In my last blog post, I completed a blog series on how to build your own simple phased array beamformer, also called an electronically steerable array (ESA). That beamformer was based on the low-cost ADALM-PLUTO, which hopefully made it accessible to everyone interested in this topic. But it was only two antenna elements – and that limits the number of things we can do with it.
In this new blog series, I want to dig into a new phased array exploration platform that Analog Devices just released: EVAL-CN0566-RPIZ, more commonly known as the “Phaser.” The Phaser is going to take us one step closer to a more practical phased array system. Its feature set will allow experimentation with beamforming and radar, principles that we couldn’t cover with the 2-element Pluto digital beamformer.
For those new to this topic, I should start with a very quick explanation of what a beamformer is and why you might want one. Much could be said about both topics! But very quickly: the goal is to steer an antenna “beam” without physically moving the antenna. Most everyone is familiar with a rotating radar dish antenna that produces something like this:
But now imagine that the sweep is complete, not by mechanically moving the antenna, but instead by electrically steering the beam. This would allow us to rapidly (almost instantaneously) position the beam wherever we wanted. And even to create multiple antenna beams, each focused on a different target.
Ok, so now we want a beamformer. But how do we make one? The basic concept is to set multiple antennas up in an array (either in a line or spaced along a 2-D plane), and then delay the timing of each of those antenna elements such that they constructively sum in one direction, and destructively sum in all other directions. The concept is not as hard as it may sound and is covered in this blog entry.
But while the basic concept is easy to understand, the implementation can be difficult. Understanding all of the issues that arise with electrically steerable arrays (ESAs) can take a lifetime of study! I think that the best way to learn this topic is to physically build one and experiment with it. As long-time Analog Devices’ designer Paul Brokaw once said: “Simulation will answer the question you asked. The lab will answer the question you didn’t know to ask.”
So how do we get started learning and experimenting with beamformers? The ADALM-PLUTO hopefully gave us some insight into the operation of a beamformer. But at only two elements, we could only do so much. Therefore, Analog Devices has recently launched the “EVAL-CN0566-RPIZ”, more commonly known as the “Phaser.”
The Phaser is an 8-element receive array, with two switched transmitters. It generally operates from 10-10.5 GHz, but by connecting to your own external antenna, it could operate anywhere from 8 to 14 GHz. The block diagram of the complete system looks like this:
The receive antenna array is integrated into the Phaser’s PCB. Each of the 8 receive antennas is the summation of a row of 4 individual antenna elements. So, we end up with an 8-element, linear, uniformly spaced, antenna array.
Each of those 8 receive antennas is individually adjusted for gain and phase at the RF frequency (i.e., ~10 GHz) using the popular ADAR1000 beamformer chips. The ADAR1000 is capable of both transmitting and receiving. But to keep the Phaser’s cost low, we have only implemented the receive portion. The purpose of the ADAR1000 is to adjust the phase and gain of each element’s received signal, such that we can coherently combine the signal for the direction in which we want to steer the antenna.
The output of each of the two ADAR1000s is then digitized by the very versatile ADALM-PLUTO. But since the maximum frequency of Pluto is 6 GHz, we must first mix down from 10 GHz to something within Pluto’s frequency range. This is accomplished with the LTC5548 mixer. And since we’re using a mixer, we need a frequency source to drive the mixer’s LO port. This LO is generated on the Phaser’s PCB using the ADF4159 PLL and HMC735 VCO. The ADF4159 can generate frequency modulated “chirps.” We don’t need chirps for the basic beamforming, but they will be very useful when we talk about using the Phaser for radar applications.
The Phaser can also transmit a 10-10.5 GHz waveform on two ports: OUT1 and OUT2. However, these are switched ports – not phase synchronize channels as the receive side has. The switched channels allow us to construct something called “virtual arrays” – which hopefully we will cover in a future blog.
Putting all of this together gives a compact and easy-to-use phased array experimentation platform:
In the next several blog entries, I’ll dig further into how you can use the Phaser to experiment with a variety of beamforming principles: grating lobes, beam squint, monopulse tracking, null steering, FMCW radar, radar target detection, etc. I’ll show how the Phaser can be a natural beginning to your ESA design and allows for the smooth progression from innovation to prototyping to production. See you in the next blog post here!