AltiumLive Munich: Day 1 Keynotes
The weather forecast was wrong! Despite my apprehension and winter clothes, there was very little snow at the Hilton Munich Airport. It could have been any season of the year inside the splendid convention facility, which was also the venue for the second European AltiumLive design summit.
AltiumLive brought together a family of over 220 electronics engineers and designers eager to learn from top industry experts and applications specialists who were equally eager to share their knowledge and experience freely. “Learn, connect, and get inspired,” was a fitting tag line for a superbly structured event.
Lawrence Romine: Welcome
“It all starts right here, right now,” said Lawrence Romine, vice president of corporate marketing at Altium, as he welcomed a packed house of attendees to the first keynote session, commending Altium users for their continuing dialogue and feedback. “Someone making tools for engineers is actually interested in talking to engineers and PCB designers, and it’s clear that you guys are interested in talking to us.”
He remarked that a new licence for Altium Designer was sold every hour of every day and that the AltiumLive event reflected not only the enthusiasm of the design community for Altium’s tools but even more importantly, a celebration of the products created using them.
The audience got a privileged preview of a cloud-based, real-time collaboration platform shortly to be released to the market as Romine was joined by Leigh Gawne, head of Ciiva at Altium and the architect behind its development. They demonstrated Altium 365, which was billed as “the connected future of PCB design” and had been announced for the first time at AltiumLive San Diego. It’s presently in closed beta.
In response to questions from Romine and audience members, and with his laptop hooked up to the two big screens, Gawne showed how Altium 365 integrated seamlessly with Altium Designer 19 and used the cloud to enable both electronic and mechanical designers to design collaboratively with component distributors and manufacturing by sharing information. Altium users could access Altium 365 through Altium Designer 19 as well as through a web browser. Previously unseen was the ability to link directly with manufacturing inside the Altium 365 platform, allowing users to communicate and check the manufacturability of designs and get pricing for manufacturing all from a click of a button.
Battling bravely against laryngitis, Romine concluded the demonstration by commenting in a failing voice, "With everyone in sync and collaborating in one virtual workplace throughout the design and development process, Altium 365 will help minimize delays and make product development and manufacturing more accurate and reliable. Importantly, Altium 365 will give designers and engineers the ability to design with manufacturing rather than design for manufacturing."
Dan Beeker: All About that Space
“And now for something completely different!” Taking over the MC role from croaky-voiced Romine, Judy Warner, Altium’s director of community engagement, introduced keynote speaker Dan Beeker, senior principal engineer at NXP Semiconductors. He gave a thought-provoking presentation entitled “Electromagnetic Fields for Normal Folks: Show Me the Pictures and Hold the Equations, Please!”
Flamboyant in his trademark bandana, Beeker redefined classical concepts of circuit theory, replacing them with the field management rules he had developed and refined over many years. He made no apology for indoctrinating his audience with the maxim “It’s All About That Space (Not Wires),” having written his own earworm words to fit the tune of Meghan Trainor’s “All About That Bass,” which his daughter had recorded, and he played loudly over the PA system. “I want you to wake up in the middle of the night screaming ‘It’s all about that space!’” My personal experience is that since attending his presentation, I can’t get the tune or his words out of my head. Thanks, Dan!
Beeker offered a quick glimpse into the idea of fields. “It’s not electrons following wires; it’s the electromagnetic field that we are building things to manage. We all are involved in developing products that generate, control, and consume electromagnetic field energy.”
He explained that fields were basic to all circuit operations. Volts and amps made things practical because they could easily be measured, whereas it was more difficult to measure electric and magnetic fields. Without digging too deeply into theoretical physics, he referred to Maxwell’s equations, which were all about the interaction between electric and magnetic fields and didn’t talk about electrons at all. He explained that this was not what we had been taught. Circuit theory suggested that electric energy was made up of electrons moving in the conductors; switches added conductors, and the current instantly started to move in the loop; the wires carried the energy; and the load instantly affected the flow of energy.
But in his view that theory was wrong. Switches added new spaces, and the moving field carried the energy. It took time for the field energy to move into that space. The moving field energy had no idea of what it was at the end of the new space. And field energy moving through a space induced current flow in the conductors. Beeker proposed that current flow was an artefact caused by moving fields in the conductors that bounded them as a result of the fields interacting with the molecules in the conductor. This interaction consumed some of the field energy, resulting in a voltage drop caused by this resistance.
This consumption of field energy resulted in increased movement of the molecules and was hence converted to heat. The dielectric consumed energy in the same way unless it was a vacuum, and electromagnetic energy moved more slowly through a physical dielectric than through space for the same reason. It took time for the interaction between the molecules in the dielectric even in the air because, in Beeker’s words, “The field has to stop and shake hands with every molecule it meets.”
In response to those classical thinkers who would challenge his theories, Beeker responded, “Why do you have so much trouble with signal integrity and EMC?” He suggested that maybe something was missing. “How many designs are right the first time? How many fail EMC? That’s not engineering, that’s just hoping.” He debunked some myths about fields, commenting that “contained” fields such as those between conductors and a plane were friendly while a “loose” field was not. He went on to analyse what was in the waves and how those fields behaved.
Considering frequencies in the analogue domain and accepting a quarter wavelength as a good antenna size, he showed that for 10-MHz HMOS, this equated to 24.6 feet. Meanwhile, for 100 GHz in 32-nm HCMOS, it was half the thickness of a standard FR-4 PCB with a rise-time equivalent to 10 picoseconds compared with 100 nanoseconds for the 10-MHz example where switching speeds were slow and efficient antennas had to be huge. Once the magic boundary of one nanosecond was crossed, most PCB designs provided an enormous source of antennas, and at 10 picosecond speeds, every structure in the system could be considered an antenna. The four-order magnitude change in switching speeds since TTL days had not been accompanied by corresponding changes in PCB or system design philosophy, and the basic laws of physics could no longer be ignored if noise and EMC failure were to be avoided.
So, how to design good transmission lines? It was all about energy management and placing enough energy in the right place on the board to make sure that energy was delivered every time the clock switched. Transmission lines were convenient paths for energy flow, and every conductor pair was a transmission line, whether trace-to-trace or trace-to-conducting-plane.
Beeker also discussed the rules of triplets. “You only get three components to use to build electronic systems: conductors, spaces and switches. And you can only do three things with electromagnetic field energy: store it, move it, or convert it to kinetic energy. It’s not rocket science!” He talked in a meaningful analogy when discussing the design of three-dimensional spaces for managing electromagnetic field movement. Beeker said, “We are all just plumbers using very leaky water pipes.”
He visualised a capacitor as a conductor geometry that concentrated the storage of electric field energy with the energy stored in the space between the plates like a lake between two rivers. Similarly, an inductor was a conductor geometry that concentrated the storage of magnetic field energy in the space around wires and gaps and could be considered a wire stretcher that added travel time to the wave. For a transmission line to be well-defined, the signal trace had to be one dielectric space away from the return, whether adjacent to a planar copper ground or a ground trace. Any deviation from the rule would increase radiated emissions and degrade signal integrity and had to be an engineered compromise and not an accident of signal routing.
Further, he used the water analogy again in discussing wave reflections in transmission lines, visualising a wave going from a low-impedance source to a high-impedance load as throwing a bucket of water out of a small hole in the wall with most of the water splashing back. Conversely, going from high to low impedance was like throwing a bucket of water out of an open window.
Beeker pointed out that transistor geometry was the driver for routing schemes, and that any discontinuity less than one-sixth of the wavelength was virtually invisible to the signal. But failure to ensure that both signal and ground copper were continuous and adjacent could result in large discontinuities that would cause signal integrity and EMC issues.
Returning to his energy management theme, he reminded his very attentive audience that all energy was moved by wave action and that the only way to reduce noise in a system was to provide adequate sources of electromagnetic field energy. The only ways to improve power delivery were to move the storage device closer to reduce travel time, reduce the impedance of the transmission line using wider traces or thinner dielectrics (bigger buckets), or add more connecting transmission lines (more buckets). When a switching element closed, it resulted in a drop in the voltage on the power supply and the resulting field energy request wave travelled until it was filled, or it radiated. Depletion waves were the most common source of EMC issues. “We have to keep the field happy and contained as far up the food chain as we can to reduce system noise.”
Beeker reviewed the primary messages of his presentation that well-defined transmission lines result in significantly improved EMC performance, electromagnetic fields travel in the space between the conductors, and movement of electromagnetic fields induces current flow in the conductors. Further, it is important to consider the time it takes for the electromagnetic fields to move through the dielectric from the transmitter to the receiver and recognise that the switching speed of the output devices determines the requirements of the power supply and the transmission line design.
He closed with a quotation from veteran EE Ralph Morrison. “Buildings have walls and halls. People travel in the halls, not the walls. Circuits have traces and spaces. Energy and signals travel in the spaces not the traces. It’s all about that space!”
Alun Morgan: Afternoon Keynote
The afternoon session of the first day of AltiumLive commenced with a keynote presentation from Alun Morgan, technology ambassador with Ventec International Group, entitled “PCB Base Material Properties and Developments: What Designers Need to Know.” It was another full house, and those attending were keen to benefit from his encyclopaedic knowledge and outstanding teaching skills.
Morgan made it clear that the base material was fundamental to the electrical, mechanical, and thermal performance of every PCB. But how many PCB designers really understood its significance, characteristics, benefits, limitations, and how to choose and specify an appropriate material for a particular application?
From the most basic of first principles to the latest developments in low-loss laminates for high-speed applications and thermally conductive laminates for heat management, he covered a broad spectrum in a plain language. “All you ever needed to know about laminates but were afraid to ask.” He covered a compendium of the technology distilled to emphasise information of particular relevance to designers and dispelled a lot of myths and mysteries along the way. Even the most knowledgeable designers could learn a lot from this presentation.
In its generally accepted sense, a base laminate comprised an insulating material—usually a composite of a resin and a reinforcement—with a conductor bonded to one or both sides. Morgan discussed PCB base material resin types and reinforcements and resin/reinforcement selection considerations; there are so many combinations and permutations. He also addressed electrical considerations such as dielectric constant, dissipation factor, dielectric breakdown strength, passive intermodulation, surface and volume resistivity, and comparative tracking index.
Mechanical considerations highlighted included processability in terms of drilling, punching, and laminating; flexural and tensile strength; coefficient of thermal expansion; thermal conductivity; thermal cycling resistance; thermal endurance; maximum operating temperature; water absorption; dimensional stability; flammability; glass transition temperature; decomposition temperature; and foil peel strength. Lastly, Morgan explained the cost considerations of specifying the appropriate material for a particular design and the difference between thermoplastic and thermoset resins. Thermoplastic resins melt on heating and re-solidify on cooling while thermoset resins—epoxy being the familiar example—harden permanently on heating.
In a schematic representation of the typical process stages for manufacturing a glass-epoxy laminate, he demonstrated how glass fabric was taken from the roll and impregnated with a solution of uncured epoxy resin, then passed through a vertical oven to remove the solvent and partially cure the resin to form a prepreg, which was then cut into sheets. The required number of plies of prepreg was then laid between sheets of copper foil in stainless-steel press-plates and placed in a heated press to cause the resin to flow, gel, and harden to produce a finished laminate.
Resisting the temptation to delve too deeply into chemical theory, Morgan discussed the basic chemistry of epoxy resins, describing how they were manufactured and the mechanism of the crosslinking process with dicyandiamide and phenolic curing agents. Flame retardancy had been a long-standing safety issue, and he explained the differences between traditional brominated flame retardants and the phosphorus-based alternatives developed to satisfy the demand for halogen-free materials.
Further, he discussed the significance of glass transition temperatures, how they were influenced by the curing chemistry of the resin, how different measurement methods could give different results, and how thermal expansion coefficients were influenced by glass transition temperature and could be reduced by the use of fillers.
Morgan also covered the glass fibre manufacturing and the weaving of different styles of glass fabric. Due to differences in dielectric constant between glass and resin, the effects that the weave texture of the fabric could have on signal integrity in controlled impedance designs could be partly mitigated using square-weave and spread-glass fabrics. On the theme of high-speed design, he also introduced the concept of the impedance triangle, demonstrating that for an equivalent dissipation factor, a lower dielectric constant enabled larger trace geometries for improved manufacturing yield.
Morgan discussed conduction and dielectric losses, explaining that conduction losses were primarily resistive losses in the conductor layers and leakage of charge through the dielectric, while dielectric losses resulted from the alternating electric field causing movement of the material’s molecular structure generating heat, which increased with increasing frequency. Dielectrics were insulators because they had few free electrons available to carry electrical current. When subjected to an electric field, polarisation occurred whereby positive and negative charges were displaced relative to the electric field. This polarisation reduced the electric field in the dielectric, causing part of the applied field to be lost. High-speed designs had driven the development low-loss laminates whereas standard FR-4 typically had a loss tangent greater than 0.015 and special-purpose materials were currently available with a loss tangent less than 0.005.
It was essential that the copper foil that formed conductor layers was securely bonded to the base laminate so that etched tracks would exhibit adequate peel strength. The foil was manufactured by electrodeposition, and as part of the process, the bonding surface was given a roughening treatment to optimise adhesion of the resin during the laminating process. Morgan explained that as frequency increased, the signal was not carried within the conductor cross-section, but predominantly in the surface as a result of a phenomenon called the skin effect. The roughness of the foil could cause problems with signal integrity, and for high-frequency applications, special grades of low-profile, very-low-profile, and ultra-low-profile foil were available.
He concluded his presentation with an introduction to the technology of thermally conductive laminates used to keep heat-generating components cooler and reduce system costs by reducing the requirement for heat sinks and cooling fans. Typical applications were the thermal management of LEDs and embedded components. These laminates consisted of a copper foil bonded to an aluminium base via a thin reinforced—or non-reinforced—thermally conductive dielectric and offered the benefits of straightforward PCB fabrication and assembly using standard equipment. Thermal conductivity was measured in watts per metre Kelvin, and materials were already available with values from 1.0–7.0 W/mK, with 10 W/mK in development.
Morgan shared an enormous amount of knowledge with this comprehensive introduction to printed circuit base materials and especially gave an insight into the effects of material characteristics on signal integrity. He may not have used exactly the same terminology and analogy as Dan Beeker, but the two presentations nicely complemented each other as learning experiences for electronics engineers, adding breadth and depth to their understanding of the physical and practical realities of the PCBs they could create with their design automation tools.
The AltiumLive team should be congratulated on pulling together an impressive programme and attracting such an attentive audience for their eminent keynote presenters. Here, I have reviewed only part of the proceedings of the first of two intense conference days. Besides these keynotes, there was a wide-ranging agenda of technical user presentations and professional breakout sessions, which I will report separately.
Visit I-007eBooks to download your copy of Altium micro eBook today:
The Printed Circuit Designer's Guide to...Design for Manufacturing (DFM)