-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueBox Build
One trend is to add box build and final assembly to your product offering. In this issue, we explore the opportunities and risks of adding system assembly to your service portfolio.
IPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
Boost Your Sales
Every part of your business can be evaluated as a process, including your sales funnel. Optimizing your selling process requires a coordinated effort between marketing and sales. In this issue, industry experts in marketing and sales offer their best advice on how to boost your sales efforts.
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Electronic Interconnection Files from SiliconPipe - No. 3
September 12, 2005 |Estimated reading time: 4 minutes
At the last Intel Developer Forum in <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" />San Francisco, CEO Paul Otellini, announced that Intel would now be applying a new metric to their products passed on performance per watt as the key metric for their products[i]. Given the weight of a company the size and influence of Intel, it is likely that others will be following their lead. It is worth reflecting on what may have been the source of this new direction and its impact, but first a bit of background is required.
Chip speeds have risen significantly in recent years, however the benefit has not been fully realized. Signal transmission speeds have fully entered the gigahertz range and considerable attention is now being paid to signal integrity. The primary objective of most research has consistently been to extract the dormant power of semiconductors. (e.g current processors operate at nearly 4 GHz, but mainstream PCB performance is nominally limited to ~ 800 MHz). (Exotic transistor based solutions with pre or post emphasis of the signal can help but there is an energy bill to pay for doing so.) The actual limits of the PCB are set by a combination of design and manufacturing issues and addressing these issues at a practical level is a complex challenge, especially when the demands of cost and performance collide.
The most intimidating challenge to signal integrity is manufacturing. Natural variation related to traditional manufacturing and assembly processes is a formidable foe for even small variations in material properties and dimensions or of circuit features can degrade a signal, thus signal integrity now dominates much discussion. So what is signal integrity really? Perhaps signal integrity can be most easily summarized as the ease or degree to which a transmitted digital signal can be deciphered as to whether it represented a 1 or a 0 when sent, a simple definition perhaps but not a simple task. Ascertaining signal integrity using current design and manufacture methods often requires a highly skilled engineer or scientist using very specialized software. Using these tools a good approximation of a solution can be generated; however current circuit materials, design approaches, and manufacturing processes do not generate ideal results. Competition for routing channels on PCBs having many signals of various strengths and proximity traveling through them translates to lots of coupling and that creates noise and degrades the signal quality. Solve the noise and cross talk problem and there are still a host of other vexing nits in the system. Manufacturing artifacts such as inconsistencies in dielectric properties and trace width, variation in circuit spacing, even copper roughening treatments, all impact signal integrity to one degree or another and at higher speeds, the additive effect of these once minor issues can and will degrade signal integrity by introducing changes in impedance and capacitance. There are also the ever present concerns of resistance, dielectric loss, conductor loss, signal skew and the electronic stubs associated with plated vias. With each element taking its toll, it is very difficult to predict and design for reliable performance. One way to cut the knot is to segregate the high-speed interconnections from the low-speed power and ground, routing them separately in materials that can be more easily controlled, like microstrip cables. These can then be routed from the top of one package to the top of another. No special skills are required to analyze the signal because simple and ubiquitous tools easily predict the results. It's a very compelling idea that cuts to the heart of the signal integrity challenge and it has already been demonstrated in the laboratory. The prototype provided sufficient signal integrity to transmit a signal nearly 3X the manufacturer's specified distance, through two connectors, with a 60% margin and it did so using less than 2% of the anticipated power.
<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />
With the now growing concern over energy use in electronic products as expressed by Ortellini, this largely untapped benefit of a clean signal channel and its inherent signal integrity seems too compelling to ignore. Power reduction is a far-reaching feature that every electronic product can benefit, from handheld devices to supercomputers. A clean, well-designed signal channel even allows for the reconsideration of how ICs are designed because the I/O driver complexity can be reduced. Additional power savings are anticipated here as well, because when I/O voltage is lowered, core voltage and thus total power can be reduced as well. Summing it all up, the power-reduction potential of improved signal integrity enabled by clean channel design should logically see increased investigation as a part of an over all solution to the energy challenge. The logic is based on a simple observation and an equally simple question for which the answer should be obvious.
Thus, as electronic product developers are presently struggling to get the heat out of the system, a logical question is: "If the heat generation can be avoided by using a clean channel for signal transmission, what need is there for exotic and/or expensive and often energy intensive thermal management solutions?" And the answer is...
Joseph Fjelstad, a founder of Silicon Pipe Inc., is an international authority and innovator in the field of electronic interconnection and packaging technologies with more than 150 U.S. patents issued or pending. He is the author of "An Engineer's Guide to Flexible Circuits" and author, co-author or editor of several other books including the most recent Chip Scale Packaging for Modern Electronics. He has also authored numerous technical papers and articles. He frequently presents seminars on PCB, flex circuit and chip scale packaging technologies at industry conferences. You may contact him at 408-973-1744 x203, or by e-mail at JosephFjelstad@aol.com or J_Fjelstad@sipipe.com.
[i] Demmin, Jeffrey C, "Thoughts on the News -The Week in Packaging (Part 1)" Advanced Packaging , 29 August 2005