Author Topic: Project: The $5 feed (beta testers, contributors wanted)  (Read 20037 times)

Offline sean

  • Full Member
  • ***
  • Posts: 164
Project: The $5 feed (beta testers, contributors wanted)
« on: July 14, 2014, 05:50:26 PM »
After life/work/family caused me to take a bit of a break from LiveATC activities, my interest was renewed over the past year when I discovered some free time. What a difference a day makes, as they say.  Since my heyday with LiveATC the Raspberry Pi has come along. Couple that with $8 RTL-SDR dongles, and it sparked my creativity and curiosity:

Was it possible to make a $5 feed?

I should clarify.. I mean $5/channel, not $5 all-in. Here's how I arrive at my baseline numbers for a "super feed" like we used to build:

PC: $250
Delta sound card: $199
8 Uniden Scanners: $250
Total:  $699 / 8 channels = $87/channel

Please.. if you have better numbers, post them!

So, I set out to figure out how we could reduce this $87/channel number to something more reasonable. How could we set up "super feeds" with 8 or 16 channels for much lower cost?  Well, we're close - but I need your help to nail down the specifics and turn this into a reality that everyone can enjoy.

As a teaser:  We're at 16 channels now for about $73 TOTAL - everything except the antenna system/splitter/preamp.  That's about $4.60/channel.  Lot's of qualifications to go along with that number, but as of now if we had a site that had 16 channels to stream, we could do it for about $5/channel, and for an operating cost of roughly $8 per year in electricity.

The next step in the process is to ask for a test site from a current feeder that has a site close to a Class B airport with strong signals. I want to start out in a known, strong RF environment so that we can rule out any oddities before going any further. If you're a current feeder, close to a multi-channel, strong RF Class B airport, please send me a message. I will provide equipment, instructions, and ultimately the entire setup will be yours in exchange for cooperation and testing of this equipment.

Over the next several months I plan to roll out the configuration in stages so that everyone can take advantage of the improvements. Some of the potential benefits of this project are:

1.  Much lower initial investment
2.  Much lower operating costs (electricity)
3.  Increased stability (Linux)
4.  Fully programmable (no need to program scanners)
5.  Ability to capture ADS-B, ACARs, and ATC audio if desired
6.  Ability to scan multiple channels like a typical scanner
7.  Software/hardware is 95% pre-configured, only minor changes to get on the air
8.  Squelch and audio level are set automatically - all channels have the same loudness
9.  Very small equipment footprint compared to a traditional PC
10. Works for MilAir (225-400) too.
11.  Up to 16 individual channels for $73!

Your thoughts, comments, questions are encouraged in this thread.  I'd love to make this a group project with other contributors.  We're much stronger as a team!

Sean
« Last Edit: July 14, 2014, 06:03:35 PM by sean »



Offline hayek

  • Newbie
  • *
  • Posts: 27
Re: Project: The $5 feed (beta testers, contributors wanted)
« Reply #1 on: July 23, 2014, 08:18:03 PM »
http://www.milaircomms.com/raspberry_pi_scanner_streaming.html
Streaming Your Scanner Audio Using a $35 Raspberry Pi Computer

Offline sean

  • Full Member
  • ***
  • Posts: 164
Re: Project: The $5 feed (beta testers, contributors wanted)
« Reply #2 on: July 25, 2014, 10:31:55 PM »
Thanks... I am familiar with the work that user has done.  This is a little different, though - we're using dongles instead of scanners, and up to 8 channels/feeds per dongle.  So no sound adapters, no scanners, and a lot of other features. 

Sean

Offline av8tor172

  • Full Member
  • ***
  • Posts: 115
    • Military Air Communication Frequenices, Live MilAir Streaming, Recorded MilAir Comms.
Re: Project: The $5 feed (beta testers, contributors wanted)
« Reply #3 on: July 26, 2014, 07:44:53 AM »
While I think your project is a worthwhile project I do have a question about your cost calculations. 

Under a traditional setup (PC & Scanners) you have "8 Uniden Scanners: $250".  Does that means you're getting scanners at $31.25 each?  Where???

My next question, are you using 1 RTL Dongle per frequency?  If so does one Raspberry Pi have the power to handle 8 RTL Dongles and DarkIce to stream 8 audio streams.

Very interesting project which I'll be watching...

Thanks
George
www.MilAirComms.com

Offline InterpreDemon

  • Hero Member
  • *****
  • Posts: 514
Re: Project: The $5 feed (beta testers, contributors wanted)
« Reply #4 on: July 26, 2014, 03:06:47 PM »
What is the passband of each dongle and what kind of front-end filtering are you doing to eliminate images?

Offline sean

  • Full Member
  • ***
  • Posts: 164
Re: Project: The $5 feed (beta testers, contributors wanted)
« Reply #5 on: July 28, 2014, 08:05:28 PM »
While I think your project is a worthwhile project I do have a question about your cost calculations. 

Under a traditional setup (PC & Scanners) you have "8 Uniden Scanners: $250".  Does that means you're getting scanners at $31.25 each?  Where???

My next question, are you using 1 RTL Dongle per frequency?  If so does one Raspberry Pi have the power to handle 8 RTL Dongles and DarkIce to stream 8 audio streams.

Very interesting project which I'll be watching...

Thanks
George
www.MilAirComms.com

George,

I was being conservative on the costs so that it didn't appear that I was  trying to fit the budget into the goal.  I've purchased lots of BC350 scanners on ebay before for under $20 a piece, usually from a police department that swaps them out or some other large agency.  Agreed, though, it's tough to find them in quantity often.

It's custom software, written specifically for this purpose, and includes the streaming libraries (lame and shoutcast) natively in the program. The current configuration will handle 2 dongles per Pi, but we're using a powered USB hub to support the dongles, wifi adapter, etc.  The hub is included in the calculation, though.  So, if you could find the right mix of frequencies, you could stream 16 different frequencies from a single Pi and 2 dongles. 

Alternatively you could stream 8 frequencies and run dump1090 for ADS-B on the other dongle, or even acarsdec.  Heck, with 2 Pis you could do 75% of the ACARS frequencies simultaneously, ADS-B, and 8 channels of audio - for around $100-$125.

The software would technically work with MilAir, although I haven't tested it.  It's running at Oshkosh now with good results - that's the first real field trial to date.  If you're interested in trying it with MilAir, send me a PM.

Sean

Offline sean

  • Full Member
  • ***
  • Posts: 164
Re: Project: The $5 feed (beta testers, contributors wanted)
« Reply #6 on: July 28, 2014, 08:09:47 PM »
What is the passband of each dongle and what kind of front-end filtering are you doing to eliminate images?

There are many different brands of dongles, with different chipsets.  They are definitely not all created equally.  With a good dongle (decent chipset), 3.2 MHz is possible.  So, with 2 dongles you could cover 6.4 MHz of continuous BW, or two chunks of 3.2 MHz.  I've looked at a lot of Class B airports, and 5 dongles would cover most of them - seems like a lot, 3 full Pi setups, but if you consider that you'd have $200 invested and 16-20 separate streams, plus either ACARS or ADS-B, it's not a bad deal.

As for front end filtering, it hasn't been an issue that I've encountered.  I've tried the various bandpass filters, but they haven't made much of a difference - only because they weren't really needed.

Sean