Now having read through your first article, and the next 3 followups about libraries,
I am somewhat left with the opinion you yourself miss the complexity of what your suggesting,
To set things clear, i am a KICAD user, and before that a PCB-Express one who got sick of the lock in,
If you want an eda that is beginner friendly, take a look at pcb-express, just as an example of a minimalism UI, it does bugger all, but what it does, it does well, and editing components is very simple, so once you get used to it, you tend to have dug yourself into a hard to escape hole,
People build up libraries over time because they tend to use the same parts, they are familiar with them, and if it needs software likely already have a code base to run it, if your like myself i add in the footprints and datasheets as i go. So next time around the part i like just works, its footprint matches, and all is right with the world, Now there is a limit on the flipside, without linking in something like a parametric search you would not want to amass too big of a library or it becomes hard to search, and even then many suppliers lie on there front page specs, (think current consumption under very specific conditions)
for schematic symbols everyone has there own preference, and in the case of something like a micro controller or FPGA, where nearly any pin has 5 functions, sometimes you want to break it up, sometimes your grouping by pin number, sometimes by function, how do you handle power pins? etc, how to best show decoupling, etc? (e.g. you may space pins on a schematic to match the capacitor symbol)
Next up design rule checking, yes differential pair DRC can be added, but in most cases that comes back to your schematic, this is where you define the connections, so unless you define in the schematic level what each output does (unlike the norm of just setting as inputs or passives so it doesn't bug you like so many do) it wont help, the rule checking can only enforce what it knows about
Its the time taken to build up these "models" of components that is the biggest drain when it comes to more complex EDA, having to read through the datasheet to figure out that nothing works unless some obscure pin has a pullup, or what the output limits are on an op amp at 2.5V when the datasheet only covers 5V to make the specs look better, and that tiny footnote in the middle of the thing that to you at the time think is something that should be bolded on the first page, This is where the challenges of EDA lie, and this is primarily schematic work, from then on you tweak to fit as you layout, I'm not saying layout can not also be challenging, but most of that already has developed DRC, setup a few netclasses define keepout areas for sensitive signals, and follow good design practices such as not breaking planes under fast signals,
Now the last parts come down to physics, this is a hard part of the learning curve, being able to visualize what happens to a signal at 0-10Ghz how noise gets into some nodes, how its radiated from others, truly a fine task for a field solver, but again this takes a lot of well defined starting conditions, such as output driver impedance's and such, to almost all EDA's its just a wire, possibly of controlled length and impedance, even calculating the heat of components takes a solid model of the components,
Home
Help
Search
Login
Register
