Advertisement

Article

Going through a mixed-signal SOC design, from A to Z (Chapter One)

(Editor's note: we are resuming our “dialogues” between these two distinguished engineers with this special, multipart series on the product development process. This is based on a real-life example and hands-on reality, not speculation or an academic perspective. If you want to see the rest of these dialogues, or other articles by these authors, you'll find a linked list here .)

[The setting: Dave Ritter has received a rare official meeting notice to join Tamara Schmitz (Dr. T) for an unspecified discussion …]

Tamara Schmitz : Hi Dave, glad you could make it … you look a bit puzzled.

Dave Ritter : I guess I am.  We usually bump into each other in the lab or you drop by my office.  This is the first time you’ve called an official meeting, what’s up?

Dr. T :  I’ve been thinking about trying to become an engineering manager.

Dave : Really?  You want to switch from doing all the interesting work to managing it?

Dr. T :  Maybe.  I think it might be a good challenge for me and I need to know more about the conception to birth of an integrated circuit.  From what I’ve seen, the best managers know—can even anticipate—issues.  Can we go over it together?

Dave : It will be easier if we focus on a product and then make comments to help you generalize

Dr. T :  Sure.  That sounds great!  You just released a great new video equalizer.  Let’s use that. 

Dave : You mean MegaQ?

Dr. T : That’s the one, the ISL5960x series of video cable equalizers, (Figure 1 )


Figure 1: The block diagram of the video equalizers, the ISL5960x family.
(Click on image to enlarge)

Dave : I’m ready.  Tell me more about what you want to do.

Dr. T :  We want to follow the project from the very start of the idea through the release and use in customer applications.  We want to take our readers through the entire conception and development, in six installments.

Dave : Wow!  I’ll have to think back.  The idea started with Kathryn and Fred.  Kathryn (Video Product Line manager in Marketing) wanted an equalizer for the security industry.  Fred (Video Applications Manager) decided it had to go a mile. I thought they were crazy.  I checked some numbers and the attenuation at a mile for an example cable like Category 5 (Cat5) was over 60dB.  But my job was to work on ‘Strategic Applications’, which means I was supposed to develop and define new ICs.  So I started working the problem.

Dr. T :  So the original idea came from the marketing manager.  Does it always come from marketing?

Dave : No, not really.  Product ideas can come from sales meetings, marketing investigations, customer visits, or from somebody in apps or design.  We’re never sure exactly where the next one will come from, but they need to get to marketing so we can see if they make sense.

Dr. T : So MegaQ must have made sense.

Dave : We already had a series of cable equalizers for the KVM part of the video business, but we didn’t have a solution specifically for security video.  Our FAEs reported that people were installing security systems with longer distances between cameras and consoles.  They were using less expensive cable.  In particular, some of them were using Cat5 network cable instead of traditional coax.  From the marketing standpoint, a video equalizer for security would make a good part especially if it could handle either coax or Cat5 cable. 

Dr. T :  The idea sounds like it is starting to become defined.  How can you determine if it is even possible?

Dave : Some of our apps people had worked on discrete solutions, so I spoke with all of them and got valuable inputs.  There was no monolithic solution, not even from our competition, that went anywhere near a mile.  That’s when I started working the numbers, which led to some more serious mathematical models.

Dr. T :  So you determined that it was technically feasible from a mathematical model?

Dave : Not right away.  The numbers looked encouraging, but I like to see it working in the lab.  So I built a prototype based on my models.  That took a couple of months.  (See the prototype in Figure 2 .)


Figure 2: The prototype of the video equalizer.
(Click on image to enlarge)

Dr. T :  And then you knew it was technically feasible?

Dave : Yes.  And so did anybody else that wanted to come to the lab.  That’s the great thing about prototyping:  it stops a lot of arguing.

Dr. T :  So we start with an idea (something our customers need), we work out the technical feasibility, and then we design the chip?

Dave : Well, it’s not that easy.  We had an idea and a prototype.  We needed to figure out how much it would cost to develop the chip and what it would be worth in the marketplace.   We needed to determine who the major customers were and how many they might use.  In other words, we had to work out the product schedule, cost, and volume to see if it made business sense.

Dr. T :  That sounds a lot like a business plan.

Dave : Exactly.  We gather our knowledge at that point, put together a preliminary data sheet, and draft a business plan.  It’s up to the business folks to decide if the project is worth doing.

Dr. T :  And MegaQ was worth doing?

Dave : Absolutely. 

Dr. T : Okay.  Here’s my high-level list so far:

  • Product Concept, from marketing (in this case).
  • Technical Feasibility, including prototype (in this case).
  • Business Plan, including cost, schedule and potential market data.
  • IC Design, putting together the transistors and simulating the result

Dave : Yep, but it doesn’t stop there.  As we’re designing we have to be developing a test strategy.  An old colleague used to say, “if it ain’t tested, it don’t work”.  That is true of ICs for sure.  We test the prototype to see if we have the concept and system details correct, we test the silicon design using simulators, we test first silicon at the bench using video test equipment, and we test in the ATE for initial characterization and then again for final production tests applied to every part. 

Dr. T :  So I should add Test Development, ATE Characterization and Bench Verification.

Dave : And there’s one more buried in there.  ATE and bench tests are both used to characterize the first silicon, and they have to agree.  We have a task called ‘Correlation’ to resolve any differences between bench and ATE measurements.  When we’re all done, we have full confidence that the ATE is thoroughly testing all the parts.

Dr. T :  My list grows ….

  • Test Development, including built-in test
  • Bench Verification, testing the first parts using video test equipment
  • ATE Characterization, making sure the bench tests and ATE agree.

But there must be even more.  At this point we have a chip that is tested and works.  How does the potential customer evaluate it?

Dave : While all this testing and verification is going on, we have to produce test boards for our customers so they can easily see the performance of our chip in their labs.  Along with the hardware we need a good data sheet and applications note to describe the function and application of the part.  And it helps to have a few articles in trade publications to get the word out.  And then there’s supporting material on our Web page.  All of these things need to get done and orchestrated for the official release of the part.

Dr. T :  So my list grows again ….

  • Customer Evaluation boards, designed and built by applications
  • Final Data Sheet with specs, curves, and apps section, written by applications
  • Promotion and Support materials, orchestrated by Marketing Communications

Clearly the problem is bigger than designing a circuit.  We have to test, verify, promote, communicate, visit customers … the whole kit and caboodle.  It’s a big deal. 

Dave : That’s why it takes so long.

Dr. T :  How long does it take?

Dave : I’ve witness projects delivered as fast as 6 months or take a few years.  It all depends on the complexity and what road blocks pop up along the way.  The product doesn’t release until everything is right.

Dr. T :  And why we need to have a few conversations to cover the whole project.  This was a great overview.  Let’s go through it with respect to your project, MegaQ, to better understand the steps. 

Dave : Works for me.  By the way, this could take a while.  Will there be refreshments?

Dr. T :  Here, have a Lollipop … (Tamara does, in fact, keep Tootsie Pops in her office to encourage visits from Dave.)

About the authors
Dave Ritter grew up outside of Philadelphia in a house that was constantly being embellished with various antennas and random wiring. By the age of 12, his parents refused to enter the basement anymore, for fear of lethal electric shock. He attended Drexel University back when programming required intimate knowledge of keypunch machines. His checkered career wandered through NASA where he developed video-effects machines and real-time disk drives.

Finally seeing the light, he entered the semiconductor industry in the early 90's. Dave has about 20 patents, some of which are actually useful. He has found a home at Intersil Corporation as a principal applications engineer. Eternally youthful and bright of spirit, Dave feels privileged to commit his ideas to paper for the entertainment and education of his soon to be massive readership.

Tamara Schmitz grew up in the Midwest, finding her way west with an acceptance letter to Stanford University. After collecting three EE degrees (BS, MS, and PhD), she taught analog circuits and test-development engineering as an assistant professor at San Jose State University. With 8 years of part-time experience in applications engineering, she joined industry full-time at Intersil Corporation as a principal applications engineer. In twenty years, she hopes to be as eternally youthful as Dave.

0 comments on “Going through a mixed-signal SOC design, from A to Z (Chapter One)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.