Reading time ( words)
Figure 3: Gerber photoplotter and its controller. (Source: The Smithsonian)
For 20 years, designers at OEMs created their 1:1 photoplots on Gerber vector photoplotters and sent them as the master tool to their bare board supplier. With the advent of laser photoplotters in the mid-1980s, the PCB fabricators added in-house photoplotting as a service and began asking for data instead of film.
What data format do you think they used? Correct—the Gerber format. After all, designers were already able to produce it, and the new laser photoplotters accepted it. So, Gerber became a data exchange format between designers and fabricators along with Excellon drill files.
In 1991, in an effort to incorporate more aspects of the imaging process into the automated systems, two people—Ed O’Reilly of Gerber Scientific and Randy Allen of AT&T in Richmond, Virginia—collaborated in the development of an extended version of the Gerber format called 274X. The 274X format allowed the inclusion of mass parameters such as layer polarity and aperture definitions within the plot file, thus eliminating what used to be known as the Gerber wheel file. It is interesting to note Gerber 274X never became an EIA standard. As the Gerber format already had become the de-facto standard, the industry simply accepted this new format as an improvement over the old one. Gerber 274X is still today the most widely used data exchange format in our industry.
Now, let’s try to apply Gerber’s innovative way of looking at things. In the 1960s, we can speculate that he probably asked himself, “What is it that electronics companies are trying to achieve? Aren’t we all trying to get our products to market as quickly as possible at the lowest possible cost and at a defined quality level?” Today, he might ask, “Is using a format derived from 55-year old technology the best way to remain competitive in the marketplace?” Can you name a single other data format or its derivative that is still in widespread use 55 years after its introduction? I couldn’t think of one either.
That says a lot for the Gerber format’s usefulness. But with every PCB design you send to your fabrication and assembly suppliers in the form of legacy CAM data such as Gerber, you are putting your company’s objectives at risk. You are not simply sending Gerber files; you are also sending drill files, netlist files, test files, BOM data, centroid data, documentation files, drawings, and instructions. You can’t assume that all of the files sent in your data delivery package are compatible. Many times, they don’t even come from the same system. Thus, in an attempt to minimize the risk of communicating conflicting information, your team is spending resources generating and validating these various files.
Even more significantly, you are asking your manufacturing partners to reverse engineer your PCB model before they can perform their value-added services. Not only is that inefficient, but it is also susceptible to error. How often do you as a designer hear your job is on hold from the fabricator or contract manufacturer because of a question they have about the data package? Why do we continue to limit our competitiveness as an industry by relying on old methodology? To be competitive in today’s market, the entire supply chain needs to commit to efficiency—beginning with design.
The issue of material cost has been milked as much as it can. Fabricators and contract manufacturers are asked to make perfect PCBs that are each a custom product but at a commodity price; they then get dumped with dumb data sets and are told, “Deal with it.” What do you think the results would be if, instead, they were treated as a critical supply chain partner and asked what could be done to improve the product’s quality and costs and accelerate the time to market? What do you suppose they would say? Probably not to keep doing things the way they have always been done.
The Next Generation
Since 1995, the ODB++ data exchange format has offered the PCB industry an alternative that enables the creation of intelligent product models (and to be fair, it’s now nearly a quarter century old). Like its predecessor, ODB++ was not originally developed with the intent of becoming a data exchange standard. Rather, it was the database format for Valor’s Genesis CAM software used by the PCB fabrication industry. As OEMs and contract manufacturers also began adopting the Valor DFM solutions, the forwarding-thinking electronics companies of the time saw an opportunity to strengthen and improve their supply-chain partnerships using ODB++.
ODB++ caught on because it was comprehensive. It allowed building a single product model that included all the necessary fabrication, assembly, and test data needed to manufacture a PCB. This product model contains layer definitions, polarity, netlist, component layers, and test points all clearly and consistently attributed so that all users of the data could benefit from the information as the product moved from design to fabrication, assembly, test, and manufacturing. It replaces several formats using one deliverable. The specification has evolved and remains free to everyone as does the ODB++ Viewer.
Figure 4: Benefits of using a single product model containing all the necessary data to build a PCB.
In a recent global study conducted by Mentor, 48.7% of all companies responding said they use ODB++ for at least one of their functions (fabrication, assembly, test, DFM, or viewing). Certainly, the industry gravitates to solutions that are practical and proven.
What Will the Future Look Like?
As electronics technology and manufacturing move deeper into the digital age, what will the machines, tools, and data look like? How will it change, and what will we need to do to adapt? Gerber would probably ask new questions at this juncture, and we need to be asking them ourselves to continue to compete and innovate as the world changes.
What can you do to improve your company’s efficiencies and quality? Ask the engineering department at your fabricators and contract manufacturers which data format they prefer and why. What information needs to be communicated to suppliers? What is the best way to communicate that information? What form should that content take so that it can be handled automatically by the recipient? If anything is still being communicated via a document or drawing, why is that so, and can it be communicated via data instead? What are the barriers to digitalization?
The truth is—as advanced as the electronics industry is—some companies are still holding on to the legacy Gerber format. It served its time well, but we currently have a more efficient data exchange format to take us to the next level. At 48.7% ODB++ usage, critical mass has been reached—the tipping point achieved. Now, it’s time for the next level.
Gerber listened to his customer. Rather than just trying to automate a single step in the process, he addressed RCA’s end goal. When laser plotters were invented in the 1980s, the machine makers continued to use the Gerber format, and it became the de facto data-exchange format. As we celebrate this achievement, let’s look at where we can go today, after half a century.
We are all still trying to get our products to market as quickly as possible at the lowest possible cost and a defined quality level. But our world of technology has changed drastically in the last 20 years—especially in the last 10—and continues to be in flux in an increasingly uncertain global market. Joseph Gerber would probably agree that with today’s rapidly changing market demands and sophisticated technology, an evolving comprehensive digitalized format is better for continuing innovation.
Patrick McGoff is market development manager for Mentor, a Siemens business.