February 05, 2007
Retrospective and Perspective on Printed Circuits – Happy Holden
Please note that contributed articles, blog entries, and comments posted on EDACafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
On the technology front would you compare five years ago to today in the printed circuit arena including from a design perspective.
Design continues to plow ahead. Fundamentally we are all driven by semiconductor technology. If you look at the semiconductor roadmaps, they have made solid progress in geometry, in increasing wafer size, in improving performance and so on. Today a 5 cent chip powers just about everything. With smaller and smaller geometries down to 45 nm the whole concept of die shrinking allows that used to be an expensive chip 5 years ago to be an inexpensive chip today. That will continue to drive the industry for the foreseeable future. We were always concerned about the 5 cent chip in the $5 package. Printed circuit technology has not kept pace with semiconductors, fundamentally because there is no finance scheme, no visibility of this at the university level. It is kind of forgotten because semiconductors being such a dominant technology that you kind of forget that the chip has to go somewhere, it has to go on a printed circuit board. We were doing pretty well powered by OEMs like HP, IBM and AT&T until the OEMs started to dismantle their internal manufacturing and began to outsource. Our development model fell away in the mid eighties and has not been replaced by anything. We are not driven by the military or the Department of Defense anymore. After 2000 the telecommunication market dried up. For the last five years printed circuits from the North American perspective has been flat or declining. If you go to Asia just the opposite has happened. There is enormous growth in printed circuits there. There is continual dominance of the technology by the Japanese where the Japanese manufacturing and the Kiretsus are still intact. They did not take apart their manufacturing and everything else like HP, IBM and AT&T. They are still going great guns with technological innovation and their own internal manufacturing for printed circuits. North America has disassembled its model. Nothing has appeared to put together the pieces it takes to handle advanced technologies. I think that there is one reason why Henry Potts got a hold of me, a printed circuit
technologist, to ask some of these fundamental questions.
Henry’s vision for Mentor Graphics is to support advanced printed circuit technology with design tools. He had a lot of questions for me as a consultant about what were those advanced technologies and where were they going. Ever since 1984 when I moved to Hong Kong working for HP, a lot of my printed circuit background has been in the growth and development of the Asian printed circuit manufacturing. When I first went to Taiwan they were barely making single sided or double sided printed circuit boards. Today in China, Taiwan, Korea the printed circuit capability is enormous in size and sophistication.
Beyond the migration of printed circuit manufacturing to lower cost geographies in Asia, what has happened on the technology front?
The very clear one where the Japanese have the lead is the whole focus on miniaturization, miniaturizing sophisticated products: mobile phones, music players, cameras so they became consumer products. This enormous sophistication has driven the printed circuit technology as well. Obviously the chips were already small but now the package and everything else has to be small. This is where the Japanese clearly provided leadership on miniaturization which drove printed circuit technology and led to much finer laser drilled micro vias, embedded passives and things like that. North America and Europe were fundamentally military and industrial products which did not have a size restriction because they were not consumer products that were going to be handheld. Many times their products were very hot. They had huge heat sinks so that were not portable. There was less of a need for miniaturization and a very great need for reliability and long life. But clearly the driving force in technology has been miniaturization along with the reduction in cost to help these sophisticated electronic products charge the engine of growth where everybody buys these things. I keep track of how many cell phones are manufactured each year and how many cell phones are thrown away. The value of the metal in thrown away cell phones is huge. It flabbergasts me that we can make 80 million cell
phones a year where each of them contains two micro via printed circuit boards. For the foreseeable future there will be continued growth beyond a billion devices a year. It is absolutely amazing. 80% of those are manufactured in Asia. Clearly, this consumer product and miniaturization are the real engines that drive innovation now. In fact I am trying to help the military catch up with some of this innovation. Thirty years ago the military used to drive innovation and technology. But now it is clearly consumer products and coming along nicely is automotive.
What are the benefits and ramifications of micro vias and embedded passives other than miniaturization?
If you look at it in terms similar to the semiconductor, how many gates you can get per square centimeter. The whole point in providing more and more functionality is to get people to buy for example new cell pines by adding cameras, colored displays and things like that.
I still carry around an older Motorola mobile phone that only makes phone calls. But I am certainly in the minority. Everybody else has mobile phones that take pictures and videos, go on the Internet, watch TV and things like that. But you still have to make these things handheld. That means you have to pack an awful lot of stuff into a small amount of space. And yet the price is not elastic. It has got to stay within an affordable price range. The cost per layer and the number of interconnects, what I call the average connections per square inch, has plummeted in the last 5 years along with the cost of semiconductors such that we can put all these integrated circuits into a small area
and still maintain a small price. Without it you are not going to get a cell phone with pictures, with TV, that is lightweight and that fits into your pocket. Where you can put it in your purse and use Bluetooth headset attached to your ear, where you can send email anywhere. We have not seen the end of this. Semiconductors are still driving the train. The semiconductors guys are still plowing ahead at reducing the size of their transistors and increasing the frequency at which they operate. This makes it easier to make everything wireless. That engine is going to drive things for many years to come.
What is it about microvias, embedded passives, flex and rigid flex that helps this along? What do they do that their predecessors did not do?
Just like semiconductors, they do the same things they have always done. They do it smaller and at a lower cost. A typical laser drill can drill 400 vias a second. A mechanical drill will be lucky if they can do a couple of hundred a minute. We have lasers that do up to 21,000 vias a second. Things like that. It is performance divided by price. Newer technology essentially lowers the cost per via and that lowers the overall cost of the printed circuit. The same is true for making smaller and smaller surface mount parts. As the surface mount parts get smaller to support miniaturization, they actually cost less for a resistor or a capacitor. A very small flip chip organic package for an
integrated circuit costs 60 cents where 30 years ago we had to pay $25 for a ceramic package for one integrated circuit. Now we are paying a few pennies for a plastic packaged chip. It has all evolved on this theme of miniaturization not just higher performance but also lower cost without sacrificing reliability for other aspects. That is why I guess electronics is the biggest industry in the entire world, if you take the total value of everything involved in electronics.
Over the next 3 to 5 years what are some of the promising technologies related to printed circuits?
That is one of the reason I came to Mentor. Mentor is focused on what are the advanced design tools to support these new products. One of my jobs is to advise the various product managers on what I see as a continuing evolution of miniaturization of what we have now. We simply do not have enough room, so we are going to three-dimensional packaging. Everything is being stacked. You have packages and wafers being stacked. You now have the components being stacked. One place for the components is also inside the printed circuit board with embedded passives. But there is also the opportunity for embedded actives. Another way to make three-dimensional packages is with flex and flex-rigid. You can assemble it flat and then fold it up into a 3D package to gain a better volumetric space saving. In addition to all that, if you are going to start things like TV on cell phones, there is enormous need for increasing bandwidth which is why the telecomm companies are starting to come back really strong. The pipes that are going to carry all of this Internet, TV and so forth have to get bigger. One way the pipes get bigger is by improving optical technology. We are on the cusp of a whole new generation of printed interconnect in that we have now figured out how to print optical waveguides for very low cost instead of having to use a piece of glass fiber, one piece of glass fiber at a time. This is like 50 years ago when somebody came up with the idea of putting all these copper wires on a flat substrate, a printed wiring board rather that wiring up a chassis using individual wires between components. We are now seeing the advent of the printed optical waveguide. That is probably the most exciting thing because if we can print these optical waveguides on a planar surface we can interconnect optical devices very inexpensively
rather than having to wire them together a glass fiber at a time.
You can find the full EDACafe event calendar here.
To read more news, click here.
-- Jack Horgan, EDACafe.com Contributing Editor.