Advertisement

Blog

20 Years of Change in Analog Simulation Tools

In January, I reviewed the Dennis Fitzpatrick book Analog Design and Simulation using OrCAD Capture and PSpice for EDN. I've also provided three excerpts from the book for EDN (Here’s the third, published last week.)

I shall not repeat all the points I made there about Fitzpatrick's book. However, given the context of this forum, I did want to discuss the chapter on mixed-signal simulation. If you're going to integrate analog and mixed-signal functions, you'd better get them right.

That's always easier said than done. Anyone remember Mentor's LSim? I was the technical lead for that project for several years after Mentor acquired it from Silicon Solutions back in the early 1990s. One of the interesting aspects of LSim was the way it handled mixed-signal simulation. All the solvers (analog, transistor timing, gate, behavioral) were unified in the same kernel. The basic component models were written using its own behavioral language, so it was at both the top and bottom of the hierarchy, so to speak. If you wanted to connect an analog net to a transistor, you just went ahead and did it. It would become a mixed node, and it was possible to define aspects such as the digital rise and fall slopes and threshold voltages. During simulation, the two subnets were evaluated and combined in a hierarchic manner.

Chapter 19 of Fitzpatrick's book describes the way this is done in PSpice. Interface nodes (defined as those that contain both analog and digital signals) are automatically separated into one analog node and one or more digital nodes by inserting analog and digital interface subcircuits — either analog to digital or vice versa. These subcircuits also have their own power supply.

An analog-to-digital converter is nothing more than a threshold detector. It could also present a resistance and capacitance to the driving circuit. A digital-to-analog converter works with ideal digital components that have just a few parameters associated with them: voltage, resistance, and capacitance. We could make very minor adjustments to the gate output voltages and rise times based on the load that the analog circuitry provides to adjust the slope of the output signal, but none of this is beyond what LSim provided. If you had a very special case in LSim, it was possible to model the node using the behavioral language.

Unfortunately, the chapter does not provide any more guidance on how this works, so I resorted to the PSpice user's manual. It shows that the interface models are basically the IO characteristics of the logic family (TTL, ECL, etc.), and that up to four levels of accuracy are provided for each. LSim provided a sampling rate on the interface that allowed any number of accuracies to be defined.

It seems as if there has been no change in the basic way mixed-signal interfaces are defined over the last 20 years — perhaps, if anything, a little simplification has occurred.

A few days ago, I talked to some folks from Analog Rails (about a completely different subject). They said that analog designers are stuck in the old ways of doing things, and that they had a very different way of handling analog design and simulation that would increase productivity.

Having been away from the analog world for two decades, I have to say that nothing much seems to have changed. What changes have you seen in the past 20 years? What changes do you want to see in terms of both simulation and functional integration?

28 comments on “20 Years of Change in Analog Simulation Tools

  1. eafpres
    March 23, 2013

    Hi Brian–it seems that products like the Ansoft family now include both EM simulation and additional post-EM tools for circuit simulation all the way to signal integrity (they say up to 10Gb/s).  Do you think tools like that can handle mixed signals more effectivly than just using SPICE?  I'm not expert in this and the differences are unclear to me.

  2. RedDerek
    March 23, 2013

    I have seen analog changes move to more of a block function. Especially when it comes to power supply circuit simulations. I find it nice to see a long-time simulation finish quicker than if it was all discrete.

  3. BrianBailey
    March 23, 2013

    I beleive I will be exploring that world as I dig deeper into understanding where things are today, what has changed and where the industry as a whole is perhaps clinging onto old paradigms that should have long since been replaced. We shall see how much trouble I cause as I start kicking around the edges…

  4. Brad Albing
    March 24, 2013

    You'd think the designers would have made that aspect changeable (“stop program with any keypress? – Y or N”). But maybe that's just me, thinking that everyone else wants designs done just as I would do them.

  5. Brad Albing
    March 24, 2013

    Just to clarify – so in some simulations, a power supply would not be detailed as to its inner workings, but would just be a “perfect” voltage source?

  6. goafrit2
    March 24, 2013

    >>  Do you think tools like that can handle mixed signals more effectivly than just using SPICE?

    Essentially, most of these systems are extensions of SPICE. There is no major circuit simulator that is not built upon SPICE. Simply, they can do better job at least in the presentation of results.

  7. goafrit2
    March 24, 2013

    >> I bought my first copy of PSPICE for $495 on a 5-1/4 floppy disk

    Today, you can get LTSPICE for free. This is the age where companies are giving away IPs which they hoarded for ages to get competitive advantage in the marketplace.

  8. goafrit2
    March 24, 2013

    >> I have seen analog changes move to more of a block function.

    I do not understand your point. Analog design has always been block function. It does not make sense to sim a transistor when you already know and understand the process. So, the trend has always been to make systems out of transistors.

  9. goafrit2
    March 24, 2013

    >> what has changed and where the industry as a whole is perhaps clinging onto old paradigms that should have long since been replaced.

    Brian, it will be good to follow you in this excursion. I do think the world of analog simulation has not changed a lot. Except that the tools have improved, the concept has largely remained intact.

  10. goafrit2
    March 24, 2013

    >> Just to clarify – so in some simulations, a power supply would not be detailed as to its inner workings, but would just be a “perfect” voltage source?

    Good question. But unless someone has made that Plug n Play power supply for you and qualify it. Otherwise, it pays to know what is inside the box. I have also come to realize that fast tech adoption makes it harder. As you move to lower process feature sizes, everything has to change.

  11. amrutah
    March 25, 2013

    Yeah, as the industry is growing, so is the complexity and density of ICs. The designswith large no. of nodes need to be more compute intensive, parallel processing, different netlists like integration of Verilog-A, VHDL netlist at one level with spice netlist at other level.  All these have lead to many analog simulation tools, with the base being SPICE.

      With technology scaling many short channel effects are seen to be degrading the reliablity of ICs, which needs to be checked.  The Analog EDA engineers have contributed a lot in this regards, many kudos!

  12. Brad Albing
    March 25, 2013

    Exactly – LinTech hopes you'll use their tools and then buy some of their parts. Seems reasonable.

  13. Brad Albing
    March 25, 2013

    OK, that clarifies my thinking a bit. Lucky for me, I don't need to do any designs at the transistor level. Just the occassional analysis of an active filter or switching power supply.

  14. eafpres
    March 25, 2013

    @Scott–my impression is that Phase 3 is where tools like Ansoft's suite or AWR's products (i.e., Microwave Office) might come into play for very high speed circuits.  I'm not convinced by the earlier statement that all circuit simulations are souped up SPICE.  Does SPICE allow for radiation?  In a mixed signal environment, for high data rates, you have GHz level frequencies even if the waveform is “digital”.  

  15. Brad Albing
    March 25, 2013

    It keeps coming back to “It's all analog.”

    Good question about the ability to simulate the effects of radiation. It does seem theortically possible. But has anyone coded the simualation for all of our favorite devices?

  16. David Maciel Silva
    March 25, 2013

    Good PSPICE model was the first form of simulation for analog circuits that existed, he opened the doors for many applications that are successful today.

    Highlight the simulators I've ever used: Proteus and Multisim, with a little more affection by Proteus, which can simulate source code for microcontrollers from different manufacturers.

  17. Brad Albing
    March 25, 2013

    Interesting, descriptive acronym. Unfortunate that it works out as that particular “nym.”

  18. goafrit2
    March 26, 2013

    >> The Analog EDA engineers have contributed a lot in this regards, many kudos!

    It is a good comment because most times with the dominant position of Intel, it seems analog guys are invisible.

  19. goafrit2
    March 26, 2013

    >> Good PSPICE model was the first form of simulation for analog circuits that existed,

    I never used PSPICE because I did not find it very intuitive. Perhaps you have explained the reason. It was the first and others might have built upon it. Today, LTSPICE is becoming a campus defacto.

  20. goafrit2
    March 26, 2013

    >> CANCER stands for Computer Analysis of Nonlinear Circuits Excluding Radiation!!

    That is truly not a clear name to have in a lab or book. The guy that came up with that must be having a bad week. Thank Heavens I never read that up and will like to forget the acronym.

  21. Brad Albing
    March 27, 2013

    Incidentally, can you compare Spice and LTSpice for me? I'm wondering about ease of use and new devices being introduced into each. And compatibility between each version.

  22. Brad Albing
    March 27, 2013

    Per my comment in a different thread, it would be good to compare and contrast differnt versions of Spice or other software tools used to evaluate ckts.

  23. goafrit2
    April 9, 2013

    >> FYI…SPICE has been available for free since the 1970's.  It's open source.

    Who invented it and why did he made it open-source? That has more value than Microsost Windows and you know how Windows turned out to be for Mr. Gates

  24. goafrit2
    April 9, 2013

    >> Incidentally, can you compare Spice and LTSpice for me? I'm wondering about ease of use and new devices being introduced into each

    I was thinking that LTSPICE is a flavor of SPICE. I know there are many SPICE versions and LTSPICE is one from LT. LTSPICE is great, it does not need a lot of resources and works  perfectly.

  25. goafrit2
    April 13, 2013

    >> Would SPICE be as widespread today if it were not freely distributed for the past 50 years?

    That was certainly the period when government funded many things and schools and companies could make things free. If a VC had funded the discovery, it would have been commercialized. That is the challenge – one is to diffuse knowledge, another is to make money.

  26. goafrit2
    April 27, 2013

    >> Exactly – LinTech hopes you'll use their tools and then buy some of their parts. Seems reasonable.

    And that strategy is working. LTSPICE has introduced many designers to LT products without any extra marketing budget. The question has been why they needed to release that tool to arm their competitors. It turns out that they are smart in their strategy.

  27. Brad Albing
    April 30, 2013

    I've had the opportunity to chat with the folks at LTC. As you noted, they certainly know what they're doing regarding marketing.

  28. George Storm
    May 15, 2013

    “Good PSPICE model” was by no means the first form of simulation for analog circuits that existed.  PSPICE is a 1984 derivative of SPICE2 that was released for the PC in 1984.  It would be correct to say that PSPICE was the first SPICE derivative that was designed to run under DOS.   Berkeley SPICE 2 dates from about 1975 (and it's actually the third generation simulator that was 'built' at Berkeley).   In order to make it easier to use with the  models of the time, PSPICE was deliberately crippled: the default method algorithm for SPICE was (and is) trapezoidal integration – but this had a tendency to crash when a slope discontinuity was encountered.  PSPICE originally (and for a long time thereafter) only allowed Gear2; this is both slower than trapezoid for a given level of accuracy.  In addition, Gear 2 tends to make unstable circuits look stable – ouch, so not so good (as well as not so old).  Oh, and unlike HSPICE, PSPICE did not introduce significant new models.

Leave a Reply

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