High Speed Digital Hardware
The following sections describe some of the areas of digital and communications hardware in which FCL has built up significant skills, experience and knowledge. Click on the topic of interest for more information.
Today, multi-gigabit per second (Gbit/s) digital signalling speeds are deployed even in many of the most basic consumer products. Here, we are assuming that the information transmission will be carried 'at baseband'. A baseband frequency spectrum extends from at or near zero (DC) up to some upper limit, depending on the type of coding used and the defined data loading. Normally the latter is intended to be a random distribution of 1s and 0s, with the probability of each occuring to be 50%, measured over a large numbers of bits. This is actually quite challenging to generate and would be difficult to use in practice precisely because it is truely random. The bit error rate test equipment (BERT) requires to compare the bits measured at the output of a communications channel with those which were transmitted at its input, perhaps thousands of kilometers away. The widely accepted solution is to use a pseudo random bit sequence (PRBS) instead. This is an artificially generated stream of bits which is not mathematically random, but close to it. The PRBS bit pattern uses a shift register with feedback so is predictable and repeats after a fixed number of bits, the length of which can be adjusted. The bit stream repeat period may be set appropriate to the signalling speed of the data channel and the test duration. One big advantage of using PRBS is that the transmit and the receive BERT can be set to the same PRBS definition so that the latter will 'know' what the error free bit stream should be and therefore be able to measure the errors.
The associated data streams propagate across physical interfaces which are hopefully well designed transmission lines. Today, differential transmission lines are more popular for high speed data than single ended types, which may still be adequate for lower speeds. The large bandwidths necessary mean that the electrical transmission properties of these are critical to achieving good signal integrity: reliable, low-error performance. Important PCB and layout properties which must be sufficiently controlled in order to achieve this include: controlled impedance architecture, connector mismatches, discontinuities and differential line lengths which cause differential delays, often called 'skew'. Electrical properties include: compliance voltages, common mode rejection, component self resonant frequencies (SRFs), Q factors and loss tangent.
We know from Fourier transforms applied to streams of digital data that, assuming no noise, perfect digital signals occupy an infinite bandwidth. That is not possible of course, so we normally have to settle for the maximum possible available bandwidth that the regulatory authorities will allow us and the minimum occupied bandwidth that will enable adequate transmission of the data channel we have in mind. To set up a radio link we require to modulate a carrier in some way using the baseband, before radiation. Bandwidth is scarce, expensive and possibly not even available and noise in one or more of its many forms is always present. Additive white Gaussian noise (AWGN) is the systemic type of finite noise which influences the integrity of digital data. This always causes transmission errors to some degree: the laws of physics do not allow error-free transmission of data. These errors, normally expressed in the form of the bit error rate (BER) parameter will always exist. Chris Angove, represented by FCL, has a good understanding of many digital transmission parameters such as these, their limitations, the related theories and how they may be applied. So the spectra of high speed data can extend to very high frequencies, perhaps well into the microwave region. We have to design the associated transmission lines to the same standards we would use for microwave transmission. Many of the parameters used and the equipment required to measure them are identical to those used for microwaves. FCL's experience in RF and microwave engineering has developed well into a wider brief including high speed data transmission lines of many different types. Some of this is summarised in the following sections.
Although we live in a digital world, most natural physical phenomena are analog. For example, radio wave propagation and any form of modulation, even if it is described as a 'digital' may be described by some engineers as analog. For example, one form of digital modulation is quadrature amplitude modulation (QAM) which relies on creating changes in both the amplitude and phase of the carrier being modulated. One type of QAM is 256QAM which is designed to distinguish amongst 256 different states of amplitude and phase, each with a unique 'symbol'. The modulated carrier may eventually be propagated by an electromagnetic wave from an antenna. It may even be carried by a cable, for example to meet one of the 'DOCSIS' specifications. Even with the propagation path (wave or cable) being analog, with today's digital technology, we will probably need to process digital signals at each end of the communications channel and possibly at intermediate points for functions such as coding, filtering and compression. Processing can of course be analog or digital but today certainly most of it is digital. Ultimately there will need to be human intervention: either input or output and this is where the analog signals are usually required. In these types of channel there needs to be at least one digital to analog converter (DAC) at the transmit end and one analog to digital converter (ADC) at the receive end. These are just two applications of ADCs and DACs. There are many others.
Early ADCs were designed to simply convert an analog input voltage into a digital form using a successive approximation register (SAR). The best conversion times were lengthy by today's standards. SARs are still used today. They are much faster than they used to be and are typically used in accurate but slow responding devices such as digital multimeters and temperature sensors. More recent ADCs are sampling types, found in many different types of communication equipment. In these, sampling theory and the work of Nyquist and Shannon feature prominently. FCL has a good knowledge of these types and many other ADC architectures including: SAR, time interleaved, pipeline, sigma delta, flash/parallel and photonic sampling.
Typically, in today's electronics hardware, there appear to be more ADCs in use than DACs. Perhaps that is because more analog parameters are processed and the data transmitted in digital form and there is infrequently the need for an analog output. FCL also has a good level of knowledge and experience of DACs. With these we need to carefully select a suitable sampling frequency and make allowances for the shaping of the analog spectrum by an inverse sinc function. The quality of the output analog waveform is influenced by the 'order hold' that is used: what happens to the analog output between samples.
Over the years, as digital signal processing (DSP) overheads and speeds have increased, ADC signal input frequencies have moved from near DC, through audio and are now well into the radio frequency (RF) spectrum. In the typical receiver architecture, it has become possible to now digitally perform many of the functions that were previously done in analog. The configuration of digital hardware is usually programmable by software, often a form of hardware description language used to generate firmware, specific instructions for configuring the hardware. Such digital implementations have the major advantage that their configuration may be updated and modified in code, provided that this does not exceed the limitations of the hardware and of course the Nyquist sampling criteria. A common piece of digitally programmable harware may be used to operate both across a range of frequencies and to demodulate or modulate a variety of modulation types, both digital and analog.
Software defined radio (SDR) refers to a flexible technique which may be applied to re-configurable radio hardware in order to set up a specific radio architecture. The SDR configuration is defined by how the associated SDR software was used to program the hardware. The SDR software may define the frequency that the target system will operate at, the type of modulation used and various other radio-like parameters. It is ofter written in a high level programming language like C or C++ and compiled before programming the hardware. More recently, applications like MATLAB® have allowed the generation of SDR code.
The dynamic range of a SDR radio is improving but is not yet considered equivalent to a good quality superheterodyne radio designed for a similar frequency. The superheterodyne architecture will not however have the re-programmable flexibility of the SDR version. Currently we are at an intermediate stage with SDR. ADC sampling speeds are still insufficient to allow direct (baseband) sampling for some of the higher operating frequencies, such as those used in digital cellular and satellite communications. So it is often necessary to use some analog circuits at the front end of a SDR receiver to convert the frequency band received by the antenna into one that is more manageable by the ADC. Usually a low noise amplifier (LNA), some filtering and a mixer are used to change the incoming frequency band to a lower one more compatible with the current technology. Actual sampling may then be performed at baseband or zero-IF, or across a band of frequencies (bandpass or undersampling).
FCL is experienced in specifying many different types of ADCs and incorporating them into multiple designs. Parameters addressed include resolution, integrated non-linearity (INL), differential non-linearity (DNL), kTB and quantisation noise, effective number of bits (ENOB), missing codes, aperture jitter, dithering and multiple carry issues. These have naturally led onto several digital signal processing (DSP) phenomena such as spectral leakage, windowing and the implementation of forward and inverse fast Fourier transforms.
FCL's initial experience in digital cellular communications was with a client company manufacturing Global System for Mobile (GSM) base station equipment designed for what we now refer to as the second generation mobile telephone service (2G). This particular implementation was around 900 MHz frequency division multiplex (FDM) using time division multiple access (TDMA). The GSM standard was first implemented in Europe but quickly became accepted and widely adopted throughout the world.
The first generation (analog) cellular mobile telephone service is generally considered to be 1G, although it was not known as such at the time. FCL's 2G work was related to the frequency hopping synthesizer boards that were part of the GSM equipment. Each board used a fractional 'N' (FN) IC. The FN device was an alternative to the other popular type of synthesizer IC at the time, the dual modulus (DM) type. FN ICs have the advantage that the step frequency can actually be a fraction of the phase detector frequency. This is unlike the DM type IC, where the step frequency must be an integer multiple of the phase detector frequency. The FN device is useful for two reasons: it allows a relatively wide bandwidth phase locked loop (PLL) to be used giving a fast lock time and it enables a common issue, the phase detector frequency modulation spurii, to be more easily suppressed.
Soon afterwards, FCL also supported another client with a similar product range of base station equipment, but this time for the DCS1800 standard in the UK. Its operation was very similar to GSM but the channels were allocated around 1800 MHz at approximately twice the frequency of the GSM 900 MHz service. In this case a DM synthesizer approach was adopted, largely because of the differing device performances observed at the different frequencies.
FCL has experience of base station and remote station hardware used for 2G, 3G and 4G/LTE digital cellular networks, especially the RF aspects. In 2011 LTE-advanced (LTE-A) was released. This specified several improvements over LTE including multiple antenna technologies and higher data rates, especially for mobile and nomadic stations. This is achieved using carrier aggregation which enables the effective channel capacity to be increased by combining contiguous or discontiguous frequency bands, even if they are in different service frequency bands.
Digital cellular communications is one area in which FCL has supported the development of RF over optical fiber (ROF) technology to extend the cellular services into tunnels, buildings or other areas which would otherwise receive poor or no coverage. The ROF work was focused on the development of remote units, usually locally powered, but connected to head units within the normal service area by ROF, optical wavelengths of 1310 nm in one direction and 1550 nm in the other direction. This work required a good understanding of the optical modulation/demodulation techniques, the constituent optical components, fiber optic cables, connectors and the peculiarities of the optical transmission domain compared to the analog (RF) transmission domain.
FCL has been working with traditional superheterodyne receiver architectures for several years now and has a good understanding of the architectures of common digital receivers and transmitters. The knowledge extends to post-superhet designs that contain zero-IF (homodyne or direct conversion), low-IF and direct IF (or bandpass) detection. FCL has a good appreciation of the properties, advantages and disadvantages of each of these types compared to the original superheterodyne.
The issues of SDR compared to the more traditional physical hardware architectures is another working area. Many of the SDR algorithms are already established so most of FCL's support has been working with the DSP implementation in developing fully integrated, re-configureable and upgradeable receiver products.
FCL has implemented Fourier transforms to convert regular baseband voltage-time pulse trains to their associated frequency spectra. It has also used the built-in Matlab® fast Fourier transform (FFT) function with the Communication Systems and DSP toolboxes, to analyse arbitrary voltage time waveforms. It is a relatively straightforward transformation then to apply the baseband filtering law as required to shape the waveform to minimise the inter-symbol interference (ISI) and to optimise the occupied bandwidth after modulation. A further transformation (frequency shift in the complex plane) may then be applied to (digitally) modulate the baseband onto a carrier.
FCL has a good understanding of quadrature amplitude modulation and demodulation techniques and the typical architectures of IQ mixers, extending to how they are used in digital receivers and transceivers. This comes with familiarity of how signals are represented and processed in the complex plane at analog, before the ADC in receivers and after the DAC in transmitters.
Whilst supporting a client's team responsible for designing terminal equipment for optical fiber submarine cables, FCL worked on techniques for dense wavelength division multiplexing (DWDM) using Mach Zehnder (MZ) optical intensity modulators. These were used to modulate optical wavelengths, with high speed digital data streams, onto optical fiber cables for carrying high capacity telecommunications infrastructure data. The modulation source was applied at baseband to each optical wavelength around 1550 nm, using the synchronous digital hierarchy (SDH) standard at a speed defined as STM-64, equivalent to a synchronous optical network (SONET) speed of OC-192, or approximately 10 Gbit/s. FCL's input was in designing a device for controlling the phase of the clock source using I-Q modulators. The bandwidth of the whole baseband after filtering was from approximately 50 kHz to nearly 20 GHz.
One of FCL's assignments with another team was to predict the performance of a proposed 'RF over optical fiber' (ROF) system end to end through the electrical domain to the optical domain and then back again to the electrical domain. This system was proposed to carry various communications channels via a remote location as a contingency. FCL was responsible for specifying the optical and electrical components for procurment, followed by building and demonstrating the prototype to the customer, all hands-on and to tight timescales. The optical components used included MZ intensity modulators, DWDM filters, circulators, attenuators and an erbium doped fiber amplifier (EFDA). The demonstration was highly successful, giving the client new expertise and enhancing their range of products.
Another ROF project was supporting the development of a remote unit designed to extend the coverage of digital cellular networks into tunnels or areas of previously inadequate coverage. It was also used for effectively increasing the capacity inside large buildings such as airports and sports stadia. Essentially the same hardware, except for a few frequency dependent components, was common across all of the European and American digital cellular services: 2G, 3G and 4G/LTE/LTE-A. The stand-alone unit was suitable for supporting single (one uplink and one downlink) band or dual (two uplink and two downlink) bands simultaneously.
Some of FCL's more recent work has been concerned with the latest generation of carrier access types for wireless networks, orthogonal frequency division multiple access (OFDMA) and single carrier frequency division multiple access (SC-FDMA). OFDMA and SC-FDMA are used respectively for the downlinks and uplinks of 4G/LTE/LTE-A digital cellular communications services. Both types support relatively wide on-air bandwidths with each band containing many closely spaced, orthogonally modulated sub-carriers. These types of access methods are well suited to propagation in multipath-rich environments, very common with mobile communications. OFDM access methods are also used in WiMAX, WiFi and digital terrestrial broadcast services (DAB, DVB-T and DVB-T2). Actually, OFDM is not a particularly new wireless access technology, it was suggested many years ago, soon after the discovery of the 'fast' algorithm for the Fourier transform by Cooley and Tukey. However, it has only been in the last few years that small, portable, battery operated devices such as radios and mobile handsets have included sufficient processing power to make it a realistic mobile/portable solution. FCL's knowledge of the principles of OFDM, extend to its practical use in a heavy multipath environment, together with link calculations and special properties such as Doppler shift, cyclic prefix, convolutional coding, Reed Solomon coding, scalable bandwidths and modulation methods.
Sampling theory is central to the understanding of many types of ADCs, DACs and other digital signal processing (DSP) components. With the help of well known and reliable references, FCL has dedicated time to researching this important subject. This has helped to provide a better understanding of many of the issues that clients' digital hardware present. This often appears in parallel with issues such as Nyquist zones, Fourier analyses and the Shannon noisy channel information criteria, also important parts of FCL's portfolio.
With todays math CAD applications such as Matlab® and MathCad®, both of which FCL has invested in, performing important transforms for signal processing, such as those of Laplace and Fourier is quite routine. FCL has dedicated resources to understanding the theories involved and, importantly, how they are used and how the results may be interpreted in the real world.
Laplace transforms have been used with Matlab® to predict the step responses of transmission lines in the time domain when they have been loaded with reactive components. Fourier transforms have been used on several occasions, again using Matlab®, to analyse the component frequencies with amplitude and phase information of periodic voltage time waveforms.
Pressure is often placed on PCB designers to put as much functionality as possible all on one board. If this is electrically and physically possible, a well designed multi-function PCB will reduce the cost of manufacture of the equipment significantly compared to having separate boards for each of the functions. A good example of how this principle has evolved over many years is to look at the typical personal computer motherboard. Much of the necessary functionality can now be included on this, usually relatively small, one board with just a few exceptions.
Today's PCB designs usually include many sandwiched conductor layers using various technologies to make connections where necessary through the layers to pads and holes on the surfaces to match the footprints of the components to be fitted. Without doubt, the most common dielectric material used to electrically isolate the layers is known as FR4®. Although this has been used for many years, it is still quite adequate today on many PCBs even allowing for the rate at which frequencies and signalling speeds have increased. FCL has a detailed understanding of the limitations of many different conductor and dielectric materials that are used in PCB design as the operating frequencies and signalling speeds have increased. Associated parameters include loss tangent, skin depth, differential mode impedance and common mode impedance.
There are many PCB design tool applications available ranging from free and open source ones to very expensive and sophisticated types targetted at the more complex designs. FCL has some experience with Altium Designer 16.1® and Cadence Allegro 16.6®. With most clients, however there are usually a few people who are experts and highly skilled in using applications like these and it is normally the job of FCL's representative to work with them to achieve the client's objectives.
FCL has good experience of the particular challenges of designing PCBs in which various different signal types on the same board come in to close proximity. This has been developed over some years of dealing with EMC and debugging issues across a range of frequencies from DC to many gigahertz. Problems solved have included: isolation, emissions (radiated and conducted), susceptibility (radiated and conducted), controlled impedance lines, transitions, matching and thermal aspects.
FCL has experience and knowledge of working with a variety of high speed transmission lines in both copper and optical fibre. In copper for PCB architecture these include low voltage differential signalling (LVDS), current mode logic (CML), positive emitter coupled logic (PECL) and peripheral component interconnect express (PCIe). High and moderate speed transmission standards encountered include Ethernet (Gigabit), SATA, RS-485, SPI, I2C and even RS-232. Experience with optical fibre have related to Ethernet (Gigabit), monomode connectorised at 1310 nm and 1550 nm and multimode connectorised at 850 nm. This extends to both baseband types such as Ethernet and modulated versions such as DOCSIS.
By high speed, as opposed to high frequency, we are generally referring to a data information rate in bits per second (bit/s). This may be extended to symbols per second if we are looking at the part of a circuit where the data is in symbols instead of bits. However, the principles remain unchanged whichever units we use: we are looking at discrete changes (usually of voltage) with time. By convention, 'frequencies' are not used to describe the speed of a data transmission but they are used to describe the speed of a periodic repetition of clock signals, whether they are sinewave or digital.
The most fundamental oscillation is sinusoidal (or cosinusoidal) so any periodic voltage-time waveform can be expressed, by a Fourier transform, as a sum of sinusoidal frequencies including zero (DC) which have integer harmonic relationships. These are most definately expressed as frequencies and, in a digital waveform, there may be significant levels at very high frequencies, many times the frequency of the original (regular but digital) periodic waveform. FCL's experience in RF and microwaves is often very valuable to clients in addressing issues like these with high speed data. The problems that do occur are usually caused by inadequate performance affecting the higher frequency components more than those at the lower frequencies. If for example, a data interface was running at 1 Gbit/s, the significant sinewave frequency components involved might extend up to and beyond 10 GHz. Solving the issues by using a philosophy targetted at 10 GHz will often naturally improve the lower frequency components. FCL has considerable experience of the challenges encountered around and beyond these frequencies.
Digital modulation is yet another one of those slightly ambiguous descriptions we often find in electronics and communications. It could be argued that modulation itself is not actually digital but analog. However, in digital modulation the voltages and phases concerned are discrete even though they might be quite close together, depending on the order of the modulation.
FCL has experience of many types of modulation, those generally described as analogue and digital. Most recently clients have required knowledge of quadrature amplitude modulation (QAM) as various forms of it are so widely used in digital communications systems. FCL has used high speed sampling oscilloscopes to display the outputs of QAM modulators as constellation diagrams. Parameters used have included error vector magnitude (EVM), inter-symbol interference (ISI), phase and amplitude imbalance, bit error rate (BER) and Eb/N0. Eb/N0 is the energy per bit divided by the additive white Gaussian noise (AWGN) power density. It is often used on a logarithmic (dB) scale.
The particular type (order) of QAM in use is identified by 'nQAM' where n is 2 to an integer power starting at one. 2QAM is the most basic but is usually called binary phase shift keying (BPSK). BPSK is used typically when only small signal to noise ratios (SNRs) are available and the data capacity of the channel might be quite low, such as communications with deep space vehicles. As the QAM order increases so the SNR must increase to accommodate the greater capacity. The highest QAM orders are found on cable systems like DOCSIS where relatively high reliable and fade free powers can be achieved.
For an information channel subject to additive white Gaussian noise (AWGN), the ratio Eb/N0 describes the ratio of the energy per bit (Eb) in joules to the AWGN noise power density (N0) in watts per hertz. This is a measure of the quality of signal plus noise combination, like signal to noise ratio (SNR) in an analog channel. FCL is familiar with the theory behind and use of the Shannon Nyquist channel capacity theorem, which uses this parameter, and how it may be applied to identify the 'Shannon limit' and the relationship between Eb/N0 and the maximum theoretical available channel capacity. This theorem provides a means of calculating the maximum possible theoretical capacity of a given channel provided sufficient (noise) bandwidth and signal power are available. For a given Eb/N0, the result is a value in bit/s/Hz. That is a measure of channel capacity against the available bandwidth. For short range communications, for example over a local line-of-sight path, or even a cable path, a relatively high Eb/N0 value would be possible which would allow a high capacity channel to be set up, typically using a fairly high order quadrature amplitude modulation (QAM). However, as Eb/N0 reduces at greater distances, the capacity is reduced and lower order QAM is required to tolerate the higher levels of AWGN. Ultimately the order may be reduced to the minimum of 2QAM or BPSK to allow some information flow, allbeit at a rather slow rate.
In the early days of RF and wireless engineering, the coaxial cable transmission line became popular because it provided a grounded screen surrounding a 'live' center conductor. The screen was often referred to as a type of 'Faraday cage' architecture and was intended to screen external interference from penetrating the cable and coupling to the center conductor. The same theory is true in reverse: the screen would block potential interference radiating from a similar cable. We know that the effectiveness of the screen depends on several factors including the type of foil and/or braiding, the metal used, its thickness and the measurement frequency. We also assume that the whole of the screen is at the same (ground) potential for the full length of the cable. As data signalling speeds and frequencies increase however, and we have to meet more stringent EMC requirements, we find that the imperfect screens and grounding are inadequate and a better form of transmission line is required.
With just about any transmission line, external radiated interference is more heavily coupled in the common mode than the differential mode. That is because the spacing of the conductors normal to the axis of the line is usually much smaller than the axial length of cable exposed to the interference. By using the signal to excite the cable in as close to perfect differential mode as possible, the common mode interference should be heavily reduced.
Unfortunately, the requirement for differential transmission lines does complicate PCB design. We may still need some form of screening. How close should the individual lines be? How long should they be and of what dimensions? How much can we tolerate them being of different lengths? These are the sort of questions that FCL has been answering for several years now. FCL has in-depth knowledge and experience of transmission line designs in many different forms.