Thursday, January 23, 2020

Energy consumption is a hot topic in the world of voice-enabled AIoT devices. With good reason.

Mark Lippett, XMOS, January 23, 2020

Voice shows the fastest adoption of any consumer technology ever. At the current rate of growth, there’ll be a further 1.5 billion new voice-enabled devices in our homes in 2025, with an estimated 5 billion units in use worldwide.

Imagine all these devices powered up and hanging on to our every keyword. At a very rough estimate, those devices will consume 65 TeraWatt hours of electricity a year, simply by being always on, listening for a keyword. That’s almost the equivalent (90%) of the annual output of the world’s largest nuclear power plant. It’s not sustainable. Intelligent IoT systems should enable us to consume less, not more. 

As voice becomes a mainstream requirement and the focus moves inexorably forward to contextual, conversational interfaces, so we’re also seeing a shift in the semiconductor industry, with increasing innovation (and demand) around energy efficient solutions.


Tuesday, January 7, 2020

Non-Linear Transmission Line Comb Generators Part-1: The Phase Noise Problem and Comb Generation

MACOM, January 07, 2020

In this two-part series from MACOM, we will delve into Non-Linear Transmission Line (NLTL) Comb Generators, first understanding the phase noise problem, and understanding a potential solution to the problem. In the second part of the blog series, we will explore NLTL comb generation, compare it to its predecessor comb generation using Step Recovery Diodes and see how the NLTL comb generation approach can enable improved sensitivity and lower bit error rates in communication systems.

Figure 1: Typical Superheterodyne Receiver
The Phase Noise Problem

Let’s start with the problem with circuits requiring low noise performance. Below we see a block diagram of the RF and IF portions of a typical superheterodyne receiver.  A weak signal is received at the antenna – 1) optionally amplified by a low noise amplifier, 2) filtered to reduce the effects of broadband noise and interferer signals whose frequencies may be close to that of the desired signal and then 3) downconverted to a lower, intermediate frequency for further processing. 

In the ideal case, the downconverter mixer mixes the received signal with a single-frequency local oscillator signal.  In the real case though, the local oscillator signal never comprises a single frequency, but is always accompanied by close-in noise sidebands which are generated in the local oscillator signal chain.  Also, the received signal may be accompanied by close-in interfering signals which cannot be completely removed by the band pass filter.