radio

569 Articles

Transmitting HD Video From A Raspberry Pi

March 28, 2015 by 30 Comments

It’s been a few years since the RTL-SDR TV Tuner dongle blew up the world of amateur radio; it’s a simple device that listens in on digital television frequencies, but it’s one of those tools that’s just capable enough to have a lot of fun. Now, we have a transmitting dongle. It’s only being used to transmit live HDTV from a Pi, but that in itself is very interesting and opens up a lot of possible builds.

The key piece of hardware for this build is a UT-100C DVB-T modulator. It’s a $169 USB dongle capable of transmitting between 1200-1350 MHz, and with a special edition of OpenCaster it’s possible to transmit over-the-air TV. There’s no amplifier, so you won’t be sending TV very far, but it does work.

On the Raspberry Pi side of the build, the standard camera captures H.264 video with raspivid, which is converted to a DVB compliant stream using ffmpeg. These are well-worn bits of software in the Raspberry Pi world, and OpenCaster takes care of the rest.

While this seems like the perfect solution to completely overbuilt quadcopters, keep in mind transmitting on the 23cm band does require a license. Transmitting in the UHF TV bands is a bad idea.

Horn Antenna

Building A Horn Antenna For Radar

March 22, 2015 by 15 Comments

So you’ve built yourself an awesome radar system but it’s not performing as well as you had hoped. You assume this may have something to do with the tin cans you are using for antennas. The obvious next step is to design and build a horn antenna spec’d to work for your radar system. [Henrik] did exactly this as a way to improve upon his frequency modulated continuous wave radar system.

To start out, [Henrik] designed the antenna using CST software, an electromagnetic simulation program intended for this type of work. His final design consists of a horn shape with a 100mm x 85mm aperture and a length of 90mm. The software simulation showed an expected gain of 14.4dB and a beam width of 35 degrees. His old cantennas only had about 6dB with a width of around 100 degrees.

The two-dimensional components of the antenna were all cut from sheet metal. These pieces were then welded together. [Henrik] admits that his precision may be off by as much as 2mm in some cases, which will affect the performance of the antenna. A sheet of metal was also placed between the two horns in order to reduce coupling between the antennas.

[Henrik] tested his new antenna in a local football field. He found that his real life antenna did not perform quite as well as the simulation. He was able to achieve about 10dB gain with a field width of 44 degrees. It’s still a vast improvement over the cantenna design.

If you haven’t given Radar a whirl yet, check out [Greg Charvat’s] words of encouragement and then dive right in!

How Cheap Is Cheap?

March 14, 2015 by 21 Comments

The Nordic Semiconductor nRF24L01 is the older sibling of the nRF24L01+ and is not recommended for new designs anymore. Sometimes, if you’re looking for a cheaper bargain, the older chip may the way to go. [necromant] recently got hold of a bunch of cheap nrf24l01 modules. How cheap ? Does $0.55 sound cheap enough?

Someone back east worked out how to cost-optimize cheap modules and make them even cheaper. At that price, the modules would have severe performance limitations, if they worked at all. [necromant] decided to take a look under the hood. First off, there’s no QFN package on the modules. Instead they contain a COB (chip on board) embedded in black epoxy. [necromant] guesses it’s most likely one of those fake ASICs under the epoxy with more power consumption and less sensitivity. But there’s a step further you can go in making it cheaper. He compared the modules to the reference schematics, and found several key components missing. A critical current set resistor is missing (unless it’s hiding under the epoxy). And many of the components on the transmit side are missing – which means signal power would be nowhere near close to the original modules.

The big question is if they work or not ? In one test, the radio did not work at all. In a different setup, it worked, albeit with very low signal quality. If you are in Moscow, and have access to 2.4Ghz RF analysis tools, [necromant] would like to hear from you, so he can look at the guts of these modules.

Thanks to [Andrew] for sending in this tip.

Measuring Filters And VSWR With RTL-SDR

March 14, 2015 by 5 Comments

Once again the ubiquitous USB TV tuner dongle has proved itself more than capable of doing far more than just receiving broadcast TV. Over on the RTL-SDR blog, there’s a tutorial covering the measurement of filter characteristics using a cheap eBay noise source and an RTL-SDR dongle.

For this tutorial, the key piece of equipment is a BG7TBL noise source, acquired from the usual online retailers. With a few connectors, a filter can be plugged in between this noise source and the RTL-SDR dongle. With the hardware out of the way, the only thing remaining is the software. That’s just rtl_power and this wonderful GUI. The tutorial is using a cheap FM filter, and the resulting plot shows a clear dip between 50 and 150 MHz. Of course this isn’t very accurate; there’s no comparison to the noise source and dongle without any attenuation. That’s just a simple matter of saving some scans as .csv files and plugging some numbers in Excel.

The same hardware can be used to determine the VSWR of an antenna, replacing the filter with a directional coupler; just put the coupler between the noise source and the dongle measure the attenuation through the range of the dongle. Repeat with the antenna connected, and jump back into Excel.

Simple Superheterodyne SW Receiver Harks Back Almost 100 Years

March 8, 2015 by 13 Comments

Early radio receivers worked on a principle called Tuned Radio frequency (TRF), patented in 1916. They weren’t very easy to use, requiring each stage to be tuned to the same frequency (until ganged capacitors made that a bit easy). The Superheterodyne design, devised in 1918, was superior, but more expensive at that time. Cost considerations led adoption of the Superhet design to lag behind TRF until almost 1930. Since then, until quite recently, the Superhet design has been at the heart of a majority of commercial radio receivers. Direct Conversion Receivers were devised around 1930, but required elaborate phase locked loops which restricted their use in commercial receivers. The point of all this background is that the Superhet design has served very well for more than 90 years, but will soon be rendered redundant once Software defined Radio (SDR) becomes ubiquitous. Which is why this study of the simple Superheterodyne shortwave receiver deserves closer study.

[Dilshan] built this two transistor and two IF transformer based superheterodyne radio designed to receive 13m to 41m bands. The whole build is assembled on a breadboard, making it easy to teach others to experiment. [Dilshan] offers very useful insights into antenna, rod coil and IF transformer parameters. To dive in to Radio architecture, check this post on Amateur Radio. And if you would like to get a closer look at Antique Radios, check this post on Restoring Antique Radios.

Retrotechtacular: Crystals Go To War

March 3, 2015 by 51 Comments

More than one of our readers suggested we highlight this beautifully-shot process documentary about the laborious and precise manufacturing of piezoelectric quartz crystals in the early 1940s. Just a few years later, Bell Labs would perfect a method of growing synthetic crystals, sending droves of brave men and daintily-handed women from the Reeves Sound Laboratories to the unemployment line.

Early radio equipment relied upon tuned or L-C circuits for clocking. These were prone to drift by a few kHz, which prompted the use of crystal oscillators for stable frequencies in the 1920s. The lives of our armed forces and those of our WWII allies depended on reliable communication equipment, so the crystal oscillators they used were top shelf, produced by hand from Brazilian crust.

Continue reading “Retrotechtacular: Crystals Go To War” →

Whiteboard Clock

Whiteboard Clock Draws The Time

February 23, 2015 by 3 Comments

[Maurice] recently built a clock that draws the time (Google Doc) on a white board. We’ve seen plenty of clock hacks in the past, and even a very similar one. It’s always fun to see the different creative solutions people can come up with to solve the same problem.

This device runs on a PIC16F1454 microcontroller. The code for the project is available on GitHub. The micro is also connected to a 433MHz receiver. This allows a PC to keep track of the time, instead of having to include a real-time clock in the circuit. The USB connector is only used for power. All of the mounting pieces were designed in OpenSCAD and printed on a 3D printer. Two servos control the drawing arms. A third servo can raise and lower the marker to the whiteboard. This also has the added benefit of being able to place the marker tip inside of an eraser head. That way the same two servos can also erase the writing.

The communication protocol for this systems is interesting. The transmitter shows up on [Maurice’s] PC as a modem. All he needs to do to update the time is “echo 12:00 > /dev/whiteboard”. In this case, the command is run by a cron job every 5 minutes. This makes it easy to tweak the rate at which the time updates on the whiteboard. All communication is done one-way. The drawing circuit will verify the checksum each time it receives a message. If the check fails, the circuit simply waits for another message. The computer transmits the message multiple times, just in case there is a problem during transmission.