Issue link: https://iconnect007.uberflip.com/i/340751
58 The PCB Magazine • July 2014 be some serious downsides to the adoption of ODB++. This is not because ODB++ is a particularly bad format: It is not. The point is that the adoption of ODB++ includes the adoption of a new image format, and image formats are notoriously hard to implement. Much has been written about just how complicated geometric software is and how much effort it takes to get it right, not to mention the years it takes to debug. So the implication that taking on this new image format is simple and low risk is at the very least misleading. Precisely because our industry's practitioners are not morons, they know this, so are reluctant to adopt a new format. They know very well how complex ODB++ is, and that it will give rise to many more problems, for many years. The reality is that Gerber works very well for transferring images. In fact, there's nothing better. Gerber X2 The most interesting point made by Coates is that Gerber files contain "no information about how the PCB layers stack up." This was a valid objection in the past, but it is no longer true, as the latest revision, Gerber X2, now contains layer stackup information. At the heart of X2 is the use of attributes. These are akin to labels which provide information that are associated with image files, or features within them. The beauty of using attributes is that they are already familiar to CAM professionals and software developers, and they sit naturally with the current capabilities of CAD and CAM systems. X2 extends the current Gerber specifiation with a series of standard attributes that are most important for efficient CAD-to-CAM communications, such as the function of each layer, whether a pad is a via or an SMD pad, and which are the component drill holes. As rather grandly stated elsewhere, X2 adds intelligence to the Gerber format. Software supporting X2 will read the whole Gerber archive automatically, with all layers in place, while identifying the function of each object. Easy to adopt and to implement, X2 is upwardly compatible with the previous Gerber version. Altium, global leader in Smart System Design Automation, has been quick to recognise the value of X2 and will support it in an upcoming version of Altium Designer. By Q4 2014, Graphicode's widely-used and highly- respected GC-Prevue viewer will also support X2. X2 maintains the trademark simplicity for which Gerber has always been known and used, and gives designers and engineers a standardized procedure that will require very little to change in their working practices—certainly none that would require approval, testing and all the rest. Equally important, this new revision does not disrupt existing workflows. If the software does not support the new capabilities, the old workflow continues to operate. Nobody is forced to buy anything. So this will be a very gentle, low cost improvement indeed, but the effects will be nothing short of revolutionary. Coates omitted to mention this latest development in the Gerber format, one of the most important developments in CAD-to-CAM automation today, given that it concerns the industry's de-facto standard format. Neither did he mention the alternative IPC-2581. Had he done so, his arguments for ODB++ might have been less compelling of course, but these omissions in an article titled Smart Data Formats Automate CAD/CAM lead to serious doubts about its objectivity. Coates also added a diagram to the article comparing Gerber to ODB++ input in CAM. This compared a badly implemented Gerber with a well implemented ODB++. I have taken the liberty of adding a proper X2 Gerber to the schematic. The result, given in Figure 1, shows that if ODB++ is a smart format, Gerber X2 is a very smart one. Conclusion When CAD-to-CAM data sets use properly implemented Gerber archives, plus correct IPC-356-A files, problems in data transfer are rare. Where a problem or bug appears, the easiest, fastest and most economical solution is to fix it. This is because issues are not down to the format itself, but more likely due to its implementation in CAD software, and they are simple to resolve. The very worst solution would be to replace Gerber with the far more complex ODB++ format, because implementing a new format is never simple, quick, and/or risk free, GERBER—THE SMARTEST WAY FORWARD continues