Lights Out!  Guerrilla Radio

by token oppmedia@hushmail.com

Turn that shit up!  In this article, I'd like to share the knowledge necessary to deploy your own remote controlled FM radio station.  Who ever said playing around in a graveyard couldn't be fun?  The goal is to have a box with all of the required components, blending in enough to not draw much attention.  Think of a weather station, a traffic control box, or a remote terminal DSLAM.  How often do most people pay these any mind?  There's no reason that these couldn't potentially be deployed at a remote intersection, or on the side of a highway, or even up on a telephone pole somewhere.

Granted, if you're going to be that ballsy, get some official looking clothes and a work truck, and be a very capable social engineer.  Otherwise, there are plenty of options, like up on the roof or balcony of a large apartment building, a hotel, a park... anywhere is fair game.  The higher, the better.  You can use Google Earth to look at terrain to determine high locations.  The general rule of thumb is that the farther you can see the roofs of buildings, the better.  The usual "educational purposes" disclaimer applies, as well as a warning that unlicensed broadcasts at any useful distance is a violation of New Jersey and Florida state laws, as well as FCC regulations.

What you will need is, obviously, an FM transmitter.  There are plenty of cheap Chinese models available on eBay, as well as Elecsky.  Warner RF and HLLY are reasonable choices brand-wise on the Chinese side of things.  The downside of these cheap little wonders is that they're very prone to "splatter," or broadcast in places on the band that they shouldn't (and that you don't want), so make sure you get a low-pass filter to avoid pissing off the FAA.  The "low-pass" should be 108 MHz and 50 ohms.  I'm not going to outright say that "you get what you pay for" with the Chinese boxes, but they're definitely not as good as the good stuff.  For a bit more, you get a bit less from Aareff feature and power-wise, but they are very well made.  I present these options, but I must tell you there are tons of options out there beyond this.  What you will want will depend on what you want to do.  A cheap HLLY is a good choice for a box you risk losing, but if you're reasonably certain your box will be safe, an Aareff is worth considering.  You may find "kits" out there that include the antenna, cables, and power supply, but be careful.  The antennas tend to be cut to a frequency away from where you want to be.  In radio transmitting, fractions of inches make a difference.  If you'll be near the center of the band (96-102 MHz-ish), you'll be all right usually, but it may make more sense to get the antenna separately.

On the topic of antennas, you obviously are going to need one.  The antenna should be rated for the power you'll be putting out.  Just like speakers, a one watt antenna will not be a good match for a 15 watt transmitter.  Luckily, you have a ton of choices.  The ground plane is fairly standard, but the J-pole is also popular.  You also have to decide what polarization you want.  Vertical tends to be good for cars, horizontal for homes, or circular to get good reception in both (but half the effective power - and expensive).  These are broad generalizations, and if you want to learn more about the pros and cons of all of your options, there are tons of sites that'll do a better job than I can in this basic primer.  Personally, I find a vertical J-pole to be the best overall.  They're cheap, low-noise floor, rugged, some gain, etc.  Google "FM broadcast antenna" for a shitload of resources on how to build, buy, or learn more.

As for cables, connectors, filters, etc... mostly you'll see 75 and 50 ohm options.  Always select 50 ohms for your broadcast equipment.  Do not mix impedances!  I won't detail the full electronics reasons, but the end of the long boring story is "broken transmitter."  Stick to 50 ohms.  75 ohms is for receiving.  Also, you're going to see the terms BNC, NMO, (N connector), etc. in regards to your connector.  Save yourself headaches and stick with one standard.  I like NMO because it's watertight and can do pretty much anything.  Adapters suck - there are extra connections that can fail and there's no good reason to want to try to connect these different standards.  Damn adapters are always the first thing to fail for some reason.  If your transmitter is NMO, get NMO cables, filters, antenna, everything.  If you're dead set on getting something out of standard - like a transmitter with BNC out - you can get cables that have BNC at one end and NMO at the other.  But avoid adapters.

By this point, these are the essentials to getting "on the air."  Transmitter to cable, cable to antenna.  As basic as it gets.  Now you can fine-tune things if you want.  Got an SWR meter?  I'm sure you do!  Not.  But here's the truth on this "fractions of an inch" deal.  At the power levels you're likely to be pushing, it's not going to matter too much.  Find an online "antenna length calculator," punch in your frequency, and cut it as close as possible to that.  If you intend to do more than 15 watts or so, then maybe it might be worth testing the SWR.  The cheap-o China boxes seem to include an SWR meter on some boxes, so that's useful.  For FM broadcast, I personally don't like any more than 1:1.5, but anything under 1:2 is probably safe.  1:1 is considered perfect.  As an example of how narrow these windows can be, I have a thin whip mag-mount that's 1:1 at 88.1 MHz, but 1:1.7 at 88.9 MHz.  In generalities, the "fatter" your actual antenna, the wider the bandwidth, meaning I can also have less than 1:1.5 from 98 to 103 MHz, but that can negatively impact the signal overall.  Don't assume that any "FM broadcast antenna" will work - they may need tweaking.  Higher channels are smaller, 108 MHz is going to be a few inches less than one at 94 MHz, so definitely check out that calculator.  Higher power makes SWR much more significant, too.

So, does the transmitter and antenna work O.K.?  Good!  Let's get to the fun part.  You will need a Raspberry Pi or some other low-power microcomputer, it will need interwebs in some form (4G or Wi-Fi), a power relay module for said Pi, a good sound output, and thick cable, super shielded crazy ridiculous audio cable... the harder to handle, the better.  Interference from RF will be an issue, and feedback can be a problem, so thicker cables from the sound card to the transmitter are very highly recommended.  In addition, get some ferrite chokes for pretty much every few inches of every involved wire.  They're cheap and they really help with interference/feedback in the system.  Not necessary, but very highly recommended after lots of hair pulling.  We will also need some kind of enclosure, which will depend heavily on where you intend to deploy this box.  Look into various "industrial" enclosures that can be easily adapted.  I'm partial to boxes meant to store weather station equipment, as they have an excuse to have technical-looking equipment mounted outside of them, as well as not drawing attention for having solar panels.

Now, this is the optional stage.  Is this going to be on the power grid?  If not, you have to determine how to get power to the box.  Wind can be good in some areas, solar in others.  You will want deep-cycle batteries as well.  I recommend two golf cart batteries.  They're six volts each, but when connected in series they provide what I think is a very good cost per amp hour of capacity.  Any deep-cycle battery will work fine, though, so long as your power source feeds them a decent charge every so often.  Also, is the box going to be mounted off the ground?  If not, some concrete may be worth considering.  A bag or two of cheap instant concrete and cement screws will look a lot better and prevent problems with the box rusting.  Remember, the more official this looks, the less it'll be looked at or messed with.  Look at stuff in your area and look at the stuff meant to blend.  Study these, they're your goal.  Unfortunately, blending makes it hard to get a lot of distance, so unless that ground level spot is on a big hill, you won't get more than a mile or two.  Every extra foot high you can get that antenna makes a difference.

What about securing the box?  Get some padlocks.  I'd also recommend sandbags to weigh it down, especially up on a roof or somewhere windy (after placement, of course).  What about mounting the antenna?  You'll need some kind of mast, as well as some U-bolts if you want the mast and enclosure together.  There are pros and cons to this, but ultimately it's up to you.  Amateur Electronic Supply (AES) sells 35 foot fiberglass masts that could potentially be used, but any pole or strong pipe will do the job just fine.  If you mount the mast separately, consider a post digger and cement to keep it secure, at least three feet deep.  The general rule is one-third height above ground below, so a 10 foot pole should be 3.3 feet deep, but this rule can be bent a bit.  Be reasonable.  Don't do 30 foot poles a foot deep.  The higher your mast, the more you'll need to consider lightning protection, so I' d keep the heights fairly low.  It's still not technically "safe" at any height, but I'm too lazy to deal with that risk most of the time.  My four antennas at 20 to 25 feet have yet to be struck in over five years, though I'd "cheat" and bury a copper wire connected to the ground of the antenna at the low-pass.  A bare copper wire a shovelful down isn't a bad idea, but it is not even close to proper.  Google "RF lightning protection" if you care.  There are other ways to mount an antenna, but you're on your own for creative ways to do that.  I can't cover everything.

Other odds-and-ends, thick wires, caulk, drill bits the width of your coaxial cable, etc.  There's a lot of little things and I may be missing things.  Common sense will help fill the gaps, and hacker ingenuity makes some stuff optional, so consider this a guideline or a framework to work from.

With your Pi (or other small low-power computer), you will need a few things - an SSH server and an application capable of playing streaming audio.  Of course, you can get fancy and set up scripts to do everything automatically upon receiving a text or email, but for now we're keeping it fairly simple.  The relay module will be used to control the power to the transmitter, so depending on whether you have grid power or battery power, the method will likely be different.  For this article, we'll focus on battery power since it covers more ground, and the grid power stuff should be self-explanatory.  Most transmitters are 12 volts, so that makes things easy.  The Pi is 5.5 volt, so that causes somewhat of an issue.  Luckily, it's a simple issue.  Buy any micro-USB car charger that outputs at least one amp for the simple solution.  If you want to get a voltage regulator and read up on USB wiring diagrams, be my guest.  It's much more proper, so at least consider it.  Ugly solution - get some alligator clips and clip a positive wire to the tip, and a negative wire to the outside of the plug.  Is your power coming from the batteries?  Well, if they're the six volt golf cart batteries, hook them together.  Easy to do.  Run a wire from the negative on one battery to the positive on the other.  Now you have one giant 12 volt battery.  Power is run from both batteries as if they're one.

O.K., so now you have some power.  I recommend testing the power brick for the transmitter before you move on, even if you won't use it.  We need to know the polarity.  Do you have a multimeter?  Great, plug in the brick and test the polarity.  If you don't, be creative.  A spare motor can be used to test it... an LED, a speaker.  Consistent results, such as same direction of spinning or the speaker "pushing" should mean that the polarity matches when testing.  Polarity is very important with DC power.  So note the polarity in the first test, unplug the power brick, and snip off the end with a few feet of wire.  Strip 'em and hook 'em up to the relay module opposite the lines from the battery.  Activate that switch and test polarity again.  If wrong, turn off the switch and reverse the wires.  That'll work.  Do note the output voltage on the power brick.  If it's not 12 volts, then consider a voltage regulator... unless it's close.  13.8 volts is pretty much 12 volts and 10 volts is close enough.  There's some tolerance here.

Now, if you have a purely AC-powered transmitter (I do not recommend this unless you're doing all grid power), you will need a pure sine wave power inverter.  Modified sine wave inverters introduce a lot of noise into the system.  This only applies to battery powering AC transmitters, though.  A real pain in the ass.  The relay modules can do both DC and AC, so that's all good.

So, what now?  Run power to the devices.  Get battery clamps or rings, or whatever they're called, and hook power to the wires.  Polarity matters!  Consider getting a "power block" for an easier way to wire things instead of a nest of wires at the batteries.  You can solder all you want if you want, but to make the entry barrier low, I'm trying to write these instructions so anyone can do it with minimal tools or skills.  Clips and clamps work just fine in a pinch.  It can get you started nice and quick.

You should test this all out.  Is everything getting power and working right?  Cool!  If not, figure it out.  We're moving on to setting up the infrastructure here.  How do you stream to that puppy?  Well, you have choices here.  Icecast can run wherever you want, even on the Pi itself.  Whatever you do, make sure the ports are open for Icecast.  You can find countless tutorials on setting up Icecast and streaming to it online.  Your goals are to stream to your Pi, so don't publicly list it and limit it to however many connections you need for how many of these boxes you build.

Once this is all good to go, well... let's test it out!  Make sure your Pi has SSH running, make sure you never power on your transmitter without the antenna connected, ever.  Make sure there's a nice fat audio cable from the Pi to the transmitter with plenty of ferrite chokes clipped on there.  In an ideal world, the Pi will already boot with all the features you'll need - SSH, Internet, Icecast, etc.  You will not be there to start these services, so get your startup services in order now.  SSH into your Pi and point it to the stream address after turning on the correct power relay for the transmitter.  You should be hearing whatever you're streaming over the radio.  Reboot and try again to make sure that your startup services work as they should.

Do you have the local test working locally?  Great - though it's usually not smooth to deploy it "in the field."  You will want some form of dynamic IP updater in most instances, such as DynDNS or No-IP.  You'll need an address to access the Pi from a remote location, one that doesn't change.  These will help there.  Also, since there's not a lot you can do about volume levels once deployed, set your volume levels now.  Play a standard MP3 or Ogg or whatever and turn on the transmitter.  Jump in your car and compare your station to your local commercial conglomerate; you want to be around the same level.  Too high and you'll overmodulate, bleeding over and causing stereo to cut in and out.  Too quiet and you'll be, well, too quiet.  This is also a good time to check around for a good frequency to broadcast on.  You want to be at least two away from existing stations.  101.3 MHz may be open, but if 101.1 MHz and 101.5 MHz have stuff on them, you'll probably piss them off.  Not good.  radio-locator.com can help you find good frequencies to start with.  Do this where you plan to deploy.  A station might be empty a mile away, but may be booming up on the hill.  That's a problem.  Once you find a station, we should start working on the box itself.

Your enclosure, if it's like what I like, is a bog old weatherproof box with plenty of room inside for everything needed.  Grid powered boxes can be much smaller.  You will need to drill holes in the box to run the coaxial wire and wires for the power feed.  You can find boxes with wire holes pre-built, but either way, run the wires through the box.  If you're mounting the mast to the box, drill some holes for the U-bolts, one near the bottom and one near the top to stabilize it.  At the top of the mast will be whatever antennas you need.  Wi-Fi or 4G network antenna can go right below the transmit antenna, if you want external antennas for Wi-Fi/4G.

If you decide to go with solar/wind, consider where to place these.  I like mounting solar right on top, which needs more holes drilled.  The caulk is to reseal these holes once they're in use to make them not leak or rust.  Same with the wire holes.  Caulk is cheap, so go nuts.  Once the holes are drilled, you can start putting stuff in there if you want (or do it at the remote site - these boxes get damn heavy).  Pre-load and hook up everything to test it.  If it works and you can connect to it and everything, cool.  Take out the batteries and load it into a truck or van.  But seriously, test everything.  The goal is to never have to look at or see this box again.  Also, if mounting the mast to the enclosure, wait to do that until at the deploy spot.  Like the batteries, it will make it hard to move.  If using solar panels, make sure they angle south or west, depending on location.

Let it sit in the sun or wind for a day, and when you just can't wait anymore, start streaming to the Icecast server, SSH to the Pi, turn on the power relay to the transmitter, and play your stream in MPlayer or whatever you choose for said stream.  Turn on your radio and, if you're within range of your deployed box, enjoy being "on the air."  A good charge should net you hours and hours of broadcasting, depending on your wattage.  Ten watts will go all night with good golf cart batteries, though don't forget to turn off the power relay to the transmitter when you're done.  The Pi is not very graceful in "low-power" situations; consider a watchdog for it.  Don't run those batteries totally dry - it'll damage them, likely lock up the Pi, and is not good for the transmitter.

Why is air free when airwaves are not?  Why should they belong exclusively to the highest bidder?  I hope that I've at least inspired some thought with this article.

Please help deploy these boxes by prisons in your area.  Prisons are the best bang for your buck, lots of people in a small area, all with radios and starved for entertainment.  Just sayin'.

Happy hacking!