A lifetime designing PCBs: Improving autorouting & design software
Article By : Charles Pfeil
These PCB CAD adventures feature punched-card autorouting and a widget ruler.
This is the sixth article in a series about my career experiences in the PCB design industry. The previous articles discuss my lessons learned as a designer, running a service-bureau, and my initial involvement with a PCB design software company.
While at Automata, the environment of design and fabrication opened my eyes to a much better understanding of PCBs. But when Automated Systems (ASI) contacted me about the possibility of working on the development of a new eCAD system, the temptation was too great.
In 1987, I was hired as the new product development manager for PranceGT, which was a new version of PRANCE running on the MicroVAX. ASI developed PRANCE in the 1960s. It ran on an IBM mainframe computer and was famous for its auto placement and routing capabilities. I felt this was an opportunity to jump fully into the software side, and it turned out to be exactly that. My role was to manage the development team and provide specifications on the user interface and functionality.
PCB design has always been a creative adventure for me. Using design software is all about enabling me to solve problems with increased productivity. When defining new capabilities, I have always kept in mind the perspective of the designer, who, like me, wants to control the automation to obtain the desired results. Working on PranceGT was an opportunity to create a user interface that enabled the designer to be more productive, as opposed to just presenting a bunch of features with little user control.
Although I did spend much of my time writing specs, it was the time working with the developers and application engineers that I enjoyed most. I worked closely with Woody Woodward, Rick Mahoney, and the late Henry Bollinger on many aspects of PranceGT, enabling us to make improvements and fixes very quickly.
I have never been in favor of the over-the-wall approach between product management and engineering. Knowing that my specs are likely to have oversights and mistakes, I engaged with the developers to work through the details and come up with ever better ways to solve our problems. This also allowed the developers to participate in the creative process rather than forcing them to just blindly implement what they’d been given.
In my career, I have been fortunate to work with several of the best developers in interactive and automatic autorouting. Henry Bollinger was one of those. In the late 1960s, he and Dick Larson developed one of the first automated place and route systems. For routing, it had a rip-up and retry algorithm which included pushing traces and jumping over pins. It’s hard to imagine that this was possible at that time: the interface to the computer was punched cards! A joyful and brilliant person, Henry was always a step ahead when it came to solving problems.
When designing the user interface for X-Windows, I wrote most of the C code for the dialog boxes and widgets. We didn’t have the kind of apps that are available today. Most of the dialogs were black and white with gray highlights. I created a “Master Widget Ruler” to measure the number of pixels on the screen. It was drawn on the back of my business card. Although the other developers would laugh at me when using it, I still felt it enabled me to measure and align the widgets better.
When Cadence acquired ASI in April of 1990, some of the ASI team put my widget design tool in a frame and presented it to me as a memento.
Figure 1 The Master Widget Ruler
At Cadence, my role changed to be more along the lines of product management and support team assistance. Jim Behrens, the new VP of our team from ASI, called me into his office and told me something that I will never forget: “My most important role is to remove the obstacles preventing you from becoming successful.” His statement has been a big influence on my career, and I believe it is the best relationship for managers to have with their employees.
My direct manager was located at Cadence headquarters, so our communications were remote. One day I got a message from him asking me to send an email each morning describing what I expect to work on that day. Although I appreciate the need for status reports, I had never been asked to provide a morning report about what would occur during that day. In my role working with engineering and customers, new tasks appear each day, and often they become the most critical ones to address; earlier priorities get superseded.
My response to my boss the next morning was an email saying, “Today, I will work on whatever I think is best for Cadence.” I wasn’t trying to be a jerk – just honest. In minutes, I got a phone call from him. The good news was that my manager listened to my explanation, understood my perspective, and trusted me. In my role, it was more practical to report after the fact rather than to try and predict the future.
Shortly after Cadence asked me to relocate to the Boston area to have broader responsibility with product management, I received a call from Jim Thomson at Intergraph, who offered me a job. I’d met him a few years earlier, when Intergraph was evaluating PranceGT as a routing solution for its PCB design system. After Cadence had acquired ASI, Intergraph bought DAZIX to gain access to their Cadnetix routing technology. And so it was that I started with Intergraph in 1992.
Related articles:
- A lifetime designing PCBs: Early design adventures
- A lifetime designing PCBs: Job shopper to service bureau
- A lifetime designing PCBs: Switching to eCAD
- A lifetime designing PCBs: From design to software
- A lifetime designing PCBs: Merging design and fabrication
— Charles Pfeil is a Senior Product Manager at Altium, working on definition of their products with a primary focus on routing tools.