aren't strawmen, weren't "saying" anything of note. There aren't any mysteries or questions here: what can be achieved is predetermined by the factors involved. The key figures of merit are power, gain, path loss, noise and receiver sensitivity. Whatever those factors are determines feasible range, and they are all well understood.
Should we be looking for evidence of LoRA style transmissions for SETI? The energy required to get a loud conventional signal to go dozens or hundreds of light years is insane. But it seems like this kind of modulation might allow interstellar range with “only” gigawatt scale power levels.
It’d be an ideal way to create a signal with the potential to reach, say, a large fraction of the galaxy without requiring a Dyson swarm to power the transmitter.
Of course there are so many modulations this might be combinatorially impossible unless we can make some rational guesses about what a rational intelligence would use for interstellar communication if they wanted others to notice.
You get added again out of a spread spectrum signal and it helps with interference too but the main issue is that with the power spread out, it's much harder to detect at long distances. It becomes even harder to detect when you don't know the spreading sequence. You may "see" some RF power in the spectrum but without knowing the chip sequence, it might as well be noise. LoRa uses a type of spread spectrum method that is even harder to detect if you aren't expecting it. There's an example of someone using it for a 100km data link for an RC plane but you know what works at an even longer range? CW, aka Morse code, at low frequencies (1-30MHz). They can wrap around the earth even at relatively low power. LoRa is nice because it's modern and you can get nice data rates that are appropriate for machine-to-machine communications but spread spectrum waveforms will never be useful when you just want any signal to go as far as possible.
Yeah.. but with 900Mhz won't there be a lot more interference from 4G towers though? I remember there were a bunch of "Crossfire failsafed near a 4G tower because i chose the wrong region" videos.
Doesn't LoRa have a pretty constrained duty cycle? I read the regs on LoRa (for the US at least) and found that you're required to limit transmission frequency to 1% (36s/h). I'm guessing you'd have to go well past that to operate an RC aircraft in the conventional manner. If you're just commanding waypoints I suppose you could be compliant.
IIRC the 1% duty cycle is an EU restriction. The US 915 MHz band has a 400 millisecond dwell time restriction instead.
LoRa is just a protocol, there are no rules inherent to the protocol itself. Certain LoRaWAN networks have rules around how often you can send messages through them, but RC links don't use LoRaWAN.
Per channel. Frequency hopping mostly bypasses this issue.
I think if you go to 500khz you can transmit all you want. It's only 125khz that has that limit.
100km @ 100mW is not impressive when it's attached to a 30dBi antenna...
It wasn't a 30dBi antenna for the control link, but a light moxon with 5.98dBi according to its datasheet. Standard 5.8GHz analog video, on the other hand, did require a ton of gain.
Why not?
The amount of "db" to make a link is called the link budget. 20dbm (100mw) + 30db antenna is equivalent to over 300 WATTS out of a 3db standard 1/4 wave monopole antenna (assuming the rx side is equal).
30 dBi means that much reduction from every other direction, unless you're doing active beam forming.
So which directions you excluding?
russian FPV drones are using LORA radios. ru were very proud of this fact until Ukrainians switched from jamming to hijacking control channel.
I transmitted this way using a raspberry pi and a basic filter and antenna and was heard one day in New Zealand from Ohio. https://github.com/JamesP6000/WsprryPi
But then you are transmitting at 0,5 bit / second?
The ideal range/speed is somewhere between 100 and 1000MHz, ideally you would want that entire range.
But how do you build a band pass filter that can cover the entire range?!
Wow! What were the specifics?
Pretty much that project, a Raspberry Pi, a ham radio license, a small filter board from QRP Labs, an MFJ antenna tuner, and then a long wire over the roof of a single story home. No other amplifier. The New Zealand reception was just once. I did somewhat more often get heard in Germany or Brazil, and quite regularly east of the Mississippi and west of the Appalachians, but never closer than about 50 miles, that is, no one ever heard me in town. This was across like 40 meters and 10 meters I think about seven or eight years ago. I ran the beacon all the time for a couple of years.
So, you thought that machine was airgapped....
i bet this technique has been used for years by the NSA.
Sleeper video of the year so far. All of that was fantastic.
And imagine that, quality content without “be sure to like and subscribe!”.
From the github repo:
>Because we rely on harmonics and aliasing, the primary frequency components emitted by your microcontroller are going to be in portions of the RF spectrum where RF transmissions are banned. Please filter your output or perform your tests in an area where you are unlikely to leak significant RF. The overall EIRP output is genreally ≪300uW across the whole spectrum spread out over hundreds of emission frequencies, but there is virtually no way a device deliberately transmitting on these frequencies could ever pass FCC part 15 compliance, even with filtering.
While low over the whole spectrum, on the desired harmonics the power is higher, just how far does it go. I'd also worry about getting into trouble experimenting with such things. Some wilderness areas utilize radio to track and monitor wildlife and resources, hopefully it wouldn't interfere with those. Though maybe it is not easy to get in trouble doing so because of the low power output.
This could be very useful as a covert channel on a hacked device, there goes your air gap!
wow, that is a great hack