Transistors were developed for two purposes – the amplification and switching of electronic signals. The word transistor is the portmanteau of ‘transfer resistor’, which means that it gets all the work done by changing the resistance value.

After the invention of the first point-contact transistor in 1947 by Shockley, Bardeen and Brattain, scientists have fabricated various types of transistors for different applications. All transistors serve the basic purposes of amplification and switching, but each transistor type works in a very different way. Consequently, their performance and price vary greatly, due to the internal structure of the respective device.

During the following years, two genres of transistors dominated the circuit design – the BJT (bipolar junction transistor) and the MOSFET (metal-oxide semiconductor field-effect transistor). These two types have various pros and cons which are discussed in detail below.

BJT transistor

All transistors are made using semiconductors. Regarding the BJTs, doping these semiconductors using various atoms, we can create P-type (P-doped or NPN) and N-type (N-doped or PNP) semiconductors which have holes and electrons as electricity carriers. This is why the term bipolar is used, to explain that both holes and electrons act as electric current carriers.

MOSFET transistor

In case of the MOSFET, only a single carrier exists. With N-type MOSFETs, electrons are the carrier, and with the P-type MOSFET, it is the hole that carries the current.

As mentioned above, BJT and MOSFET transistor structures determine performance and price. The differences in structure are significant. The BJT is a 3 terminal transmission device having an emitter, a collector and a base. The MOSFET is 4-terminal device- the body of which acts as one of its terminals- and it features source, gate and drain terminals in addition to its body transmission terminal; although in practical circuits the body is tied with the source, so it acts as a 3-terminal device.

In the case of current conduction, BJT creates two p n junctions in the collector-base and emitter-base boundary. The collector-base junction is much thinner than the emitter-base junction. Due to this, the collector current is almost equal to the emitter current and only a little amount of current flows through the base. Conversely, in the case of MOSFETs, the gate current is ideally zero (infinite input impedance) and gate voltage creates a conduction channel for the carriers.

Due to these structural differences, these two transistors do not perform equally in all situations.

Traditionally, BJTs are very suitable for high frequency applications, whereas MOSFETs are low frequency device. This is due to the mobility of the carriers. In BJT, collector current is controlled by the base current, but MOSFET current is controlled by the gate voltage.

In cases of high power devices and load varying devices MOSFETs are suitable, because during the switching period greater loss occurs in the BJTs. (A big disadvantage for BJT in these circumstances is the idle-state current.) Current always flows through the base however small it is. In MOSFETs the gate is isolated by an insulator and current through the gate is virtually zero (unless quantum tunneling occurs).

In microprocessors where millions of transistors are used, this can cause huge loss. Modern CMOS (complementary metal–oxide–semiconductor) architecture consumes almost zero power in the idle state for both 0 and 1 logics. BJT is a positive temperature coefficient device, which means if the temperature is increased then collector current flow increases, which in turn again increases the temperature. That’s why BJTs cannot be used in parallel configuration because the transistors will be destroyed; however, it is completely safe for MOSFET to be used in parallel, because they have negative temperature coefficient.
BJTs are least expensive in price. Hence, low power circuits such as LED drivers (using the popular Darlington pair) are often built using the BJT. Also, the BJTs typically have higher fan-outs (number of transistors attached with the collector output) than the MOSFETs.

In times of digital circuit design, when using TTL (transistor-transistor logic) built with BJT, it is not a fault if any pin if an IC (integrated circuit) is kept open. Yet, when using the CMOS series of digital ICs, it is important that all the dangling pins are strapped with either Vcc or GRD (ground), using pull-up resistors.
As MOSFETs are voltage driven, the accumulated charge in the open pins may create noise and make the output unstable.

Image via AppleInsider/Double Helix Cables

Patents, intellectual property and proprietary technology can create billions of dollars for companies that own them. The threat of competition and someone else copying your trade secrets are always at stake. That competition creates great products though. Companies must innovate and stumble upon new ways to sustain and create value for their customers, and if they’re a public company, they must constantly answer to shareholders who demand increased earnings and profits. We have seen an example of competition and proprietary technology recently with the release of the Apple iPhone 5.

It’s pretty much been commonplace that when you buy a new gadget, there will be a cheaper, third-party adapter cord or cable that will work just as well for transferring data or power. That does not seem to be the case with the new Lightning cable for the iPhone 5. While tearing apart the Lightning cable, Peter from Double Helix Cables discovered that this new Lightning connector had an authentication system inside. Apple had placed an authentication chip between the power pin on the Lightning plug and the USB contact. It’s not uncommon for authentication systems to be integrated into advanced electronic systems, but this is the first time that this type of chip has been seen in a piece of electronic equipment as simple as a cable used for charging.

What did this chip control? It could be used simply for power regulation, but it didn’t have the characteristics of a voltage regulator IC. Further investigation by this cabling expert led Peter to believe that the 8 pin layout on each side of the plug meant the pins must be dynamically assigned. This allows users of the iPhone 5 to flip the connector either way when plugging the Lightning cable in. The chip may play a part in dynamically assigning each pin. Most importantly, he came to the conclusion that the cable would not work without this special miniaturized device. Would third-party knockoff Lightning cables work for the iPhone five 5 then? Probably not.

Reverse engineering is the competition’s key weapon when trying to understand how a software program or tech device works. It’s sometimes easy, and at other times can never be done. It may end up never being done in this case. Peter mentioned that no one was able to reverse engineer Apple’s MFi (Made For iDevices) program before.  Apple requires developers who want to create accessories for Apple products such as the iPod, iPad or iPhone to buy a specific license to do so. A Developmental License or a Manufacturing License will allow registered companies to create or manufacture MFi accessories.

If the chip does have some type of code on board, it would make it difficult for a third-party company to replicate. If the chip is involved in any necessary functions for the iPhone 5, the chances of reverse engineering it are pretty low. Apple has gone on the record as saying this is a “smart” connector.

The verdict is still out on what this chip’s role is in the Lightning cable. It may be one way for Apple to protect it’s devices. It could also be as simple as Apple’s way of adding value to the user, by allowing him to connect the cable either way into the phone. Regardless of the answer, Apple stands to make a hefty profit from just the cable if third-party manufacturers can’t replicate a solution.

via AppleInsider/Double Helix Cables

The designs produced by Computer aided design can also be used in manufacturing through a process known as Computer aided Manufacturing. This actually means that CAD and CAM are related aspects of computer design and manufacturing.

Computer Aided Design is a form of computer based tool that is used to draft and design. CAD is used in various fields including electrical engineering, architecture and mechanical engineering among others. This software enables designers to create precise drawings and sketch plans of various products. CAD provides the needed flexibility required to create different dimensions with little effort.

CAD can be used by both professional and amateur designing enthusiasts. The software comes with a number of templates and symbols for various designing and drafting functions. This makes it suitable for a number of applications. All 2D and 3D designing functions can be done with CAD. It is a useful tool especially for designing professionals such as architects, and engineers among others.

CAD is applied in various professions including aerospace, automotive, machinery, shipbuilding, consumer goods and others. In electronic engineering, CAD is used for manufacturing process planning and design of digital circuits among other applications. In architecture, CAD is used in the design of all types of buildings as well as providing an assessment of steel-framed structures. CAD enables architects to create 2D and 3D building designs that provide an almost real replica of the original design. CAD is significant in various engineering processes such as creation of conceptual plans, laying out and analyzing of manufacturing components and others. CAD applications can now be installed in personal computers to enable individuals design while in their homes.

On the other hand, CAM or Computer Aided Manufacturing refers to an automated method of converting product designs and drawings into code formats readable by machines that manufacture various products. This design tool compliments CAD or Computer Aided Design. CAM offers a wide range of applications in various manufacturing fields. The technology involved in CAM has gone through rapid changes from those used to control CNC machines to the modern CAM tools that can control whole sets of manufacturing processes simultaneously.
CAM allows the direct communication of instructions and procedures to manufacturing machines. CAM can control machines used in robotic mill processes, welding and lathes among others. The computer automation tool allows various machines to move materials from one point to another by allowing the completion of various steps in a systematic manner. Finished products can also be moved to other locations for packaging, final checking and synthesizing.

Some of the applications of CAM include drilling, PCB imaging and in other industries spinning, glass working, metal working, woodturning and providing graphical optimization of whole manufacturing processes. Three dimensional solids with intricate details and intricacies can be created with a CAM system. Various products such as candlesticks, bowls, baseball bats, table legs and crankshafts can be created using a CAM system.

CAM is applied in a number of fields such as industrial, aerospace, mechanical and electrical engineering. Various applications such as kinematics, thermodynamics, fluid dynamics and solid mechanics depend on CAM systems.

LVDS was originally developed as a high-speed protocol that allowed for data transfer rates of up to 650 Mbps. Today, the industry has far exceeded this target and LVDS is capable of data transfer rates approaching 3.2 Gbps. One of the primary reasons for this is the fact that prices for main memory modules of the DDR2 and DDR3 variety have dropped so drastically in the past few years. Likewise, bus architectures have undergone a major transformation with the advent of high-speed data transmission standards such as HyperTransport and PCI Express.

All of this innovation presents a serious challenge to the PCB designer, who must now take into account parasitic effects and EMI issues that can impact signal integrity and cause circuit failure. Advanced circuit simulations can be used to identify macro-design weaknesses, but it’s equally important to design each circuit with as much symmetry and structure as possible in order to achieve a predictable result. Wherever possible, it’s always a good practice to follow the chip vendors’ guidelines and specifications, since they have a wealth of knowledge about their products and know how they should be integrated into a PCB.

Inevitably, there are situations where the PCB must be designed using non-standardized layout rules. In these scenarios, the designer may need to perform additional detailed simulations for each critical circuit. More importantly, he or she must identify and model the routing deviations to assess potential problem areas.

Route Topology Symmetry
Low-voltage differential signaling (LVDS), unlike single-ended routing, utilizes two parallel traces between the receiver and the driver. A small termination resistor is placed between the traces that circulates a current, which the receiver detects as the signal voltage. This approach delivers superior signal integrity, since it effectively eliminates ringing and greatly reduces electromagnetic interference effects. In addition, LVDS is far more resistant to noise glitches and other discontinuities that can degrade signal integrity.

Impedance matching of the differential pair is the key to using LVDS effectively. This means more than just ensuring that the trace lengths are identical. The differential traces should be routed from the transmission point to the receiver using as much symmetry and repetition as possible. If they’re absolutely necessary, deviations should be modeled in greater detail to assess their overall impact on performance.

Signal Conditioning Techniques
Typically, PCBs are designed using multiple vias and thru-holes to route traces between the various signal layers. These three-dimensional deviations, if identified correctly, can sometimes be compensated for by using additional intelligence to mitigate impedance variations along trace lengths.

• Pre-Conditioning
  Conditioning relies on transmitting a distorted signal to compensate for discontinuities that can decrease signal integrity and cause failures. Common methods include modulating the signal amplitude of signal transitions over time and boosting the high-frequency components over the frequency domain.
• Timing Skew
  For high-speed data buses, fly-by routing of DDR3 DIMMS is becoming much more common. The fly-by technique necessitates that all of the command signals be routed to connect in a daisy chain rather than in parallel like previous memories. Since each module is routed serially with the next one in line, the control signals will arrive on the bus at different times. To adjust for these delays, additional skew can be added to compensate for the delays to ensure that the signals arrive on the bus at the proper time.
• Impedance Matching
  A relatively recent trend is the emergence of FPGA devices that can detect and adjust to variations in impedance. FPGA-based impedance matching involves adjusting the termination resistor value on the fly to tune the receiver so that it yields the best noise immunity. These techniques are still in the early stages of development at the moment.

Additional Simulation Methods
The more you know about each device component’s internal structure, the easier it will be to model and simulate your design to ensure that it functions properly. Some vendors will supply their internal device schematics in their spec applications. These components can then be modeled using SPICE or can even be added directly as a circuit schematic in your network package.

Scattering parameter modeling treats the network as a collection of black boxes comprised of basic, unknown electrical components such as resistors, capacitors and transistors. Different scattering frequencies are then applied to the network to study the behavior of the transmission model between the ports. Since little is known about the internal structure, this technique may require thousands of iterations to arrive at a satisfactory solution. If detailed hardware measurements are added to the model, the number of perturbations can be reduced to improve accuracy.

System On a Chip (SoC) motherboards are becoming the standard in PCB design due to miniaturization. These systems are being developed to service the workstation, server and PC markets and are used extensively in telecommunications networks. As such, it’s especially important that the designer pay special attention to matching the trace impedances to ensure that the design will be functional and will perform as desired. Whenever possible follow the component vendors’ recommendations and layout rules to prevent complications. If deviations are necessary, additional modeling of the changes should be undertaken before completing the layout.