Quality Camparison Allegro Vs. PADS

[Carl Schattke]

What differences in quality are there between PADS and Allegro design rule checking and what gaps do either tool present

I am interested in doing an indepth study of this and would like to collaborate with any users of either tool.

I want to evauluate if there is anything missing from the checks of one vs. the other.

What kinds of errors can you make in PADS or Allegro that the DRC's will not check?

Are there rules that you might not check due to difficulty in setting them up?

Let's be fair and only compare most recent versions.

[padmaster]

As far as I know, Allegro has a full compliment of drc checks and constraint settings. That includes individual layer settings and individual pin type settings for individual layers. The only down side I have noticed to having many constraint settings is that it boggs down the system. But then, I would imagine that would be common between software packages.

Haven't used PADS enough to comment on it.

[Carl Schattke]

Well the results are in.
I put a spreadsheet list of over 150 things wanted in an ideal design system.
Then I rated both Concept/Allegro/APD vs Viewlogic/PowerPCB
on a scale of 0-3 0=not capable, 1=requires workaround, 2=completes task, 3=completes task with extraordinary ease or capability
I then compiled a list of the most important features with regards to 100% checking, Complete schematic integration, capture of Design rules both physical and electrical, Library Revision control for templates and components, automation, productivity and ease of use and assign a value to each feature with regard to each catagory and I rank the catagories on ascale of 1-5.
I then multiply the capability# x the feature score# x the value# and then add them all up.

My matrix can compare any two CAD systems and determine the resulting value.

My score for Cadence was 3661 and my score for PowerPCB was 4345. This may be surprising to some but not to those of you who have used both systems to do your work.

Pads scored better the biggest difference in ease of use productivity 33%, Schematic control of the design 35%, and revision control of library and templates 288% since the Cadence suite really has nothing to compare with the DX designer tool.
In the important area of 100% checking Pads was 17% better.

Cadence scored slightly higher in integration autorouting support 4%

I did not rate customer service, availability of pre-built libraries, integration with SI tools, availability of service bureaus, cost of ownership, however these could be plugged into the matrix as well.

I am also interested in getting the data for other design systems that are being used and plugging in the numbers for them to see how they stack up.

In general Cadence has some work to do adding features that are all ready available in Pads.

[YeeLin]

Do you mind sharing the comparison that you mention in a spreadsheet? Or where can I download from this site?

[Carl Schattke]

Allegro lacks an acute angle checker in the DRC mode so you won't find your acid traps until the design gets a check from a CAM system like Valor.
It also does not check for broken plane connections, so you could have an open in a dense thermal pin area.

To YeeLin, it's been awhile since I copared Allego to Pads, I would be willing to go over the data with you, but the 150+ questions I asked will remain mine.
Keep in mind that thiis was geared for the typeof designs our group performs and your priorities may differ.

[Wildwpe]

Allegro lacks an acute angle checker in the DRC mode so you won't find your acid traps until the design gets a check from a CAM system like Valor.
It also does not check for broken plane connections, so you could have an open in a dense thermal pin area.

To YeeLin, it's been awhile since I copared Allego to Pads, I would be willing to go over the data with you, but the 150+ questions I asked will remain mine.
Keep in mind that thiis was geared for the typeof designs our group performs and your priorities may differ.

ISBN 0-9741936-0-7 "Right The First Time"
Author Lee W. Ritchey
Copyright Date:May 2003
Page 92 -
Question "What about acid traps formed by right angles bends?"
Answer "This hasn't been a problem in the history of PCB manufacture. It is a puzzle where this rule came from, but it is not valid"

So, maybe that's why the tool doesn't flag angles as a DRC? I would love to know if this gentleman is wrong in this assumption. He also states that right angle bends on routes do not cause EMI problems. He does run some interesting experiments that are documented in this text.

[Carl Schattke]

Wildwpe,
You opened the can of worms. Yes, acid traps is the common name for acute angle checking. There is a higher failure rate an artifact of processing boards years ago. Having worked in a fabrication facility about 30 years ago, and inspecting thousands of boards I can assure you that acid traps are a location of higher errors than other areas and they did affect yield. Of course you can manufacture boards with just about any pattern, but why create a potential weak point. PCB manufacturing equipment and plating process equipment has come a long way in those 30 years. Aesthetically it's hard top argue with the elimination of acute angles, but I don't trash bin the boards if one slips through the checking process. Lee's contention is that eliminating them is a waste of time. He can design and build boards that way, I'll choose not to. I want my software to point them out.

Right angle corners are a tiny source of impedance discontinuity. So are the bumps on a trace under a microscope. Why don't these matter? Well at common PCB frequencies it's to small a difference to matter.
Lets look at a situation where you have a long trombone pattern for length tuning. It's better to use the space with two u-turns rather than an alligator saw tooth pattern with hundreds of U-turns. Why, because the cumulative effect of those tiny impedance discontinuites can get to the measurable point at high frequencies. I agree with Lee that in general right angle traces will not be a source of EMI radiation since the area is typically well short of the length required to become an antenna.