Issue link: https://iconnect007.uberflip.com/i/1526666
SEPTEMBER 2024 I PCB007 MAGAZINE 61 the current enterprise pitch. It will be lowered in the next generation to pitches like 0.6, 0.7. Being able to maintain similar or better power distribution with lower pitch helps in cost and performance. e main issue to lowering the drill to metal (D2M) in subtractive power layers is the potential to create the failure conductive cathodic filament (CAF) through the glass bundles. It happens when two metals and a bias are too close together and some separa- tion occurs (infrequently) on the glass fibers, or they have cracks or hollow areas. Aer ther- mal cycling, sometimes the path is created, and once you create that path, if there's a bias between these two metals, eventually they can create a short. is is the number one prob- lem in the industry for D2M, which is drill to whatever the metal feature is. We must elim- inate the potential glass path to bring these metals closer to the vias. Otherwise, you can't lower your pitch, or you must go to another interconnect type altogether. is wouldn't be such an issue for very few failures in the field, but there is no screening method to tell when this will happen in other parts of the lot once it's happened on a single board. Suppliers hate the word recall. HDI layers don't have this CAF impact as significantly, but they still have it in the sub- panel vias. Most HDI spacings are between 55 and 100 µm. e D2M values change very lit- tle between most advanced suppliers, between 0.2 mm and 0.25 mm depending on power and application. At Wus, we mechanically remove the spacing between the drilled hole and the nearest antipad along with the glass path. en we back-fill it with liquid prepreg, a dielec- tric printable (with a very high vacuum coat- ing machine) material with the same FTIR match as the laminate but with no glass (but higher filler for CTE advantages). So, now you can have the D2M much closer, approximately 0.13 mm. So, if you need to go to 0.6 mm pitch, you can do it with your signal processing, and your power. Matties: In looking at the markets, it is realis- tic that the West is resigned to servicing DoD and prototype medical. Maybe we won't see much change from what we already have. Yes, but then performance should be the issue. e difference in performance is becoming dramatic around what's capable in Asia and the West. When commercial products significantly outperform current military products, this will be a big technical challenge. Some are trying to address this through government legisla- tion and funding. But in the bigger picture, it doesn't make a lot of sense to me. Focus on the real separations of technology, dies, and so- ware. A single X-ray picture of any PCB tech- nology with a cross-section gives away all the technology of that board, independently of where it was built. Matties: The U.S. has many brownfield sites. Maybe we will continue to see more invest- ment in captive facilities, but I would not expect to see job shops chasing big commercial work. Okay, but then let's be honest as we look at our technology and individual company roadmaps Joe Dickson