I am practising PCB CAD for 25 years now -say- 10% of my engineering time. The rest is pure research, schematic entry, parts resource, etc. So you can imagine that with this 10% only I want to design high end boards as well. Everybody can imagine that "management tools" and such only is hindering me by doing my 10% job effectively.
To my opinion CAD has a total stand-still for at least 10 years if you look at the *real* thing truly critical. Nothing has changed. I find CAD companies only money makers, working with 100x100 pixel per inch ugly symbols, limiting you all the way to make a proper graphical design. Anyway, I find that the old P-CAD system was the only system that lets you design freely. It could not do anything, but what it did was super fine and intuitive to use. Just schematics and board design with a simple netlist in between, and an easy and intuitive way to create component libraries. Wonderful. Since then, only crap new options arrived inside a sluggish unworkable enviroment: The Chinese bought the old owners out and completely f*.. it up. (forgive me, have no other word for this)
This day -for quick designs- I still use P-CAD. The finest SMD designs are worked out with it, operating even over 3GHz... Indeed because I learned that skill - indeed the hard way.
So for all CAD companies out there I would say: Get back to your roots and see what real needs there are to get a proper PCB out these days. Forget about the crap-options, make it graphically nice, fast and intuitive for your users. I think it is artwork: So make the programm in such a way that it INVITES you to create nice designs. Adopt to new technologies -for example- to route differential lines WITH CURVES, not 45deg pairs. Add length matching, without that stupid spreadsheet. It *can* be done intuitive by selecting a bunch of lines in the schematics and then doing so in the layout. It is a shame- for example- in Cadence I still cannot appoint a differential pair into the schematics. I HAVE to use a stupid spread sheet to do that. It is very very inproductive, time consuming and still dangerous and prone to make errors very easy. Make the schematic the holy grail again so that finally it can be recognised again as a proper schematic. Get rid of all those cubes with labels on them.
In 2003 to 2005 we did a couple of reference designs for Xilinx with P-CAD. (UG037 and UG172, for example) Damned fast boards, 1,2Gbit/sec x 64 (!) completely created with P-CAD, without board verification tools, done about 10 years ago. LVDS was pretty new back then.
I could go on for ever, but that's too much aimeed to one example, and of topic I believe... but I made my point ;-) Tell us, how do you feel about this CAD topic?
Totally agree, Caleb. Over the years, I've noticed that those who indulge in this "surfing," as you call it, are also those who never seem capable of getting on with their tasks. It becomes an excuse for delaying.
The producers, on the other hand, have an uncanny ability for getting stuff done in spite of this sort of handicap.
Hey @Caleb, your post here reminds me of a quote I often shared with my students in the past, from Alexander Pope: "A little learning is a dangerous thing; drink deep, or taste not the Pierian spring: there shallow draughts intoxicate the brain, and drinking largely sobers us again." Fundamental human flaw?
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...