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!
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.
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.
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.
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”
[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.
Morse code used to be widely used around the globe. Before voice transmissions were possible over radio, Morse code was all the rage. Nowadays, it’s been replaced with more sophisticated technologies that allow us to transmit voice, or data much faster and more efficiently. You don’t even need to know Morse code to get an amateur radio license any more. That doesn’t mean that Morse code is dead, though. There are still plenty of hobbyists out there practicing for the fun of it.
[Dan] decided to take a shortcut and use some modern technology to make it easier to translate Morse code back into readable text. His project log is a good example of the natural progression we all make when we are learning something new. He started out with an Arduino and a simple microphone. He wrote a basic sketch to read the input from the microphone and output the perceived volume over a Serial monitor as a series of asterisks. The more asterisks, the louder the signal. He calibrated the system so that a quiet room would read zero.
He found that while this worked, the Arduino was so fast that it detected very short pulses that the human ear could not detect. This would throw off his readings and needed to be smoothed out. If you are familiar with button debouncing then you get the idea. He ended up just averaging a few samples at a time, which worked out nicely.
The next iteration of the software added the ability to detect each legitimate beep from the Morse code signal. He cleared away anything too short. The result was a series of long and short chains of asterisks, representing long or short beeps. The third iteration translated these chains into dots and dashes. This version could also detect longer pauses between words to make things more readable.
Finally, [Dan] added a sort of lookup table to translate the dots and dashes back into ASCII characters. Now he can rest easy while the Arduino does all of the hard work. If you’re wondering why anyone would want to learn Morse code these days, it’s still a very simple way for humans to communicate long distances without the aid of a computer.