Skip navigation
1 2 3 4 Previous Next

Analog Dialogue

93 posts

Last month, a horse named Nyquist won the 142nd Kentucky Derby. Of course, I immediately thought of the groundbreaking engineer and communications theorist Harry Nyquist, whose work is fundamental to data conversion
applications. He’s a true engineering celebrity of the sort that I wrote aboutin an earlier Editor’s Note. It was a thrill to see the name of one of our engineering heroes being honored in the mainstream international media.


Harry Nyquist came to the US from Sweden in 1907 at the age of 18, and promptly earned BSEE and MSEE degrees from the University of North Dakota and then a Ph.D. in physics from Yale. He went on to do amazing things at AT&T and Bell Labs, and his name lives on today most notably  in the Nyquist sampling theorem, also known as the Nyquist–Shannon sampling theorem. So when I saw the horse Nyquist, my mind immediately started to think up engineering jokes: I wonder if Nyquist will run twice as fast as the next fastest horse…How do we know that the horse we see is real and not an aliased image…Will his lane at the starting gate be the Nyquist Zone?  Then I discovered Nyquist the horse is actually named for a hockey player. My short-lived excitement was crushed.


In this issue, our feature article authors Colm Slattery and Ke Li take a look at a particularly timely issue, as more and more environmental regulations to control and monitor liquid waste are appearing worldwide.  Electromagnetic flow technology is the choice for this particular application, and the trend is toward an oversampled approach, which challenges the ADC requirements. In addition to oversampling design considerations, their detailed article discusses the significant power
challenges involved in the application.


In the 141st running of the Preakness, Nyquist came in third, so no Triple Crown bid for this horse. But those of you involved with  mixed-signal system design, know that Nyquist, the sampling theorem, can never really be beaten. It is also known to occasionally bite you.


The data sheet for the amplifier that I've chosen for my application specifies a small signal bandwidth along with a large signal bandwidth, and they are quite different specs. How do I determine if my signal qualifies as small or large?


Whether you consider your signal small or large, proper selection of an amplifier requires (among other things) knowledge of how fast it needs to be. And the larger the signal at the output, the harder it gets a good, high-fidelity reproduction of the input signal.


Learn more in "Citius, Altius, Fortius" as published in the April 2016 issue of Analog Dialogue.




We welcome your feedback in the comments section. Has this article been useful or helped you in your design? Would you like more information on this topic?


Best regards,

Gustavo Castro

When designing a motion control system that uses a gyroscope as it feedback sensing element, understanding the system's dependence on the gyroscope's noise behaviors has a number of rewards.  Most notably, these insights can help system architects establish application-relevant performance requirements, which will help them focus early evaluation efforts and resources on the most appropriate class of device.  While we all want our designs to be as small and as inexpensive as possible, they also need to perform appropriately, through all relevant environmental conditions.


A wise man once offered that "a problem that has been well defined, is already half-solved." With that thought process in mind, we present, Designing for Low-Noise Feedback Control with MEMS Gyroscopes, which describes some of the key attributes of noise in MEMS Gyroscopes, while describing how they impact a simplified application that tightens a camera's pointing angle error from red to green (see below).  I hope that this helps you and I look forward to your comments!  Thank you for visiting our blog and for checking out the latest issue of Analog Dialogue. To gain access to this new article, just click here.

Our company bills itself as “bridging the physical and  digital.” For those of you working on circuits and systems that cross that bridge, both of our feature articles this month include hard hitting design information for you.


Mark Looney offers practical insights for using MEMS gyroscopes as feedback sensing elements in motion control systems. MEMS gyros are a booming tech area, with both gyros and accelerometers essentially taken
for granted in smartphones, tablets and game systems nowadays. Of course, they have many applications beyond consumer electronics, such as automobiles, space, and industry. Mark probes a bit further beyond simply designing for “low noise” and looks at quantifying how gyroscope angular rate noise will directly affect system level behaviors. I don’t know about you, but the word “gyroscope” brings to mind one of my favorite mystifying toys as a kid—a flywheel combined with a gimbal that, when given a good spin, would balance and stay somewhat upright for a little while on just about anything, including a piece of taut string or the tip of a pencil. The gyroscope technology that Mark Looney writes about doesn’t contain microscopic spinning flywheels, which is comforting to me, because if a gyroscope is going to be driving my car of the future, I’d just as soon not have a toy at the wheel.


Daniel Burton’s article discusses classic vs. integrated solutions for protecting high-precision op amps from those nasty, real-world overvoltage conditions that can ruin the op amp’s performance, or even take it out altogether. The trick is to protect the signal path adequately without creating an unacceptable loss in performance. Daniel does a good job of
describing the always present real-world phenomena that can cause a harmful overvoltage event on the op amp’s input. There’s danger crossing that bridge between the digital and real worlds, and you have to protect your circuit
“babies” so they can keep doing the jobs they’re designed for. It’s a lot like sending your kids out into the world—a notion which currently is at the top of my mind, because my triplets are all graduating from college this month. This
life-event has my mind going down some weird paths, both celebratory and anxiety-ridden. With kids, as with your circuit and system designs, you do all you can to anticipate the real-world dangers ahead and prepare them adequately—as Daniel clearly points out. Do this well, and for years to come your designs will help protect, inform and entertain our world.


Lastly, I want to thank those of you who participated in our March survey. We appreciate you taking time from your no doubt very busy schedules to offer some feedback. Your comments were overwhelmingly positive and it’s gratifying to hear that Analog Dialogue content helps you do your job. It’s clear that many of you use our publication to help solve some of your technical design problems and to find new design ideas, and we aim to keep it that way. Your feedback is helping us make the magazine’s content more accessible in today’s varied media while still remaining true to its engineer-to-engineer roots, a commitment we'll carry forth into our next 50 years.

Users are facing more and more difficulties in limiting the signal chain noise, like in implementing filters, to take advantage of high performance ADCs without limiting the ADCs’ capabilities.


One article discusses the design challenges and considerations associated with implementing analog and digital filters into the ADC signal chain to achieve optimum performance. The data acquisition signal chain can utilize analog or digital filtering techniques, or even a combination of both. Since precision SAR and Σ-Δ ADCs are commonly sampling within the first Nyquist zone, this article will focus on low-pass filters. It is not the intent to address specific low-pass filter design techniques in this article but rather their application in ADC circuits.


The full article is available in ADI Analog Dialogue.

Analog design engineers often throw around names like Chebyshev and Butterworth and Bessel when referring
to types of filter designs and most of us probably get an immediate mental image of the amplitude vs. frequency response when we hear these common filter designations. But, one might wonder, who are these folks who’ve lent their names to famous filters?
Read about them and their relative levels of fame in the Note from the Editor in the April 2016 issue of Analog Dialogue.

Sometimes you just can’t beat having a reference book on your office shelf for quick access to those formulas, graphs, concepts, circuits, and general technical knowledge that you occasionally need, but maybe don’t want to commit the brain cells to memorize. Technical books are alive and well and ADI has been a fairly prolific publisher over the last 50 years. Read about what’s available and where in the Note from the Editor in the March 2016 issue of Analog Dialogue here.

Moore’s Law, just like Analog Devices, has recently celebrated its 50th anniversary. Moore’s Law, as you probably know, is essentially the principle that the number of transistors on a chip doubles every 18 to 24 months. This breathtaking scaling has worked well for the digital world—i.e., memories and microprocessors—but not totally for interfacing with the actual physical world, which is of course analog. Read more in A Note from the Editor in the February 2016 issue of Analog Dialogue.

An increased number of applications in industrial, instrumentation, optical communication, and healthcare industries use multichannel data acquisition systems that result in increased printed circuit board (PCB) density and thermal power dissipation challenges. These applications also demand precision measurements, reliability, affordability, and portability. System designers make trade-offs among performance, thermal stability, and PCB density to maintain optimum balance and they are continually pressed to find innovative ways to tackle these challenges while minimizing overall bill of material (BOM) cost.


You can read the entire article titled: "Integrated Multiplexed Input ADC Solution Alleviates Power Dissipation and Increased Channel Density Challenges" as published in the January 2016 issue of Analog Dialogue. This article highlights the design considerations for multiplexed data acquisition systems and focuses on an integrated multiplexed input 4-channel/8-channel, 16-bit, 250 kSPS PulSAR® ADCs AD7682/AD7689 solution to address these technical challenges for space constrained applications such as optical transceivers and wearable medical devices.


We welcome your feedback in the comments section. Has this article been useful or helped you in your design? Would you like more information on this topic?


Best regards,



Are you kidding me — it’s 2016 already? It seems like just yesterday the world was stressing over the Y2K bug. Anyone remember that? Folks were worried that our dependence on computer technology would be civilization’s demise, when the calendar rolled over to the new millennium and every computer would crash. After we got safely past that apocalyptic hurdle, it seemed like the dam broke in terms of technology dependence — i.e., embedded computing — PDAs and smartphones took off around that time, and the rest is technological history. Virtually every major industry has been revolutionized by technological breakthroughs in the past 16 years, or maybe you hadn’t noticed.


Way before the Y2K non-disaster, I remember the predictions of the death of analog, as digital gradually took over. But analog and mixed-signal design is alive and well today especially since these complex embedded high-speed and low-power computing systems that we’re so dependent on need to connect to, communicate with, or sense the real world. And the real world speaks analog. 


In this issue the integrated multi-channel data acquisition solution article by Maithil Pachchigar, discusses how to solve the analog interface challenge in power- and space-constrained applications, such as wearable devices and the internet of things. These are fast-moving markets on the edge of what’s possible, and analog interface design solutions like those discussed in the article are essential to making the products work. Whether we’re talking home automation, medical technology, automotive mechanics, fitness, smart clothing—or some new market concept that we currently aren’t aware of—they all require a degree of real-world interfacing along with masterful analog design techniques, design tools (also addressed in this issue), parts, and systems.


I don’t think our society is at all nostalgic for those days 16 years ago when we feared that our growing dependence on computers would ruin everything. I never thought my wife would become addicted to smart technology, but it’s happened. It’s just way too much fun. Not to mention that it saves lives, keeps us safer and healthier, and, well, it makes life interesting since we never know what innovation is coming next. So, thanks to our readers for the amazing designs, products, and systems that you all have created over the years. Here’s to a new year of continued creativity and design excellence. We hope you continue to stay in touch with Analog Dialogue through the New Year, (note that this issue begins our 50th volume set – more to come on that milestone) and let us know how we can better help you to keep innovating. 


Read the January 2016 issue of Analog Dialogue here.


Jim Surber, editor

Why is the effect of common-mode signals at the output of an instrumentation amplifier larger than the CMRR specification? Even if you haven't run into this type of issue but normally work with differential signals, it is worth taking a couple minutes to make sure you're not misplacing your expectations regarding common-mode errors.

Learn more in "O CMRR, CMRR! Wherefore Art Thou CMRR?" as published in the January 2016 issue of Analog Dialogue.



We welcome your feedback in the comments section. Has this article been useful or helped you in your design? Would you like more information on this topic?

Best regards,

Gustavo Castro

In the December issue of Analog Dialogue we conclude the four-part series on software-defined radio design. The authors bring the algorithm and hardware together, and to life, and take their radio for a real-world test drive. This is the culmination of a journey that has taken us from simulation to prototyping to production-worthy design, and now we see it in action. This issue also features an article on the critical role that precision bipolar DACs serve in calibration and control functions in a multitude of applications from motor control to industrial automation. Several system block diagrams are explored with a focus on the design considerations for the DAC functions involved therein.

Part 4 of the “Four Quick Steps to Production: Using Model Based Design for Software Defined Radio” concludes the article series by introducing the final step in a SDR design - generating C and HDL code out of the Simulink model and integrating the code into the SDR platform’s software and HDL infrastructure.  This article highlights three very useful and powerful tools: MATLAB Coder and HDL Coder provided by MathWorks and Board Support Package provided by Analog Devices. By reading this article, you will learn how these tools work together to complete the final production.


When it comes to content, Part 4 walks through the steps of partitioning the ADS-B model presented in the previous parts of the series into hardware and software components, optimizing the model for code generation, generating C and HDL code using MATLAB Coder and HDL Coder and finally deploying the code onto the SDR system.  Then it introduces the Analog Devices HDL Workflow Advisor Board Support Package (BSP) that provides seamless integration of Simulink generated IPs into the Analog Devices AD9361 SDR platforms HDL reference designs. The end result is a fully functional SDR platform capable of decoding live commercial aircraft ADS-B traffic. This example system shows that the Model-Based Design workflow in combination with the Analog Devices AD9361/AD9364 integrated RF Agile Transceiver programmable radio hardware can help design teams develop working radio prototypes more quickly and less expensively than using the traditional design methodologies.


Please check out Part 4 of the article series to learn more about the final steps of taking a SDR system from simulation to production.

In Part 3 of the Four Quick Steps to Production: Using Model-Based Design for Software-Defined Radio article series we employ two very useful tools provided by ADI. These two tools not only help us verify the ADSB algorithm with live data, but also have been well received and widely adopted by ADI partners and customers.


First is the MATLAB and Simulink IIO System Object. The IIO System Object is based on the libiio library and is designed to exchange data over Ethernet with an ADI hardware system connected to a FPGA/SoC platform running the ADI Linux distribution. With this tool, the users can easily stream the data from ADI hardware to MATLAB, and then the post processing will happen inside MATLAB. Before this tool was developed, in order to verify the data from real hardware, the users had to first save some data on the Linux side, and then import the data into MATLAB. Based upon this interface, we have created several MATLAB and Simulink models for users to try out the hardware in the loop simulation, which is an important step in model-based design.


The AD9361 filter design wizard helps you design the 128-tap FIR filter on the Tx and Rx paths of AD9361. This digital FIR filter is very much required to maximize the system performance, but designing it is quite complicated, since there are various combinations on the signal paths, consisting of analog filters, as well as several digital half band filters. Before this tool was developed in late 2013, customers who needed to design an FIR filter had to ask for help from the ADI designers, which was not a very efficient way to solve each individual problem. Nowadays, with this tool, the users only needs to input the basic filter specifications, and the tool will make the design for them, so that everyone can easily implement their own design and change it as often as they want. So far, this tool has been a required dependency by two MathWorks hardware support packages: Zynq SDR Support from Communications System Toolbox and Analog Devices RF Transceivers Support from MATLAB and Simulink.


Please check out the Part 3 article to see how to use the MATLAB/Simulink IIO System Object to perform hardware in the loop simulation, and how the AD9361 filter design wizard is used to improve the SNR on receiver path.

In this new article, the single-ended to differential circuit is even made more versatile with adjustable common-mode and greater output dynamic range. The article discusses how it is accomplished, the considerations in choosing the right amplifiers, and the stability and bandwidth of the configuration. With the versatility of a simple single-ended to differential circuit, high performance and precision, this circuit can have wide applications in the analog front end of data acquisition circuits.


Versatile, Precision Single-Ended-to-Differential Signal Conversion Circuit with Adjustable Output Common Mode Boosts Sy…