Excellent IEEE Foothill Meeting on Software Defined Radio Held October 5
|October 21, 2013||Posted by COMauthor under COMSOC||
The IEEE Foothill Section was pleased to welcome Dr Jeffrey Pawlan to DeVry University Pomona on Saturday October 5, 2013 for a technical presentation on “Software Defined Radio (SDR)”. Dr Pawlan is one of this current year’s Distinguished Lecturers selected and his visit was sponsored by the IEEE Microwaves Theory and Techniques Society. His presentation to us was the first on his recent tour of IEEE sections in Southern California.
Dr Pawlan noted that his interest in high frequency phenomena and equipment began over 53 years ago, when he first attended the International Microwave Symposium in Los Angeles. His interest in RF and microwave equipment developed, along with his professional career. He is a licensed amateur radio operator (WA6KBL), broadcasting in the 10MHz band, from a base station set up at his home in San Jose, CA. (Just look for the 71 foot tall antenna, with the Yagi array mounted on top.)
His presentation covered some of the highlights on the path of radio development; in particular radio receivers, from the earliest broadcast days until now. As Dr Pawlan views this path, there are four generations. The problem has always been to determine how a radio receiver should deal with a limited bandwidth in a finite electromagnetic frequency spectrum. For Generation 1, this meant that a mixer was used, together with a local oscillator(LO), to get an IF signal. This was amplified to get the audio signal.
Generation 2 began with the inventive work of Edwin Armstrong, whose work included the development of the frequency modulation techniques(FM) to reduce the effects of noise, and the superheterodyne process. Now the LO signal was mixed with the signal received from the antenna, together with another LO signal, phase shifted by 90 degrees. The sum of these two signals (called I/Q for in phase and quadrature phase signals) was processed. Image rejection ratios were now investigated. This was equivalent to a Hilbert transform; all done to obtain a better signal to noise (S/N ) ratio.
Generation 3 involved the use of separate analog-to-digital convertors (ADC) for the each of the I / Q channels. In effect, the processing is taking place on the IF in the VHF range. Then software (SW) was used to implement the Fast Fourier Transform (FFT) routines for the combined digitized complex I/Q signals. This FFT spectrum would show the expected frequency band, and well as the image band. The next step here is to adjust the amplitude and phase (A/Φ) of one of the I/Q channels (via the LO feeding that channel) to reject as much as possible of the undesired image band.
Dr Pawlan demonstrated that by way of the custom software “WINRAD”. This was developed by a Digital Signal Processing (DSP) experienced engineer and amateur radio enthusiast Alberto (12PHD) in Italy. Here we were shown how small errors in A/Φ were spotted in the FFT of the received spectrum, and nulled out. This real time demonstration was executed by setting up an Internet link from his home receiver (RF front end) in San Jose, using a static IP address. Then, he used this Internet transported signal to drive a SDR sitting on a cart here in DeVry Pomona . In turn, this SDR (on cart) output was sent to the sound card in his laptop computer. Now, errors in non-uniform amplitude and nonlinear phase in the I/Q channels could be examined. By tweaking the A O/ in one channel, definite changes in the signal tone could be noted and heard by our IEEE Foothill audience. This was also displayed on the projector screen as real time output from WINRAD.
We then continued with a discussion of Generation 4 for SDR. Here, after the receiving antenna and a broadband, low noise amplifier, the signals can be converted via ADC into a buffer for all-digital processing. This includes digital mixers, digital oscillators, and digital filters. As Dr Pawlan noted, a professor in San Diego has reported development of an oscillator using FPGA (field programmable gate array) that has a phase noise -180dB below the carrier signal.
Dr Pawlan spent a short segment of his presentation discussing some commercial parts for a SDR. These included the Linear Technology devices LTC 2206/2207/2262 high speed, 105 Msps, 14-to 16-bit ADC. A follow up question would be to examine whether off the shelf parts like this could be productively used to supplement a SDR design for a receiver to detect a weak BW limited signal in a high interfering noise environment. Many of our attendees no doubt have other questions for a future visit of Dr Pawlan to our IEEE Foothill Section.
All our attendees agreed that this was a good Section meeting. Our thanks to Dr Pawlan for making his trip to DeVry Pomona so worthwhile and informative for our members.