Skip navigation

Captain’s blog. Stardate 72216.8

I’ve received an urgent request to mark the premiere of Star Trek Beyond with some thoughts about how ADI technology could have played a role aboard the original USS Enterprise. The idea is, as a certain crew member of the NCC-1701 might have said, “Fascinating.”


Live long and prosper. At least just a bit more.

Watch enough Star Trek episodes, and even a casual viewer will notice that crew members in red shirts seem to die a lot. Had those red shirts come from Hexoskin, ADI technology could have helped gather location and environmental data, beamed them to the cloud for analysis, and relayed a text to the wearer’s communicator to get him the heck out of danger.


I'm givin' her all she's got, Captain!

For all the matter/anti-matter horsepower of the Enterprise’s warp drive engines, poor Montgomery Scott, the ship’s chief engineer, was always being berated by Captain Kirk because there was never enough power to get out of whatever situation had been written into the script. Engineering can be a thankless job.


If only Scotty had access to our very own EngineerZone. Instead of screaming through the intercom, he could have calmly posted a question about dilithium crystals and had a wealth of solutions pop up right there on his Android or iOS communicator.


Are you out of your Vulcan mind?

When it came to bedside manner, Leonard McCoy, the Enterprise’s cantankerous chief medical officer was no Hawkeye Pierce or Derek Shepherd. You got the feeling he didn’t really want to see patients and they didn’t want to see him. If only they had the remote monitoring capabilities offered by ADI and LifeQ. Non-invasive body monitoring devices would capture essential physiological data and advanced bio mathematical algorithms would combine to help manage and maintain the crew’s health and wellbeing. Plus, it would be great to hear McCoy snarl, “I’m a doctor, not a micro-electro-mechanical-bio-mathematical engineer.”


Captain’s blog. Supplemental

It’s hard to improve on something as unique and special as Star Trek. Yet, for all its visionary technology and innovation, I still think a little help from ADI might have made it even better. And the Enterprise and her crew might have not only boldly gone where no man had gone before, they might have also gone Ahead of What’s Possible.


Trekkie or Trekker? Whichever you call yourself, you probably have your own ideas about how to use ADI technology in the 23rd century? And you don't have to wait 200 years share them. Comment here today.


Click to Tweet: Doomed Redshirts, balky warp drives & cranky Dr. McCoy. See how ADI technology could help on the USS Enterprise.


Image Source:


I’m a doctor not a blogger! Oh wait - I am a blogger.

In 1966 Gene Roddenberry let his creativity run wild and brought his vision of the 23rd century to life. He assembled a mixed race crew aboard the USS Enterprise for a five-year space mission. Their mission is well known to us, “to explore new worlds, to seek out new life and new civilizations; to boldly go where no man has gone before.” And they went. The crew of the Star Trek took us on many adventures and expanded the limits of our imaginations. Fast forward to 2016 and Star Trek is a phenomenon. Multiple television series and movies, conventions and let’s not forget the countless catch-phrases.


One year earlier, in 1965, Ray Stata followed his vision and co-founded Analog Devices. What started as an op amp-based company in Cambridge, evolved into an integrated circuit industry first thought leader and is now a highly successful, global organization. ADI products merged into the consumer market in the 80’s with the popular devices of the time: Sony disc players. On the heels of our 50th year celebration, our products take flight regularly and soar through space to capture photos of “new worlds”.


This Friday, Star Trek will celebrate fifty years of success with the premiere of the new movie,

Star Trek Beyond. Seeing the similarities between our own ADI story and Star Trek, I asked employees to share their thoughts on two topics: If Star Trek were conceptualized today, what ADI technology would you use or re-imagine to build the USS Enterprise? OR Are you are Trekkie? Did Star Trek inspire you in any way with your decision to become an engineer?


Over the next three days we will share our thoughts with you, starting with this blog. Without further ado I bring you the minds of the ADI crew, as they continue their mission of staying “Ahead of What’s Possible”.


Mark Cantrell Product Applications Engineer, offers a few improvements for the Enterprise.

"The world of Star Trek is a dangerous place.  And it is not just the Romulans and Klingons shooting at you that can cause grievous bodily harm.  The Enterprise itself handles enough energy to flip the massive star ship past light speed with an ominous whirr.  Every panel in engineering controls lethal power and every Jeffries tube has access to raw antimatter.  All this raw energy means that the need for personnel safety in controlling the ship has to advance as fast warp technology itself.


ADI’s isolated control and communication technology is in its infancy right now at ADI with the iCoupler digital isolators.  These devices will be essential to safely control the ships systems, and isolate sections of the ship in case of a severe hull breach.  At the rate that ADI is innovating its safety devices, they will be more than a match for the awesome power of a star ship, so Captain Kirk does not run out of engineers before he gets to the final battle.

Next time you want to stick a pointy tool into the antimatter stream and flip the polarity, the great-great grandson of iCoupler will be in that pointy tool to keep you from getting vaporized."



Julie Barbeau,Sr. Product Development Engineer, answers the call, are you a Trekkie?

"Are  you kidding me? I love Star Trek! Note the present tense.  I got the original series on DVD for Mother’s day this year.  I’ve introduced my children to Star Trek and they love it too.


Gene Roddenberry was a visionary.  In the 1960’s he foresaw hand held communication devices and today we have cell phones.  Handheld scanning equipment for medical diagnosis is now a reality.  Star Trek has allowed engineers to dream of what might be and then go and invent it.  The tricorder hasn’t been invented yet and we can’t transport people or things by reducing them to molecule and reassembling them in a new location.  But thanks to Star Trek, we think about these and other seemingly impossible things.


I wish I had more time to write about this.  For now, it will have to suffice to say that it was a major influence on me and my little brother who also became an engineer."


David Kress, Director of Applications talks about future opportunities for ADI products on the Enterprise. 

"By Stardate 2450, Analog was producing a variety of products that were useful on the voyages of the Enterprise -- some of these were visible, some not.



Communications -- One might wonder how the Enterprise could communicate so rapidly with Starfleet over light-years of distance. The physics of quantum entanglement was well-established that enabled particle pairs to communicate their state instantaneously over great distances. The ADI Isolator group had extended isolation technology to include the quantum entangle-o-meter with up to 8 bi-directional channels able to encode and decode spin direction changes at over 1 GHz -- with isolation distances reaching beyond 100 light-years. Relay stations could extend that range indefinitely.


Food -- Although many have postulated that human bodies might be modified so as to obtain sufficient nourishment with just a couple daily pills, the need for energy and calories is difficult to deny. The Analog technologies that were used to grow the perfect tomato in 2016 were greatly extended to grow organic items with minimal resources and no net waste. Since the Enterprise, including its occupants, was a closed system, all waste, from exhaled CO2 to dead skin, was captured and recycled. The advanced ADI sensors could detect trace elements and re-use them where needed to produce replenished foodstuffs. Probably not a gourmet meal, but it kept everyone more than healthy.


The Enterprise -- Analog Devices was one of the key contributors to the development of the warp drive system used in the Enterprise. This grew out of Analog's participation in the LIGO gravitational wave observatory. LIGO continued to develop more and more sensitive equipment, with our help. One of the key premises of sensors is that any sensor can also be an actuator (think photodiode and LED). So, as the LIGO and its successors advanced, the technology, which could sense gravity waves, was eventually able to create them. This was combined with the concept of the Alcubierre Warp Drive (look it up), which relies on the manipulation of gravity. These theoretical concepts were developed in the 21st century but had to wait for ADI to begin making the sensor/actuator chips from dark matter to realize the necessary energy levels. The primary work was done at the Analog Space Garage, the latest extension of our ventures that take you Ahead of What's Possible."


Are you a Trekkie? I already have my ticket for Friday. How about you? Share your ideas about how ADI technology can improve the Star Trek universe in the comments and be sure to check back tomorrow for Ed Grasso's EngineerZone Spotlight blog: Where No Semiconductor Has Gone Before.


Click to Tweet: “Forget seat belts, let’s put in a warp drive.” ADI engineers go under the hood of the USS Enterprise.


Image Sources:



As the leader of the team that developed the AD9361 product, I'm often asked where the idea came from.  Was it a "light bulb moment"? Divine inspiration? Uh, no. It was the result of a series of failures.


The AD9361 software-defined radio (SDR) is now one of ADI's most successful products. It is in hundreds of applications. Chances are you've used it and don't even know it. But how did we get here?


I was leading a fledgling RF transceiver development group at ADI. We had a family of WiMAX chips. (Remember WiMAX?  Me neither.) When the WiMAX market didn't happen, we had to figure out what to do next. One thing I knew I'd never do again was to chase a market that may not develop.


This was when the industry was starting to migrate to the 65nm technology node. Along with the remarkable capabilities of this new technology, came what seemed at the time, an incredibly high development cost. (This is amusing in retrospect given where costs are now!) With these soaring costs, my management wasn't going to be very patient with more failure.

RV Blog Price of Bread.jpgAnyway, the aha! moment came when I realized the implications of the really good switches that we could make in 65nm CMOS! (I'm picturing readers switching to some very important cat videos now - stay with me!) 


By good switches, I mean switches that have low series resistance when closed and small parasitic capacitance when opened. It turns out that those good switches play a huge role in making SDR possible. Here's a short list of things these switches enabled:


  • A "universal" local oscillator generator (LO) up to 6GHz. The core of the LO generator is a single-core LC VCO that tunes from 6GHz to 12GHz and is enabled by digitally-switched capacitors with a high Con/Coff ratio and pretty good quality factor.
  • Inductor-less RF signal paths. Traditionally, RF signal paths have needed resonant circuits to tune out capacitance in order to keep power dissipation low. With the very low parasitic capacitance of deep-sub-micro CMOS, we were able to eliminate the inductors.  This is important because nobody ever told inductors about Moore's Law so they're not only expensive but they get more expensive with Moore's Law because inductors don't scale but the cost per area does. And, this octave VCO is followed by cascaded divide-by-twos (made with switches, of course) from which we can generate any frequency up to 6GHz.
  • Widely reconfigurable continuous-time sigma-delta (CTSD) analog-to-digital converters (ADCs). Aside from the bandwidth and the power reconfigurability, these are remarkable ADCs because they have inherent anti-aliasing which is good because ...
  • ... after inductors, active analog filters probably scale most poorly with Moore's Law. With ADCs that don't need anti-aliasing filters we can use relatively low order filters and, because of those switches, they have 200:1 bandwidth tuning ratio.


Of course, it goes beyond just the switches. Lots of other things are enabling, including the direct conversion radio architecture with heavy use of "digitally-assisted-analog". Direct conversion is enabling because, compared to other radio architectures, it consumes the least power, eliminates impossible-to-integrate components such as IF filters and relaxes RF filters since there are no out-of-band images.  Direct conversion is used in virtually every consumer wireless application and ADI is bringing it to the rest of the world. 


The architecture also scales. For example, the same architecture as the AD9361 is used in the AD9371 but at twice the bandwidth and between 10 and 30dB better performance in most metrics that matter.


That is just a taste of the benefits and challenges of direct conversion. Look for more on this topic and other
under-the-hood topics in future blog posts. Now, you may watch cat videos.




Click to Tweet: Was ADI’s successful AD9361 software-defined radio born of divine inspiration? ADI fellow Tony Montalvo: “Uh, no. ” @Adi_News #EngineerZone

Quite often I get questions from customers on how high they have to drive their signal generators in order to get close to full scale. This RAQ answers some of them.

Tony Montalvo.jpgA new Radio Architecture and Design blog series, will makes it's debut on the EngineerZone Spotlight next month. This new blog will be authored by a team of Wide Band RF Transceiver experts.


ADI Fellow, Tony Montalvo will kick off the new series by sharing his technical and applications knowledge about Radio Architectures and Design and giving customers a look "under the hood". I had the pleasure of interviewing Tony to find out more about this new blog series, RF Transceivers and how he transitioned from an Ultimate Frisbee enthusiast to a highly recognized innovator at Analog Devices.


LA: I understand you are an ADI Fellow. That’s quite impressive. What does being a fellow entail?

TM: Out of 9000 employees there are about 30 Fellows so it’s a huge honor to be named a Fellow.  It’s humbling to see my name on the same list as some of the other Fellows who have created this industry and, in some cases, literally written the book.  The core criteria are innovation and impact.  That is, innovation has to be impactful and the easiest way to measure impact is revenue.


LA: Tell us a bit about this new blog series?

TM: Our products bring some very complex technology to applications that have traditionally been served by discrete components.  I thought it may be interesting to our customers to learn a little about our philosophy of radio, how we got here and what’s under the hood.


LA: What inspired you to become an engineer? 

TM: This isn’t one of those I-took-the-TV-apart-when-I-was-six stories.  I wandered until I found something that grabbed me.  As a kid I was consumed by skateboarding and punk rock.  I studied physics in college (Loyola University, New Orleans) so I must have had some technical tendencies.  Right after college I went to Columbia University in NY to get a masters because I couldn’t figure out what else to do.  I was a somewhat uninspired student of solid state physics and quite an inspired player of Ultimate Frisbee. Finally, I took a circuits class that changed everything.  The professor, Yannis Tsividis, was describing a circuit and wiggling his fingers to show how voltages varied in time.  For example, one finger would wiggle at the input of an inverting amplifier and another finger would wiggle more and out-of-phase at the output.  Suddenly, everything fell into place and it was finger wiggling that did it.


LA: Tell us about your years after graduate school and what brought you to ADI.

TM: After I finished my MSEE, I went to the Bay Area and became a Flash memory designer at AMD.  This was the first generation of Flash memories so it was fertile ground for a budding analog circuit designer.  We had to figure out things like how to generate the programming and erasing voltages in a way that would maximize the lifetime of the memory cell.  It was great work but I realized that it would become repetitive once those fundamental problems were solved.  After four years, I left to get a PhD at North Carolina State University.  My dissertation was about an artificial neural network chip that could learn based on examples and the “synapse” weights were stored on a floating gate which is the same idea as a Flash memory cell.  It was cool. It worked but turned out to be a solution looking for a problem.  Now, some 20 years later, neural networks are back in vogue and are at the core of many of the speech recognition systems like Siri.


After I got my PhD, I joined Ericsson in RTP, NC as an RF IC designer.  I knew literally nothing about RF at the time.  They were growing fast and luckily had a very low bar! Ericsson was an amazing learning experience for me.  I was surrounded by radio experts and soaked up a ton of knowledge. My claim to fame was architecting a receiver IF chip that eliminated a bulky 2nd-IF filter.  Some of the most experienced RF people in the world told me that it was impossible, that it “would never work”.  It did work and millions shipped into Ericsson handsets.  I’ve learned that unless someone claims that a product “will never work”, I’m probably not pushing hard enough.


Although I learned a lot at Ericsson, it became clear that as an IC designer I’d be in a better position at a semiconductor company.  I left Ericsson after 5 years and joined ADI in 2000.


LA: I recently learned that your office is on the North Carolina State University’s campus. How did that come to be?

TM: I started ADI’s site in Raleigh. Since I’m an NCSU grad and since I was an adjunct Professor at NCSU, I chose to locate the site on campus. Remember that this was in 2000 when things were booming.  Very soon after we signed the lease the boom ended and it was just me.  That wasn’t so comfortable but luckily the company was patient and we VERY gradually grew.


Being an adjunct professor, I taught an RF IC design course every other semester for many years.  This was a great experience for me for a few reasons.  First, you have to go much deeper in understanding something in order to teach it.  Second, my communication skills improved tremendously which comes in handy when interacting with customers.  Finally, I hired some of the best students!


LA: What interests you about the RF sector, specifically transceivers?

TM: Impact, first.  Everyone can relate to wireless communications these days so if we can do something that makes people’s experiences better it’s easy to see.  Second, the industry is changing quickly.  For example, not too long ago voice was the killer app for cell phones and coverage was spotty at best.  Now, just a few years later you have nearly instantaneous access to all of the knowledge from all time in your pocket.  And we’re far from done.  If anything, the change is accelerating and that what makes this so much fun.


LA: As an engineer you have experienced many successes and failed attempts. Please share one example with our audience.

TM: I’ve generally been able to make products work but that doesn’t guarantee success.  There have been several products that were technical successes and commercial failures.  I’ve learned that the most important thing is identifying something valuable to make.  Once you identify a valuable problem good engineers can find clever solutions.  I’ll talk much more about that in the blog.


LA: Do you have any advice for tackling an intimidating project?

TM: You can’t finish until you start.  And, if it’s not intimidating at the start you’re not trying hard enough.


LA: What are you reading right now?

TM: I just finished Oliver Sacks’ autobiography “On the Move”.  He’s an incredibly accomplished neurologist who had a passion for bringing these complex topics to the masses.  For example, the Robin Williams movie “Awakenings” was based on his work. What really resonated with me was his early years.  He was a wanderer.  He didn’t have a plan.  He just did what interested him and things worked out well for him. That describes my path pretty well too.


Thank you, Tony for taking some time to share your experience with our audience. We will be sure to stay tuned to the EngineerZone Spotlight blog for the first installment of this mystery RF Transceiver based blog.




CLICK to TWEET: A Fellow's Path: An interview with Tony Montalvo #EngineerZone @ADI_News

There are a lot of tables. There are a lot of specifications.... do I understand them all, do I understand their impact, which ones interplay? I want to be sure about picking this converter within my solution. It can be daunting

...and you are only on page 4 of the datasheet.


Even though, I personally enjoy reading, I prefer pictures. I prefer to see it. To play with it, to understand through seeing....seeing is believing.


Fast, with no hassle.

This is the essence of Virtual Eval. Seeing, playing and understanding. Building confidence that this converter is the right one, or other times, perhaps not fitting what I need. Up to now seeing meant, ordering the evaluation board and hooking it up on the bench - Virtual Eval is much faster and more efficient in selecting the right component for your task.


Virtual Eval is an interactive model of your converter that's one simple click away.

Click and see for yourself.


Precision Sigma Delta ADCs

Our precision sigma delta ADCs build in lots of flexibility around our first class ADC core performance. Our customers like the ability to tweak and optimize via these programming options. With functionality comes a fair level of complexity and many, many use cases. Virtual Eval cuts through the complexity and offers you the ability to see if the ADC can operate to fit your needs. It makes understanding easier and faster.


Here's a bit of a show reel of how it works:


Launching with coverage of the AD7190 family of 24-bit Precision Sigma Delta ADCs; AD7190, AD7192, AD7193, AD7194, AD7195 with fully integrated front end multiplexer, PGA and a host of digital filtering functions, let's peak at AD7193 through the browser enabled playground of Virtual Eval.



iStock_000015939174XXXLarge_jpg_with_text.jpgAD7193 has a multiplexer. Multiplexers and Sigma Deltas can muddy the water. Are you sure you've understood how fast I can multiplex, how I can reject a given frequency? AD7193 excels in gaining up signals and making a DC measurement very accurately with low noise and high resolution. Part of this is accuracy is in tuning the filter to reject known interferers.

A classical case is rejection of 50Hz, 60Hz, or if your design is global: rejecting 50&60Hz.

Scenario: 100mV input range on 4-channels, Gain of 8 PGA, need to reject 50 and 60Hz in tandem. How does AD7193 do this, which choices do I have? Virtual eval will show you for AD7193


Select blocks for configuration by clicking on the block diagram and then run the simulation. The tool will allow you to browse the change in noise between data rates or gains. It will tell you if your analog input voltage span is too large for the gain selected and even if your serial clock is too slow. All the little things.


Timing diagrams showing the instructions to be written to the ADC and the expected times to receive the output from the ADC depending on the filtering rates and the number of channels selected. It's like putting an oscilloscope on the digital interface and looking at the format and timing of the ADC output, where the RDY signal falls to indicate a new conversion is ready, the period between conversion ready signals.....all graphically shown


Frequency Domain plots of the rejection offered by the digital filter at specific rates like 50/60Hz are also available. You can see what the specification would be for conditions outside of those shown in datasheet tables. For instance, you know your temperature range is going to mean the clock is going to have a tolerance of +/-2Hz instead of +/-1Hz as the datasheets have used as a standard. AD7193 Virtual Eval will calculate that for you and show the expected rejection.


Output Session settings to file when you're finished. This way you can keep a record of the configuration you need for your project and help you fill this in for your design requirements document.

Here's an example of the file that I export at the end of my settings- I've also attached it to this post.

It's a text file and you can open it in Word, Excel or Notepad, whichever you'd like.



Fast forward to your internal design reviews, a question comes up about a certain voltage rail, input voltage span - is that ok for the converter, can the converter handle it do we need to change settings?.....pop open Virtual Eval to check there and then, even share on the projector in your meeting -a quick check....peace of mind.


That's it for now. Stay tuned for future products releasing on the Virtual Eval platform. AD7124, AD7175 are in the pipeline. Virtual Eval: It's a good one, we like using it at ADI too.


twitter.pngClick to Tweet! "Seeing is believing with Virtual Eval #EngineerZone @ADI_News"

Let’s talk a little bit about power. Well, maybe more than just a little bit as this is a fairly long post.


Every circuit requires power to function, and from my experience as an applications engineer, power is usually one of the last things to be addressed in the design phase of product development. While commonly viewed as a commodity these days, the truth is that there are many subtle nuances in most power designs that must be accounted for to allow an application circuit to reach its full potential. And more and more, design engineers are not experienced in power or just don’t have the time to devote to it. Many books have been written on power design, but very few people have the time or scope in their jobs to fully understand and work with it. At the same time, data rates, bit counts, and noise floors have made power supply requirements more stringent than ever. But how do we design a power supply to meet these ever-increasing specifications with some confidence?


It takes about 3-5 hours for an experienced application engineer to perform a switching power rail design manually. Most of the external components need significant adjustment of their effective values to account for effects of DC bias, saturation, thermals, and many other factors. Unless these items are taken into account, the power rail will not meet the desired specifications and might even be unstable. Furthermore, most companies want not only a design, but also a simulation model that verifies that the design works as desired to reduce risk of board spins. This simulation model must also have the effective values of the components to provide accurate results.


ADI’s power applications team handles many requests for these designs. We developed a set of design tools that we use internally to assist in making sure all component and system effects are addressed, design time is minimized, and that a design will indeed meet the specifications required when built. It wasn’t too long before we realized that such a design tool would be a huge benefit for customers to use as well.


ADIsimPower and ADIsimPE is the result. Although not widely known, this tool set is groundbreaking in many ways. Using a very simple 3-step process, anyone, regardless of power experience, can get a schematic, Bill of Materials (BOM), starting layout and in many cases an exact simulation model in just a few minutes instead of hours or even days. Designs can be optimized for efficiency, cost, area, or parts count as needed. Advanced users can do far more to tweak a design (read the datasheet first!) such as change the major components and fine-tune the specifications. When built to the BOM, an eval board will meet the specifications completely, and the tool predictions and simulation results will match the measurements in the lab when built. Seven distinct topologies are fully supported and even combinations, including SEPIC, Cuk, and the inverting buck-boost. Many of our multi-rail devices are also supported.


Let me illustrate with an example just how fast and easy it is to complete a complicated power design from end to end. A very common application is providing bipolar power for a signal chain. Let’s assume that we have a 12V input rail with a tolerance of +/-5% (i.e. 10.8-13.2V), and that we need to deliver +5V and -3.3V at 200mA each. The maximum ambient temperature will be 55C. Let’s further assume that we need ripple of less than 5mV in this application and that lowest cost will be the design goal for optimization. Here goes:


Step 1: Part Selection

The first step is to use the rail specifications and choose the best part from ADI’s portfolio. To do this, we need to run the selector guide with the five basic parameters that every power rail needs specified. The tool is located at or can also be found by following the links from the Design Tools section of ADI’s website.


Now this selector guide does not operate like all the others out there which simply match the Vin, Vout, and max load currents with datasheet limits. Instead, ADIsimPower computes a partial design for every single power part that ADI manufactures. It determines the entire set of ADI parts that will meet the specified parameters in all of the topologies and displays them with a considerable amount of useful information for making decisions. The selector guide is a little bit conservative and will not suggest a part unless it is sure that the part will work to spec. More on that in a little while.


Let’s enter the positive rail first and see what comes up:

After entering the design parameters, we click the “Find Solutions” button and after a few seconds the following list comes up:


Hmm, lots here. Note each part has an estimate of the cost, area, parts count, and efficiency. While not exact (it is a partial design calculation), these numbers are surprisingly accurate and certainly very helpful to rank our parts by what is important as a design goal. Each of the informational columns can be sorted-on (click the header), and best in class is highlighted in bold. Furthermore, each applicable topology is addressed. There are many additional features on this page, including a very powerful filtering function. Clicking on “Quick Filters” will bring up a rather complicated dialog box that lets you select or exclude important features such as Tracking, Light-load efficiency, Clock synchronization, etc. These allow the user to very rapidly narrow down the list of possible parts and to choose the best candidate with the desired optimization. I’ll save a detailed description of these for another post.


For our purposes, we know that we need a negative rail too, and we can see that the ADP5070 or the ADP5071 in a SEPIC topology can do the positive rail and has a negative converter too. So let’s enter the -3.3V rail and check if those two parts can work for that solution as well:

Upon entering these design parameters, we get the following results:

Looks like we should target the ADP5071. So far, we have quickly narrowed the list of candidates from 39 parts to just one in about one minute. Try that searching through a standard selector guide!


The next step is to download the ADP507x tool by clicking the “Download Tool” button. A zip file will be made available, and after downloading it, you can extract the Excel spreadsheet inside (please extract the tool—Windows won’t allow saving the sheet back to the zip file) and begin the detailed design process.


Step 2: Detailed design and optimization

After downloading the tool, you will be asked to enable macros when you open it. This step is needed because most of the intelligence in the tools is contained in VBA code. The tools are safe, digitally signed and certified, so there is nothing to worry about here.


Now we enter the inputs again, this time both rails simultaneously. We use the basic inputs dialog which displays on startup or when the “Enter Inputs” button is clicked. Here is what we need:


We have one more piece of design information to enter to ask the tool to keep the ripple at or below 5mV. To do this, we need to access the “Advanced Settings” dialog. I don’t want to go too much into all of the possibilities of the advanced settings, but here is what we need to set things to for our design:

Note that I have set the “Vout Ripple” on both rails to 0.1%, with design specifications of 5mV and 3.3mV. I have also checked the boxes to allow a two-stage LC filter to be used. Why did I do this? The tool will output a design that meets the specs even without this, but I happen to know that achieving lowest cost is easier with the extra filtering than the alternative of really bulking up the output caps. While you may not have the benefit of my experience, you will find that this tool set is so fast and efficient that you can easily try several ideas with an investment of only a couple of minutes to optimize things to your liking. The rest of the settings can be altered as needed, but for this example, no further changes are required.


Now we submit the design by clicking “Submit & Goto Basic” and then “Submit/Run”. After a few seconds, the tool completes the design and displays the results. First, we get a schematic:

This schematic is from the eval board, and not every component might be needed for every design, but it is complete. Our tools target an eval board, so taking this approach makes sense and it is easy to tweak the schematic for a final end application. Next, we get the Bill of Materials:

Surprise! Upon doing the very detailed design calculations, the ADP5070 works for both rails. We can save some money by using a lower current part. Remember that I said the selector guide does a partial design and is conservative? Unlike so many of the other design tools out there, ADIsimPower almost never will suggest a part that won’t work, which saves a lot of time from going down dead-end streets.


Note that we have exact part numbers, cost and area here, too. The cost is for 1k or 1-reel pricing, and obviously each company negotiates their own prices for components, so it may not be exact for your organization, but it is a very good starting point.


The next section shows the computed design performance. With multi-rail tools, this section gets pretty long, so I’ll only show channel 1 for brevity:

Note that the tool predicts 2mV of ripple and a crossover frequency of 26.5kHz with 57° of phase margin.


I don’t want to toot my horn too much, but I want to point out that this is a SEPIC design. This topology seems to unnecessarily scare most design engineers (including myself many years ago) because it is a 4th order power stage which makes the control loop analytically intractable. Fortunately, we model the loop in its entirety and are able to numerically solve it so we can recommend a stable design with the most loop bandwidth possible. Even with a second LC filter stage in the loop.


Ok, we are rolling now. It took a lot longer to read this than to do it in practice. Let’s finish the design. Of course, you can print the results now, but perhaps you or your management have a few doubts still? Would a simulation help ease some concerns? We can help with that.


I want to make an important point here. Simulation models show you what will happen if you choose a set of components. Simulation is an analysis tool that doesn’t help choose the component set that meets any specifications. Simulation only helps verify a design. AdisimPower is a synthesis tool. It chooses all components to ensure the design specs are met. AdisimPower uses a very advanced and proprietary database of component specs along with significantly more accurate and detailed models of the power part than can ever be expressed in a datasheet to produce robust designs that meet the predictions when built.


May I make another point here? Our competitors seem to think that a nominal model of the external components is good enough, but I can state that it isn’t. I’ve tried designs in competitor’s tools, and frankly, I am disappointed that I can by default get unstable and totally unworkable designs from them and certainly ones that don’t meet the required specifications. ADIsimPower tells you what is really needed. If you are using a 22uF ceramic output capacitor, it won’t have 22uF if the output voltage is 5V. You may need most of that capacitance to meet ripple specifications and stability. Unlike all the other design and simulation tools out there ADIsimPower uses the effective value at the operating point for its computations as well as in the exported simulation model. Ok, I’m getting off the soapbox now.


Step 3: Design verification

To simulate the design, we just click the “Simulate with ADIsimPE/SIMPLIS” button. This brings up a dialog box asking what exactly should be simulated. You can choose what you need, but let’s focus on the most common runs. For our design, let’s check the loop stability and transient response. Here is the box:


The default is a very good choice for verification. We choose “Export & Run Simplis Model”. Now, you have to have downloaded and installed ADIsimPE for this to fully work. Assuming that you have done so, when you click the button in a few seconds you will load the simulation model loaded with your design specs.

When the simulation loads all that is needed is to run it. Let’s do that. First, here is what we get to simulate:

Isn’t it nice that we already have the effective component values pre-loaded for us? Take a look at 1Cout2. This is a 10uF cap, and you can see that at the operating point, it only has 5uF of effective capacitance. This is how we can match the tool output, the simulation results, and later lab measurements.


We are ready to run the simulation. Now, with most spice models, it’s time to go get a cup of coffee and wait for a few minutes. With ADIsimPE, the simulation completes in about 15 seconds, and includes a true AC analysis that other simulators cannot even perform. I’ll cover the details of why this is so in another post, but the reason for this is that ADI’s switching models are developed in SIMPLIS instead of spice.


Here are the results from the AC analysis:

Notice that the SEPIC gain and phase closely match the tool predictions: 26.4kHz and 64° of phase margin. That is about as good as it gets!


Here is the transient analysis:

Zooming in to look at the ripple on the SEPIC rail, we see 1.2mV of ripple. Pretty close to the 2mV the tool predicted and again a little conservative to account for tolerances:


There are many other things that can be analyzed, but at this point, the design is looking pretty solid. We have a complete paper design. With only seven minutes of time invested, we now have:

  • Selected the optimum IC
  • Generated a full schematic/BOM
  • Obtained all the operational parameters
  • Verified the performance meets spec


If you would like to test the design yourself, you can order an evaluation board. Every tool has an eval board, and you can order one at no cost by clicking the “Order Blank Evaluation Board” button in the spreadsheet and filling out the information.

Many customers would like an example layout as well. Each tool includes the layers:


So there you have it. Even if you have practically no knowledge of power, you can very rapidly get a complete design and layout using ADIsimPower and ADIsimPE. Give it a try!

Wizard Tools? We're not talking about mythical magicians wielding screwdrivers. Analog Devices' suite of Design Tools help you simplify calculations and tackle complex engineering challenges with ease.


Welcome to week five of ADI's 10 Weeks 10 Tools! For those of you just tuning in allow me to recap the previous weeks.


Week One: We introduced you to ADI's Design Tool Wizard. View the video here. With the help of ADI Design Tools – you will be able to find the right products for your design, easily and accurately. These tools simplify your design and product selection process and simulate results to discover real world performance.


Week Two:  Anne Mahaffey shared imagery and user feedback about the Analog Filter Wizard. Read her EngineerZone Spotlight blog, Active Filter Design in a Snap.


Week Three: Scott Hunt wrote, The Diamond Tool: No Longer in the Rough where he shares questions related to the Diamond Plot, the datasheet graph that we use to describe the common-mode range.


Week Four: Anne Mahaffey returned to tell us about the Photodiode Wizard in her blog, Photodiode Circuit Design - tedium is for computers. This wizard tool is used to design a transimpedance, amplifier circuit to interface with a photodiode.


Stay tuned for more about the ADI suite of Design Tools such as, ADIsimCLK, ADIsmPE, Virtual Eval, ADIsimPLL and ADIsimRF in the coming weeks. Until then please visit the Circuit Design Tools & Calculators site or share your experiences with ADI Design Tools in the comments.


Click to Tweet: "The Engineer's toolbox should include ADI Design Tools #10Weeks10Tools"

I'll just get straight to the point - if you're faced with a photodiode measurement circuit design, you should check out our Photodiode Wizard design tool.

tedious.jpgPhotodiode Circuit Design Wizard | Analog Devices

Photodiodes are semiconductor devices that convert light into current - which doesn't sound very complicated, until you are faced with the challenge of figuring out how to measure that current. 

Photodiodes generate relatively small amounts of current, and require amplifier circuit with a large gain.  But, photodiode trans-impedance amplifier (TIA) design can get messy and tedious - the interaction between the high impedance photodiode and the op-amp is prone to oscillation or overshoot, and selecting an op amp (and calculating passive values) that results in a well-tuned circuit can be elusive.

Textbooks, App Notes, and articles can be educational, and extremely helpful when designing a photodiode amplification circuit.  But, there's no way around the math - the equations can be tedious, and finding a suitable op-amp can require multiple iterations of calculations.  (See the High Impedance Sensors section of the Op Amp Applications Handbook for an excellent review of design considerations and calculations)

Tedious iterations of calculations - sounds like a good job for a computer!

Photodiode Wizard Gif Smaller.gif

Have you used Photodiode Wizard for a photodiode TIA design?  Did you find it useful?  Let us know in the comments.

Related reading:

Using SPICE to simulate you circuit?  Simulate ADI devices easily in ADIsimPE.

Why use an Instrumentation Amplifier Tool and What Does it do?


Instrumentation Amplifiers are great for measuring small signals, applying high gains, and rejecting common-mode noise. And Analog Devices in-amps have provided leading-edge performance for years, with products from AD620 to AD8421. However, there is one question that comes up constantly in the in-amp forums and tech-support mailboxes. This question is related to the Diamond Plot, the datasheet graph that we use to describe the common-mode range. It comes up in several ways, most commonly:


            “I am not getting the gain that I programmed.”

            “The output saturates early.”

            “There is too much error.”

                Or for those who already know about the diamond plot

          “How does the common-mode range plot change for my supply voltage/gain/REF pin voltage?”


These are perfectly valid concerns for in-amp users. There shouldn’t be any mystery about your circuit’s measurement range, even if it is caused by an internal limitation. However, until now, it was difficult to illustrate how changes in circuit configuration affected the Diamond Plot. Even when the equations are published, the designer has to go into each individual datasheet of the products under consideration to calculate whether the expected signal is in range, which is not an efficient use of engineering time.


The Diamond Plot tool ( was designed to fix this problem designing with in-amps. It calculates the headroom limits of any chosen in-amp, plots the configuration-specific Diamond Plot and compares it directly against the signal range to determine whether the circuit will operate properly. In the background, it calculates all of the ADI in-amps given the same conditions in order to recommend alternatives that will work. It even has a miniature selection table so that you can compare key specs interpolated for your chosen gain (check it out by clicking the “Recommended” tab and choosing “Filter this list by specifications”). Interpolating specs for a specific gain is something that the parametric selection table on is not able to do.


An Illustrated Example


“Sounds great. Now how in Shockley’s name do you use the darn thing?”


I just happen to have brought a completely made-up circuit to illustrate. Imagine that you are trying to sense current on the low side using a shunt and the AD8226 in-amp, which you wisely chose because of its rail-to-rail output and input range that includes ground. Sounds simple, right?



But you are dismayed to find that the output voltage is not linearly related to the current in the load. What is happening here? Let’s take a look inside the simplified AD8226 schematic.



We can analyze this problem from the endpoints to see where the problems are.

First, if the measured current is zero:

Inputs: Both inputs are at 0V, no problem. That’s why we chose AD8226.

Preamp Outputs: There is a +0.6V shift from the input to the RG pin due to the PNP input transistors, Q1 and Q2, so NODE 1 and NODE 2 are at 0.6V. Also no problem.

AD8226 Output: Wants to be at 0V, saturated at about 50mV (no load). Problem.


Second, if the measured current is 10mA:

            Inputs: -IN and +IN are at 0V and 10mV, respectively, no problem.

            AD8226 Output: Should be at 10mV*200 = 2V, no problem.

            Preamp outputs: Centered at 0.6V, but the difference between NODE1 and NODE2 should be 2V, so the output of the positive-side amplifier A1 is at +1.6V and the A2 output should be at -0.4V. Since the output of the op amp A2 certainly can’t go below 0V, the A2 output is saturated. Problem.


The end result is that this circuit is blind to small currents (where the output is saturated) and large currents (where the preamplifier is saturated).


With much less work and higher accuracy, the Diamond Plot Tool could have told us this:

It is clear to see that the signal (red) crosses the boundaries of the plot, both on the low-end of the signal range, and on the high end of the signal range. As expected, the legend tells us that the output range is violated for small signals, and the internal node limits are violated for large signals.



It’s not a very good design tool that tells you your circuit is broken without helping to fix it, so let’s see if we can make some modifications to get the circuit to work.

First, we can increase the voltage at REF to 200mV to fix the problem with zero input signal. Interestingly, the tool has now found 4 other options, one of which is the pin-compatible AD8227.

Second, reducing the gain to 80 puts the AD8226 just within range for the 10mA input.

We can often leave it at that with no significant loss of performance, but in cases where an extra gain stage is needed to get the gain back up to 200, a simple non-inverting op amp can be used. Just remember to terminate the feedback network to the level-shift voltage at REF rather than terminating to ground, otherwise the level-shift voltage will be gained as well.


That’s all Very Reactive. Can I use the Tool Proactively During the Design Process?


That’s the intention in the end. This tool should save time, not only diagnose what’s wrong. The first criterion of any design is the most fundamental, it needs to work. It is only after that criterion is met that other parameters can be optimized. Stop designing in-amp circuits backwards, optimizing the product selection before knowing if it will work. Start with the Diamond Plot tool and work from a narrowed-down list of in-amps that will work.


It’s better to design the first time with the Diamond Plot in mind than to later find you need to redesign.


Have you used The Diamond Plot Tool to design an in-amp circuit or to solve an existing circuit problem? Tell your story in the comments.

As a Web Tools Engineer, I'm constantly working to ease the design challenges of our customers through our design tools. Sometimes, I'm able to respond to customer feedback about an existing tool, and use that feedback to make improvements to that tool. Other times, I'm able to work with a team of engineers to create a new tool that simplifies a particularly tedious design challenge.


The feedback we receive from customers is one of the more rewarding aspects of my job. Often times, this feedback gives me a better understanding of a design engineer's pain points, which helps me to make these tools even better.


Negative Feedback.PNG

Actually, we'll take whatever feedback you have...

(And, you can get your own mug here)



We acquire quite a bit of feedback about Analog Filter Wizard - the feedback is overwhelmingly positive, which is, in my humble opinion, mostly due to how much effort we made to ensure that the tool was truly easing several pain points in active filter design.


The most frequent feedback we collect for Analog Filter Wizard is:


"It's very easy to use"

This is, by far, the most common feedback we get for Analog Filter Wizard. It's very simple to specify your filter requirements, and quickly create a circuit design. It's also very quick and easy to tweak the circuit - for example, optimizing for noise, power, or voltage level. And, it's wicked fast.


FW Baic Low Pass Gif.gifAdding gif creation to my skillset...






"The tolerance view is very helpful"

Analog Filter Wizard has a particularly handy feature that allows the designer to view the effect of component tolerance on the behavior of the circuit.







"Instantaneous insight into real circuit behaviors"

OK, so we don't get this feedback as often as the first two, but this is one of the things that I think makes Analog Filter Wizard so useful. You can quickly go from specifications to actual circuit behavior quickly, and immediately see how non-ideal circuit behaviors are going to cause your circuit to deviate from the ideal transfer curve. No surprises at the bench.


Multiple Feedback.gif




Have you used Analog Filter Wizard for an active filter design?  Did you find it useful? Let us know in the comments.



ADI's Best of 2015

Posted by lallison Employee Feb 1, 2016


2015 brought many improvements, celebrations and innovations, most notably; 2015 marked Analog Devices 50th anniversary. To reflect on this milestone, we analyzed the year's content to bring you the most sought after resources from our website, community, news and social channels. This list represents your favorite article, blog, news story, video and more.

We present you with... the Best of 2015.

Top Downloaded Technical Article:

"Designing Power Supplies for High Speed ADC"

In this article, Rob Reeder describes techniques to measure a converter's AC power supply rejection and provides some useful tips to maximize PC board noise immunity to supply changes. Ready to tackle this challenge head on? Read on...


news item.png

Most Viewed News Item:

"Analog Devices Improves Monitoring and Protection of Smart Grid Transmission and Distribution Equipment"

This announcement introduced a 24-bit data acquisition SoC series designed to improve the performance of protection, monitoring and power quality measurement equipment. Read all about it!



Most Used Design Tool:

"Analog Filter Wizard"

Users went to the Analog Filter Wizard to design low-pass, high-pass or band-pass filters with actual op amps in minutes. Try it for yourself.


Most Viewed Blog:

"Panic, Cry, Do it Anyway"

EngineerZone started a new blog series in 2015, The Engineering Mind. In Panic, Cry, Do it Anyway; Zach.Klimczak shares three steps for tackling complex engineering problems. Read on...


Top Tweet:

top tweet.png


Top Downloaded Data Sheet:


It's our highly integrated RF agile transceiver, the AD9361. The device combines a RF front end with a flexible mixed-signal baseband section and integrated frequency synthesizers, simplifying design-in by providing a configurable digital interface to a processor.

Most Viewed Video:

"Thermocouple 101: What is a Thermocouple?"

This one was uploaded in October 2011, but is still going strong. Out of hundreds of videos on our YouTube page, this one garnered the most views in 2015. Take a look...


Most Popular Facebook Post:

top FB post.png

In closing I leave you with the Oddest Search Query of 2015: "do i need a signal processor on a Harley?" I certainly hope that individual wears a helmet.


To see all of this great content and future contributions be sure to visit our websites and social channels.

Read a technical article on Analog Dialogue

Follow our Blogs The Engineering Mind, EngineerZone Spotlight and Analog Dialogue Blog

Like us on Facebook

Follow @ADI_News on Twitter

Visit our new News Room!

Visit our website,

Since ADCs straddle the Analog and Digital domains, the common belief has been to separate the analog and digital domains as much as possible. This methodology has its roots firmly planted in the CMOS era where the supply domain bounce caused performance degradation in the Analog area. It is for this reason that ADC eval boards until recently have had LDO's and had an almost universal ban on having DC-DC converters in them. The DC-DC converters were thought to introduce switching spurs and corrupt the ADC measurements. LDOs on the other hand, do not have switching circuits, but they are a very inefficient way to design a PDN (Power Delivery Network). A lot of our customers tend to re-use our customer evaluation boards as-is, and have expressed frustration at the inefficiencies (both electrical and thermal) posed by the LDOs. Not to mention the system cost due to the expanded BOM. For example, take a look at the AD9680 customer evaluation board PDN.


Figure 1 : Default PDN of the AD9680 Customer Evaluation Board

This PDN was a "safest design" approach to dealing with GSPS sampling while at the same time, maintain "optimum" ADC performance.
I disagree with the notion to keep this level of separation for all the different domains. Trust me, we do have to be aware of the analog and digital domains and give them the respect they deserve. But it isnt necessary to go all out and separate all analog and digital supplies in the name of "performance". In "DC – DC Converters Offer Efficient Power Delivery Networks for GSPS ADCs" I talk about how, with careful design and layout considerations, we can still have optimal performance with reduced BOM. This is a tremendous cost savings to a systems designer who may have multiple High Speed ADCs in his system. See figure 2 for the simplified PDN.


Figure 1 : Simplified PDN for the AD9680 Customer Evaluation Board

I was told to not make this article product specific, hence no names mentioned (in the article). A more detailed article will be published in Analog Dialogue in February 2016.


Let me know your comments.


Future Day - October 21, 2015

Posted by IanB Employee Oct 19, 2015

This Wednesday October 21, 2015, fans of the iconic 1985 movie “Back to the Future” celebrate ‘Future Day’. October 21, 2015 was the target date from Back to the Future I and II where Marty, Doc and the Delorean™ time machine fast forwarded to a point in time that was 30 years into the future from October, 1985.  There were many technical prognostications that were assumed we would have today.  Some of them are indeed a reality, while others are still a bit futuristic.


You decide.  Have these events and technologies come to fruition?


1985 - Prediction                                                                                     2015 - Reality

“Roads? Where we’re going, we don’t need roads…”           We still pretty much need roads.  We haven’t even solved pot-hole issues.
Travel by aero-cars that propel passengers in the sky          No.  However, the driver-less car was not part of the vision in the movie.

Flat screen televisions and 3-D displays                                Yes!  A good prediction - flat displays and 3-D movies are commonplace.

Auto-lacing sneakers                                                              Not yet. This still seems like a potential big seller to today’s youth.

Mr. Fusion – Turn cans and banana peels into energy          No. Alas, I have not seen a cold fusion system on the shelves yet.

Biometric Identification for hand and fingerprint                   Yes, for opening locked doors, accessing computers/smart phones

Hoverboard – Like a skateboard, but without wheels            No.  Gravity still seems to be fairly pesky 30 years later
Ubiquitous Communications                                                  Yes!  But with a mobile focus, fewer faxes and phone booths

Flux Capacitor - 1.21GW Power Consumption (1985)           ??

Chicago Cubs sweep World Series                                         Although it was not a sweep, the Chicago Cubs end 108 years of losses and

                                                                                                win the World Series in 2016.            

What about other technologies such as High Speed Analog to Digital Converters?

How have they fared in technological advancement over the last 30 years?

Let’s compare today’s state-of-the-art high speed ADCs to the AD9000, Analog Devices’ highest speed flash ADC of 1985.


                                                  1985 - AD9000                                                2015

High Speed ADC                                ‘High Speed’ 6-bit A/D Converter 77MSPS             AD9625: 12b 2.6GSPS

                                                                                                                                                    AD9680: 14b dual 1GSPS  

                                                                                                                                                    AD9652: 16b dual 310MSPS

Full Power Bandwidth                20MHz                                                          AD9625: 3,200MHz

SNR                                           33dB                                                             AD9652: 73.7dBFS                    
SFDR                                        38dBc                                                             AD9680:85dBc

Input Capacitance                     35pF                                                              ~1pF

Voltage Supply                          Need both +5V, -5.2V                                    1.25V, 2.5V

Package                                    16L Ceramic DIP (2.5mm pin pitch)              AD6676: 80L Wafer Level CSP (0.5mm pitch)

Datasheet Pages                       8                                                                    60-100 + links to other specifications

# of registers                              0 (no control communication)                       100's - 1,000's


Looking back 30 years, we have come a long way for both technology in general and ADC performance.

What will the next 30 years bring?  Until then, enjoy the future...


Ian Beavers


Analog Devices and MIT are in full swing with SOLVE and HubWeek. Throughout this week we will be highlighting stories from ADI engineers that embody the four pillars of innovation - Learn, Cure, Fuel, and Make.


We complete our SOLVE blog series with an entry by Jarrett Liner, a Systems Application Engineer, located in Greensboro, NC, talking about planes (or airborne platforms), and how the engineering concept SWaP can affect these platforms and their performance in the air.

edit suitcase.jpgEver wonder how airlines determine how many bags should be allowed on a plane? If you are an avid traveler then this may interest you.


One area of common concern for commercial (and defense) airborne platforms is maximizing payload efficiency. Every ounce of weigh, cubic centimeter of space and mili-watt hour of power is carefully planned. The focus is on Size Weight and Power or SWaP. Advances in RF technology can provide a leap-frog advantage for commercial and defense airborne platforms, manned and unmanned. To learn more about recent advances in semiconductor technology and component integration, check out this new article: SWaP: The RF solution that can mean the difference between flying high and being grounded.


In the consumer world, this will help explain why we pay for baggage, why overhead compartments are configured the way they are on each type of plane, and why flight attendants follow such stringent rules on the number of bags on the plane compared to under the plane.


So, next time you’re asked to cheEdit plane.jpgck your rollerboard, think twice before complaining – it probably has something to do with SWaP.


In the defense world, the implications of advance RF technology also relates to safety and more. When airborne platforms need to stay, well, airborne, engineers will continue to investigate and refine technology that will keep us all safer (and maybe convince us to pack lighter).




Click to Tweet: Catching a flight? Learn more about SWaP before you pack. @ADI_news engineers #SOLVE to making flying safer.

Filter Blog

By date: By tag: