Multiple PCB Boards, One DxDesigner Project: The Coil Winder (Part 2)
Posted by dave_brady Jun 23, 2009Multiple PCB Boards, One DxDesigner Project: The Coil Winder (Part 1)
Posted by dave_brady Jun 22, 2009The new 2007.x version of the DxDesigner / Expedition design flow has more power and flexibility than I was aware of. I was working through some PCB designs and I made a comment to one of my MGC engineering friends: “The new DxDesigner is great! But wouldn't it be nice if I could design multiple Printed Circuit Boards (PCB) within a single project?”
Well, he was kind to me. He gently informed me that single project, multiple PCB design was a reality and that I simply lacked the tool expertise to exploit these advanced features. Once the basic concepts were explained to me, and trust me, there is no real “magic” here, I was able to quickly create complex, multi-board system design with the DxDesigner - Expedition design flow.
Once I worked through the process I realized the true limitation: The limitation was me, I have a long history of EDA tool usage and my arrogance manifests in my steadfast refusal to attend training or read the product documentation (sound familiar?). I have long held the belief: If I am unable to figure out how to use the software to accomplish my objectives then, the software isn’t very good (because I am 
). OK, every once in a while life slaps some sense into all of us 
Since it is impossible for me to believe that I am alone in my perspective and attitude regarding EDA tools I thought the best approach to recover from my arrogance was to create a demonstration highlighting the multi-board, single project capabilities of the 2007.x Expedition design flow.
Since I am a true engineer (nerd, geek, pick your adjective) I have created a demonstration around one of my hobby projects: My Wind Turbine. Due to the length of the design process I will break this demonstration process up into multiple segments within each blog, and multiple blog chapters. If you follow my Blogs for the next few weeks you will not only learn how to use DxDesigner and the Expedition flow to create multi-Board PCB Systems, you will also learn how to build the electronics for a coil winder used in the construction of home built wind turbines.
In part 2 of this Blog I will review the design requirements as well as the creation of the DxDesigner Project and the top level system diagram. I like to use top-down system refinement as my design method so you will also see how DxDesigner may be used to accomplish more than simple PCB design.
As we plow through this process please share your perspectives!
Recently I took some time to examine the capabilities of Verilog-A in the context of PCB design. Verilog-A is an analog behavioral modeling language that has syntax & style conformity with Verilog (for digital designs).
After spending some time with Verilog-A, modeling and simulating components of a windmill power generation system I built, I developed some perspective:
I don't think there is anything you could not model with Verilog-A (really, it has that much functional depth)
At a behavioral level Verilog-A models are easier to develop, understand and share than SPICE models
For primitive level circuit topologies I remain a fan of the schematic & pure SPICE approach.
The ability to mix SPICE and Verilog-A models in one design netlist and simulate is absolutely critical in the PCB domain (Almost 100% of available analog models are some flavor of SPICE)
All of my evaluation experiments were based on HyperLynx Analog with the Eldo simulation kernel and the Verilog-A simulation option.
I've been attempting to create a simulation model of the electrical generation capabilities of my windmill to facilitate development of an "intelligent power storage and distribution" controller. The elemental characteristics of the wind power generation head are:
Three Phase AC generated
Frequency increases with wind speed
Power (Voltage & Current) increase with wind speed
Stator Temp increases with wind speed (current)
Max Stator Temp is 180 degrees C
From a modeling perspective I added the following requirements
Must support Transient Analysis
Input to model: DC voltage establishing wind speed (allows wind speed to be a control variable through transient analysis)
Must be easy to adjust (As I get better characteristic data from my windmill I need to improve the accuracy of the model)
I started to create the model with SPICE:
Yes, it is possible (not easy) to create the model
I don't see any technique that would allow the model to be easily adjusted
The resulting model is not easy to decipher
Then I turned to Verilog-A as an absolute novice.
A few web searches and I had dozens of Verilog-A examples in front one me
An examination of the Mentor Supplied Verilog-A help "Modeling Cookbooks" and I had code ready to copy and paste
The model design:
Clearly the wind generation head could be modeled as a form of a Voltage Controlled Oscillator (VCO)
I started with a Verilog-A VCO design example I found on the web
I modified the model to include a Power (Amplitude) generation equation (linear).
I then parametrized all of the equations to create an "easy to modify" model.
The results:
Generates a proper 3 phase AC output with Voltage & Frequency varying with wind speed
Fits the characteristic data (limited data set)
Further data collection may invalidate the linear power model
Simulates quickly (simulation run of 10 seconds completed in the blink on an eye)
Works well with SPICE models (test bench connected a 3 phase rectifier modeled in SPICE)
Areas for model improvement:
Model of Power is incomplete
Good physical models for power extraction from the wind exist (non-linear) and should be incorporated
Verilog-A: Analog modeling made easy:
Verilog-A is easy if you have any HDL experience and any form of software coding experience (Basic, Perl, Tcl, etc... does not need to be C++ experience)
Verilog-A has the potential to significantly benefit PCB design teams.
Difficult to create SPICE models are fast and easy to create with Verilog-A
Verilog-A is not some new unproven technology ==> Stable and proven
Verilog-A simulation is commonly available
Verilog-A shares a common benefit with HDL languages in that it is self documenting (although some style improves communication significantly)
Why is Verilog-A easy?
Verilog-A characteristics:
Ports: Inputs, Outputs and Bi-directional (inout)
Parameters: Constants that may be modified outside the model by the simulator
Variables: just like any other software language (typed)
Equations: Create any equation your mind is able to dream up
Functions: Create any function you desire
Control statements: If-then-else (hint: research "begin-end" blocks before your expressions become too complicated)
Native support for the electrical domain (discipline)
Yes, Verilog-A supports simultaneous differential equations
A component modeled with Verilog-A is no different than a component modeled with SPICE from a "design creation" perspective with tools like HyperLynx Analog
With co-simulation of SPICE and Verilog-A
Summary
I picked a model I was having difficulty creating using pure SPICE. I turned to Verilog-A and quickly created a simulation model that met my requirements and used the model in a mixed Verilog-A & SPICE design.
In my case the difficult model was needed as stimulus (source) to the system PCB design process. Once I had some experience with Verilog-A I found that quickly creating missing simulation models was easy with Verilog-A: High current DC relays, battery isolators, 3 phase knife switches (shorts all three phases to break the windmill and lower RPMs) and my current focus: energy storage (DC battery models).
Adding Verilog-A to your PCB functional simulation "tool box" requires a minimal training investment (with the typical engineering background) and greatly accelerates design flow productivity. For "in-house" OR component supplier model development teams Verilog-A should be considered as a primary analog modeling language. The self documenting benefits of Verilog-A over SPICE should be one monumental motivation but I suspect the real attraction will result from the rapid model development for complex components.
If anyone is interested in the Verilog-A model for the windmill electrical generation head drop a note and I will send it to you. Obviously, we make no warranties and assume no liabilities for the use of the model. 
Do You have Experience with Verilog-A in the PCB design process? Share your experiences and insight by commenting to this blog or Joining the MGC PCB community and creating your own discussion thread. As always, don't be shy because you are not using HyperLynx Analog, all experiences are welcome here as the goal is to move design productivity and first pass success forward for all.
I happened to be on the MICROCHIP web site downloading and testing their SPICE models for HyperLynx Analog when I came across the PIC microcontroller. I decided to go ahead and order their PICKit 2 starter package (the cost was right )
It's been a while since I have done any assembly level programming and I reckon I had an itch that needed some scratching. To give you and idea of how long, the last microcontroller development system I purchased was the Rockwell AIM 65 (6502 based). I have it sitting here in a box and every couple of years I break it out, consume the kitchen table and have a little fun.
The PICKit 2 arrived last Thursday and I was amazed when I unpacked it... it is so tiny! No more consuming the kitchen table... it fits neatly on my desk in front of my laptop. That would be both the programmer and the development board consuming no more than 2.5 in x 7 in of desktop space. The programmer is USB based and connects to a 6 pin header on the development board. It really is a sweet little package.
MICROCHIP does a fantastic job supplying software, examples, datasheets, application notes... everything you need to get going. In fact, the Integrated Development Environment (IDE) MPLAB is AMAZING!! It includes comprehensive development tools including a language sensitive editor, Instruction Set simulator, debug and device programming capabilities in a well organized and easy to use interface. MPLAB is also FREE to download off of the MICROCHIP site
The development board contains a 16F690 20 pin PIC microcontroller. This is a Flash based device that has Analog to Digital converters, timers, built-in oscillator, EEPROM, Watchdog timer, Brow-out reset, RAM registers, and a nice allocation of program memory. What was amazing to me was how inclusive the PIC is, you really only need a 6 pin header (for programming) and a power supply capacitor to configure the microcontroller... you don't need an external oscillator although it is configured to accept one.
The PIC architecture is a Harvard based RISC architecture originally designed by General Instruments in 1971. The PIC 16F690 is a 14 bit instruction design with approximately 36 instructions. I took a look on the DigiKey site and the cost is extremely reasonable at $3.05 for single units.
Today is Saturday and I have the instruction set well understood, the Analog to Digital capability operational, complete control over the bi-directional ports and I am ready to make significant progress on my needful project.
One year ago I decided that I was going to design and build a power generating windmill. Over the summer I completed the fabrication and erected my first 40 foot tower with a 10 foot diameter windmill... it was pure dedication that allowed me to complete the project. A couple of weeks ago the wind really kicked up (60+ MPH) and my home made composite blades snapped sending one of the blades over 300 feet away. So, while I was redesigning my blades and hub I decided that I needed some form of feedback control system to pulse brake the windmill in extreme wind conditions. I started the design as a pure analog design (since that is where my current skill set is strongest) and I was running into some interesting boundary challenges as the alternator I built increases in voltage and current as the wind speed increases causing my frequency monitor design to increase in complexity. While I was enjoying the design challenge I was not looking forward to the fabrication as I like to keep my designs very simple. The introduction of the PIC microcontoller into my design will not only simplify my frequency monitor design but will also provide useful in extending my design to include some smart pulsed battery charging techniques as well.
Yes, this blog had really nothing to do with Mentor Graphics Board Design Tools. Clearly, I use our Board Design suite to do more than create Video Blogs and Workshops!! In the not so distant future expect to see a Video Blog combining my Analog designs with a PIC microcontroller into a PCB designed using the MGC tools.
There is nothing easy about Analog functional design! However, that does not imply that learning a new PCB functional simulation tool should be difficult!
If you have been following my Blogs you will see that I have been attempting to provide Video Blogs. The challenge is providing high resolution video over the internet on-demand and quite honestly while I do record high resolution once the Videos have been uploaded and magically massaged through compression algorithms the resulting resolution is disappointing
Fortunately, I am a decedent of that rough & ready community that simply refuses to give up. Through the adoption of new Web 2.0 technologies I think we may have achieved a break through and produced the step by step instructions in a high resolution format that will service your quick adoption needs:
Workshop 1 takes you all the way through the import of a component Vendor web based SPICE model, circuit creation, simulation setup to analyzing the simulation results => A comprehensive overview
Workshop 2 builds on the results of Workshop 1 demonstrating DxDesigner's ability to explore multiple circuit topologies within a single project, simulation test bench management as well as advanced waveform calculations
Tell me what you think! Am I drinking my own cool-aide or does this format provide the high resolution, on-demand, assistance you need to quickly adopt new powerful technologies?
The following videos will show you how to:
Import a web based SPICE model into HyperLynx Analog
Auto-generate a symbol for the model
Assign a preexisting symbol to the simulation model
Check the symbol properties
Map the symbol pins to the simulation model ports
Create a test circuit to test the new model
Quickly setup simulation sources
Setup the DC, Transient and AC analysis
Run the simulation
View and analyze the results
I hope you found this Video Blog useful
The Title is not a joke... I am going to prove this is a true statement!
For a moment let's simply the FPGA - PCB Interface design to just PCB Signals (traces) and FPGA signal pins. Let's ignore the power pins, the configuration pins, the ground pins and all of the possible electrical configurations for each signal pins (IO Standards, drive strengths, etc..), pin banks and pin bank signal configuration rules.
For a FPGA - PCB Interface design with "N" PCB signal traces and "N" FPGA signal pins: How many possible PCB Signal to FPGA Pins assignment possibilities are there?
The Answer is: N Factorial (N!)
Let me prove it to you visually.
2 PCB Signals, 2 FPGA Signal Pins = (2 x 1 =) 2 Assignment Possibilities
3 PCB Signals, 3 FPGA Signal Pins = (3 x 2 x 1 =) 6 Assignment Possibilities
4 PCB Signals, 4 FPGA Signal Pins = (4 x 3 x 2 x 1 =) 24 Assignment Possibilities
O.K., at this point I trust you see the pattern and understand that creating further graphical representations is more confusing than illustrative. Let's switch over to just running some numbers:
10 PCB Signals, 10 FPGA Signal Pins = 3,628,800 Assignment Possibilities (That's right 3.6 Million Possibilities!!)
50 PCB Signals, 50 FPGA Signal Pins = 3.0E+64 (3 x 10 raised to the 64th power) Assignment Possibilities
100 PCB Signals, 100 FPGA Signal Pins = 9.3E+157 Assignment Possibilities
500 PCB Signals, 500 FPGA Signal Pins = 1.2E+1134 Assignment Possibilities
1000 PCB Signals, 1000 FPGA Signal Pins = 4.0E+2567 Assignment Possibilities
These are HUGE Numbers. To put these numbers in perspective consider that educated estimates put the count of all of the grains of sand on all of the beaches of the world at 7.5E+18 and on the order of 1E+22 to 1E+24 Stars in the Universe. Clearly, the number of possible PCB Signal to FPGA signal pin assignment possibilities DWARFS all of the grains of sand on all of the beaches of all of the planets revolving around all of the Stars in the Universe combined for modern FPGA devices with JUST 50 Signal Pins!!!
(Think about that for a moment.... assume 1 planet with "sand" per star and then assume 10 planets with "sand" per star and you have the same result..._staggering_. Your job is to find one grain of sand in all of the sand in the Universe )
Why are these numbers important? We know that if a design implements a good FPGA - PCB Interface design PCB trace lengths will be reduced in length by half:
Of course, trace length is directly proportional to signal trace propagation delay and inversely proportional to Max. Signal operating frequency: Reducing trace length by 50% supports a 100% increase in the Maximum Operating Frequency of a signal over that trace. The benefits of a good FPGA - PCB interface design are not limited to simply meeting timing constraints:
Reduce trace cross-over, reduce PCB Via count => Improve PCB reliability
Reduce routing congestion ==> possibly reduce PCB signal layer pairs and the associated PCB manufacturing costs
Gain a competitive advantage offering higher quality products at lower costs!!
The first aspect of this Blog was to simplify the FPGA-PCB interface design challenge to the physical assignment of PCB trace signals to FPGA signal pins. I immediately threw out pin banking rules, FPGA pin electrical configuration (I/O Standards) and all non-signal FPGA pins. In the real world these simplifying assumptions are not allowed. The million dollar question quickly becomes: How should a design team effectively navigate the staggering complexity of FPGA-PCB interface design to develop good interface designs that will result in a significant competitive advantage?
The Answer is Automation. In this case the automation is not simple and requires some form of Artificial Intelligence (I use A.I. loosely to incorporate all forms of knowledge based systems). As a design engineer I do NOT want to spend months studying each FPGA device's pin banking rules, I/O Standard options, special pin assignment rules, etc... I want all of the knowledge captured in the automation I use to help me create "good" FPGA - PCB Interface designs. The human brains ability to process massive complexity visually is awe inspiring hence I need the automation to present me with FPGA pin assignment knowledge in the graphical context of the PCB physical layout so that I may leverage the magic of the human bio-computer. Yes, I am lazy, I do not want to deal with each trace/signal of the FPGA-PCB interface individually, I want to be able to operate across buses and bundles (arbitrary collections of signals/traces)and quickly unravel these elements to minimize trace cross-over and trace length simultaneously
Fortunately, Mentor Graphics identified the complexity barrier to FPGA - PCB interface design years ago and has made a tremendous investment in knowledge based automation to allow you to navigate astronomical complexity and develop high quality FPGA - PCB interface designs much faster than you ever thought possible. Take a look at I/O Designer if you are challenged with leveraging FPGA devices in your PCB product design process and you want or need a substantial competitive advantage!
I do try to stay away from promoting specific product solutions through community Blogs. However, as you have read, I have studied the complexity of the FPGA - PCB interface design challenge in detail and the complexity is overwhelming. My objective writing Blogs is to give the design community insight into accelerating their total PCB design development process. With that goal in mind, I believe the complexity of the design challenge is so beguiling that many in the design community have chosen to simply ignore the opportunity. I wanted to take a moment and confirm the design complexity challenge and set everyone's mind at ease by letting you know there is a very easy to use intelligent system solution available to you that will deliver design acceleration to meet your ever decreasing time to market design windows. Please consider this a public service announcement!
I started using SPICE when the SPICE decks were created on card punch machines and submitted to far off main frame computers (I hope I am not alone 
)
I progressed through vector graphics terminals connected to main frames and when MicroSim first released PSPICE I thought heaven had opened those pearly gates.
I had the opportunity to use Linear Technology's LTspice and I like it. Can't beat the price (free). Easy to use, good convergence and good performance.
Obviously, now my PCB functional simulator of choice is HyperLynx Analog. HyperLynx Analog is actually more than one SPICE simulation engine delivering a built-in sim engine out of the box and scaling up to the IC proven Eldo sim engine to meet high capacity and extremely high performance requirements and then scaling to ADMS for true industry standard language baed mixed signal simulation leveraging both Eldo and ModelSim.
Independent of the specific simulation product we each use we all face common challenges such as locating the latest component model on the vendors web site, establishing component distribution profiles for statistical analysis. I've read and been engaged in conversations discussing the differences between the distribution profile between a 1% tolerance part and and 5% tolerance part based on manufacturing practices. (This topic would make an interesting Blog and discussion by itself)
Consequently, we (MGC-SDD) are inviting all PCB Designers using any SPICE simulator (any product including non-MGC SPICE based products) to use our Web Community to discuss and promote PCB functional simulation. The goal is to help our industry move out of the dark ages of prototyping as the iterative key to design development into the "modern" process of functional simulation. With PCB manufacturing costs rising due to High Density Interconnect, high speed requirements, Time to Market pressures and Six Sigma corporate initiatives I am stunned by how few engineers use PCB functional simulation.
Although SPICE is older than most EDA companies the adoption of PCB functional simulation tools as a standard PCB design practice is pathetic. Most of us who use SPICE tools are Geeks even to our geeky engineering family. Of course this is ridiculous, SPICE tools are no more complicated to use today then a web browser and from a business perspective the absence of functional validation through simulation is nothing less than a catastrophic loss of competitiveness. Independent of your awareness if you are using PCB functional simulation today you are an evangelical minister of core engineering discipline driving tomorrow's designs into reality today.
Based on my experience there is not a good central location for PCB functional simulation experts to come together and share their experiences independent of product centric sites. This is an open invitation to you to join me in creating a global community of PCB functional simulation expertise in an effort of mutual enlightenment.
Obviously, for this effort to be effective we must all avoid the inevitable product specific jousting matches. So while you should feel free to discuss the specific product you use (by name) and give product specific details and instructions I would appreciate your help in avoiding degrading any product. While I understand that marketing collateral is often helpful in instructing others on the basics I would suggest that instead of regurgitating marketing propaganda on this site you simply link off to the appropriate information.
As an example if you would like to know WHY I prefer HyperLynx Analog please visit the HyperLynx Analog Overview Presentation
Thanks, and lets have some fun helping others discover the productivity and creativity unleashed in our design processes through PCB functional simulation.
Caution: There is zero probability that I won't offend both DxDesigner AND Design Capture/DesignView users with this Blog. Of course, it is not my intention to offend anyone! However, I am sensitive to reality. 
The Old Days at Mentor Graphics
I worked for a little startup company based out of Pittsburgh,PA named Trimeter Technologies. Trimeter was using AI based pattern matching techniques to implement logic synthesis. Trimeter was primarily leveraging Mentor Graphics NetEd as the design capture package to feed their synthesis process. During this time I had grown to appreciate the simple elegance of NetEd coupled with the power of the embedded scripting language. I really liked NetEd.
Trimeter was acquired by Mentor Graphics. I moved from my beach apartment on the Balboa Peninsula in Southern California to Beaverton, Or. It was not too long after joining Mentor Graphics that I became involved with Design Architect. Talk about POWER; Design Architect ( in my opinion ) was and is the most powerful schematic capture package ever created for global enterprises. I recognized that the power of Design Architect came at a cost; you needed an infrastructure team to configure and maintain the environment. I have been an unabashed advocate of Design Architect to global enterprises for many, many years. The fact that I had developed into an D.A. expert cemented my enthusiasm permanently.
A Little Personal Background
After five glorious years working for Mentor Graphics focusing on ASIC & FPGA synthesis technologies I needed a change, both professionally as well as geographically (I was born and raised in Los Angeles, CA in the beach communities and while Portland, OR has the most beautiful summers I simply required more sunshine!). My search led me to Intergraph in Boulder, Co (massive amounts of sunshine!!) prior to it's transformation to VeriBest.
I had the honor of working with a core team at VeriBest to create DesignView. DesignView was an extension of Design Capture focusing on system level design while providing engineering flexibility to explore multiple design architectures "in situ" leveraging any combination of design creation techniques including industry standard Hardware Description Languages. In more than one sense Design View was the precursor to FPGA on Board system design before FPGAs had become mammoth PCB layout challenges due to the large BGA packages.(We were ahead of our time)
I loved Design Capture What a productivity product. I take No Credit for the awesome productivity enabled through Design Capture as that work had been on-going well before I joined the effort. Sure, I was able to suggest some tweaks here and there that I took great pride in. The underlying success of Design Capture was a seasoned engineering development team that was manifestly focused on customer satisfaction by providing "state of the art" performance, quality and functionality.
Obviously, I also took great pride in Design View as a consequence of many sleepless nights invested in the architecture of requirements to drive system design to new heights based on my hardware design experiences. Yet, DesignView turned out to be a community project with active participation from dedicated customers driving us forward. It was the synergy of the customers in harmony with the development organization mutual passion that led to the development of the most under rated product I have been associated with. I've never met an electrical engineer who has used DesignView who did not love it. Unfortunately, I can't claim that DesignView was ever a success in either financial terms or market adoption terms.
Moving forward past the MGC acquisition of VeriBest
While I loved Design Capture and Design View and had more than a little of my ego invested in the success of these products post aquistion there was no avoiding the fact that Design Capture was Not a global enterprise product. This realization was not a shock to me as Design Capture was never designed to be a global enterprise schematic capture tool. Design Capture was designed to be high productivity work group tool set scaling from small to medium sized installations.
It was around this time that I set off on a global adventure to interview every Design Architect customer and understand not only their infrastructure requirements but to document the customizations that each enterprise had implemented to fit their specific needs. This was a monumental task and until recently I suspect it was my opus.
Post Innovative Acquisition
Keep in mind that for years I had been a dedicated opponent of ViewLogic including a few years dedicated to the competitive replacement of ViewDraw determined to eradicate this antiquated pretender. To state that I did not like ViewDraw (DxDesigner) would be a complete denial of my emotional state: I hated DxDesigner. How much of my emotional state was firmly rooted in a technically based analysis is impossible to say in hind sight. I am able to tell you that the only aspect of DxDesigner that I did like was the scripting; I thought the use of exposed APIs and integration into Visual Basic was nothing short of magnificent. The real problem I had was that I simply could not make it work.
Fortunately, I was able to differentiate from the product and the people. The People who came to MGC from Innovative I really liked: smart, creative, hard working. It took me quite a while but after speaking to many DxDesigner users I came to the conclusion that not only was I biased but that my EDA tool use history had made the adoption of DxDesigner a near impossibility. The usage paradigm was the antithesis of everything I wanted in a schematic capture package.
I'll be the first to admit that it took me a lot longer than many to understand the potential of DxDesigner. The simple fact that eluded me was that DxDesigner was inherently extensible to enterprise deployment with some investment while maintaining it's native appeal to small and medium sized companies. Of course, this investment could not be realized over night but over the course of a few years I made incremental (and hopefully significant) contributions to the DxD evolution.
The New World: The launch of 2007 DxDesigner
DxDesigner does certainly enjoy some advantages. First, it operates across PCB layout systems, it operates with Expedition (including Xtreme), it operates with PADS Layout, Cadence's Allegro and Zuken's Board Designer.
Now, with the launch of 2007.1 DxDesigner has improved the PCB entire design flow radically. While others have been focused on the global enterprise requirements for DxD I have been working with the HyperLynx Analog team (DxD is the native environment for HLA) to greatly improve the out-of-the-box "ease of use" of DxD. Who wants to use a PCB Functional simulator where the first task is configure the schematic capture package? We solved these out-of-the-box issues with the release of Expedition Enterprise 2007.2 and recently with the PADS 2007.3 release.
Also, toiling in the background have been members of that aggressive VeriBest DesignView team to enhance DxD to remove Design Capture and DesignView user's objections to DxDesigner. Honestly, I really like using 2007.x DxDesigner. My objections have been satisfied. I find the new version of DxDesigner easy to work with requiring no out of the box configuration and yet entirely flexible. In my work I switch between Expedition and PADS flows routinely... often multiple times in the same day and I have no issues!
I am not going to tell you that every aspect of Design Capture/DesignView that I believed set it apart from the pack has been added to DxDesigner (That would not be truth). What I will tell you is that the evolution of DxDesigner into the "engineering cockpit" promoting rapid advances in design productivity is well underway and that the 2007.x release deserves your attention.
The Opportunity is NOW!
As I mentioned earlier in this Blog one of the key factors for technical success with DesignView was the user's community passionate commitment to a shared objective. If you are a Design Capture or DesignView user who finds some resonance with this Blog you are likely one of those champion souls who made Design Capture and DesignView the technical successes they are today. Now would be a very good time to add your voice to the DxDesigner evolution program!
DxDesigner: It's not your Dad's schematic capture package
One of the issues I always had with DxDesigner was that it lacked vision, it lacked scope. I don't design schematics, I design products. To accomplish product design I need a "cockpit", a hub, from which I may easily progress through not only design creation but detailed design analysis leading to physical design and ultimately manufacturing while easily managing all of the necessary design permutations. From my perspective the 2007.x DxDesigner has made this fundamental leap in it's evolution into the space of the system design cockpit. While I know the maturation process is not complete I realize that the significance of the advancements beckons the expertise and enthusiasm of like minded contributors in the user community to accelerate the metamorphosis. I'm not suggesting you "switch", I am suggesting you "take a look" and make your own determination. I am betting that if you are like me you will be impressed with what you see and will want to get on this rocket ship and help direct it to amazing horizons.
You Could Not Pay Me to Write This Blog
For those of you who are unfamiliar with me you may conclude that some marketing "dweeb" has crafted a "story" to promote corporate objectives. I suggest you take a look at some of the other Blogs I have written. In short order you will realize that I am an independent soul. If you review my Forbidden Blog you will understand that I present some interesting challenges to my management. It is impossible to promote Design Acceleration and not be an independent thinker challenging the comfort zones of the folks you hold closest! I'm telling you honestly no one asked me to write this Blog, no one suggested it would be a good topic, in fact, I suspect this Blog will make some in my chain of command a bit queasy.
Why would they be "queasy"? Because some of you will read this Blog and decide that we (MGC) are announcing the End Of Life of Design Capture and DesignView. It will not matter how many times I tell you that I have not received that message and that there is no intention of conveying that message, the Sales staff's phones will be ringing and answers will be demanded.
Frankly, I am the only person I know who makes the argument internally to End Of Life Design Capture and DesignView. I do not make the argument lightly or due to a lack of love for Design Capture and the installed base. I make the argument because I am a pragmatist and I know that focusing our resources on DxDesigner will only accelerate the maturation of the system design cockpit that we need to produce our next generation of designs. Not to worry, I am soundly rebuffed every time I make the argument. Design Capture and DesignView will not pass the way of the TRex until you (the users) make the determination for yourselves!.
Feel free to post your insights and opinions concerning the evolution of DxDesigner. {Even if you think these are the ramblings of an old dog barking at the moon}
PCB Functional simulation is only as good as the waveform viewing and analysis features that are standard with the product you use. In honor of the labor day holiday here in the U.S. I decided to do a little labor and demonstrate how easy waveform analysis should be. You might take this as a shameless marketing ploy However, I really do think that waveform analysis is so fundamental that I put together a few video tutorials on the topic.
This first is a simple summing amplifier composed of basic Op Amps and the goal was to determine when the reality of circuit design would limit the feasibility of the implementation I chose (a standard reason for simulating a design):
The second example uses the same summing amplifier design while focusing on categorizing the real errors using a statistical histogram plot:
The third tutorial is a bit more nostalgic using a Op Amp as an oscillator (a timing circuit generator) using parametric sweeps to allow me to easily find the correct component values when coupled with the ability to quickly calculate the operational frequency of the design (along with the duty cycle):
I have barely scratched the surface of the power of good waveform analysis we should all expect from our functional simulators! Accelerating design development demands excellence is waveform analysis. Having a fast simulator and then spending hours or days analyzing the results is NOT productivity!
Feel like sharing your opinion? click here to discuss
These tutorials were inspired by my previous Blog What are the characteristics of the Best SPICE simulator?