VLSI Testing — digital and mixed analogue/digital techniques: S.L. HURST (IEE, 1997, 532 pp., 65)

This text is another excellent addition to the IEE Circuits, Devices and Systems series. The book provides a comprehensive introduction and a reference for all aspects of IC testing, from practical test strategies and industrial practice, to economic and managerial issues. Unlike many books on the subject, it also gives in-depth coverage of analogue and mixed analogue/digital testing. I particularly like the succinct summaries and the long list of references given at the end of each chapter; these allow the reader to confirm he has understood the salient points and provide indicators to more advanced readings on the subject.

S. J. HARROLD Department of Electrical Engineering and Electronics, UMIST

Two more companies offering analogue and mixed signal IP emerged at IP99.

Analogue and mixed signal circuitry is widely regarded as being so dependent on the details of the process on which it is to be manufactured that it is difficult, if not impossible, to transfer to a customer as a product rather than as a service.

But both Pivotal Technologies and ChipIdea want to enter the market for blocks of analogue IP using internally developed tools to ease the transfer between processes.

KC Murphy, chief executive of Pivotal Technologies, claims modern CMOS processes at the same minimum line widths are very similar so porting blocks between them is easier than it appears.

According to Erlend Olson, chief technology officer of Pivotal, the company has ported an 8bit analogue to digital converter to a new process in seven days. It has seven patents pending on techniques to isolate its analogue blocks from the rest of the circuitry on a chip.

ChipIdea of Lisbon, Portugal, is also trying to enter the market for analogue and mixed signal IP. It has been servicing foundries including ES2/Atmel and Austria Mikro Systeme, with its baseband codecs for wireless applications and converter designs. But it wants to go beyond using the foundries as a channel to market and address systems customers directly.

Joao Piteira, product manager at ChipIdea, says the company has been developing hard IP targeted at particular processes, but is trying to develop an internal tool to ease the shift between processes.

Jose Epifanio da Franca, president and chief executive, says the company is working on an ESPRIT project called RAPID to develop tools which help retarget a design between different applications.

Earthquakes (e.g., Seed, 1968; Biscontin et al., 2004) or storms (e.g., DeGraff, 1994) can trigger the gravity sliding of soft sediments. Attempts to describe the mechanical links between earthquakes or storms and gravity sliding have tended to emphasize the role of fluid pressure. For example, Biscontin et al. (2004) and Hutton et al. (2004) suggest that landslides can be triggered by excess pore pressure generated by earthquake loading. Similarly, Kokusho (1999), Kokusho and Kojima (2002), and KoKusho and Fujita (2002) suggest mat gravity sliding of sediments can occur above layers of water created by liquefaction during earthquakes.

This paper demonstrates a simple, reusable, and inexpensive experiment for the triggering of gravity-driven sliding of soft sediments, which illustrates the importance of liquefaction and overpressuring. It is a development of the method presented by Peacock (2003) to illustrate the development of overpressuring in soft sediments, and uses the same materials, simply modifying the set-up of the experiment. The experiment is therefore an ideal teaching tool for illustrating a range of geological processes.

SIMPLE ANALOGUE EXPERIMENT

The set-up of the experiment is shown in Figure 1(A). Sand, mud and water are mixed together in a transparent, sealed 2-liter plastic bottle ( 121 mm wide, 82 mm thick and 200 mm high). Various types of sand have been used with success, but the mud should be course enough to settle within the space of a few hours. Lucustrine silt has been found to work better than esturine muds, while soil commonly contains too much organic matter. A few grammes of salt are added to speed up the settling of the mud, and a few milliliters of bleach are added to stop organisms from growing in the bottle. The relative quantities of the sediments do not appear to affect the results, but there should be enough sand for liquefaction to be significant, and the mud should not be 10 mm thick. The water above the settled sediments should be tens of millimeters deep. The following procedure is followed:

1. The bottle containing the sediments and water is shaken vigorously until all of the sediments are liquefied.

2. The bottle is held still for a few seconds so the sand can start to settle.

3. It is tilted so the top surface of the sand has a dip of a few degrees.

4. The bottle is left on a stable surface for several hours, until the mud has settled and the water is clear (Figure 1A). The mud must be allowed to settle undisturbed because even slight disturbances during or after settling cause premature liquefaction.

5. The sand compacts and liquefies when the bottle is gently squeezed or knocked.

A typical sequence of events illustrated in Figure 1. Water is expelled from the sand and trapped under the mud, developing a layer of water (Figure 1B). Extension fractures rapidly develop upslope, with buckles developing down slope and increasing in amplitude as sliding of the mud continues (Figure 1C). Thrusts commonly develop in the buckles to accommodate shortening, with strike-slip faults accommodating variations in movement within the mud. The water layer beneath the mud starts to escape into the overlying water, forming muddy plumes (Figure 1D). Such a sequence typically takes several seconds.

The experiment presented in Figure 1 has the following characteristics: (1) it is a closed system, in a sealed bottle; (2) liquefaction occurs as the sand is disturbed, rather than as water is pumped through the sand (e.g., Nichols et al., 1994); (3) it is simple and inexpensive; and (4) it is reusable. It is therefore ideal for demonstrating gravity sliding, especially for teaching purposes.

GRAVITY SLIDING AND THE DEVELOPMENT OF OVERPRESSURED FLUIDS

Overpressured fluids (e.g., Thomeer and Bottema, 1961; Osborne and Swarbrick, 1997) operate in various geological processes. For example, the crucial role of fluid pressure in controlling thrust faulting was recognized by Hubbert and Rubey (1959), who showed that large thrust sheets can move on gently-dipping fault surfaces only if the fluid pressure is close to, or exceeds, that of the overburden. The overpressurized fluid effectively supports the mass of the overlying thrust sheet. Evidence for overpressured fluids in thrust sheets includes sub-horizontal veins in the walls of thrusts, these having initiated as cracks caused by overpressured fluid (e.g., Sibson, 1989; Teixell et al., 2000).

The iC-NQ is a high-resolution sine-to-digital convertor with alignment functions for analogue sine/cosine sensor signals supplied, for example, from optical or magnetic position encoders.
The iC-NQ is a high-resolution sine-to-digital convertor with alignment functions for analogue sine/cosine sensor signals supplied, for example, from optical or magnetic position encoders. Adjustable settings include decimal as well as binary resolutions up to 8192 angular steps per period, which can be output as incremental A/B quadrature signal or as absolute value via the fast BiSS interface. The iC-NQ is suitable for many sensor bridges, eg magnetoresistive sensors for length or angular position sensors and for optoelectronic scanners.

Drift-free instrumentation amplifiers allow to directly connect measuring bridges to the inputs without requiring further components.

For conventional sensor signal levels the amplification can be selected in ranges between approximately 20mV and 1.5V peak-peak.

The adaptation for photocurrents can be performed via external resistors.

As a result of the fully integrated signal conditioning typical signal errors can be immediately corrected within the analogue signal path.

Matched D/A convertors, which are adjustable via the BiSS interface, are available for the calibration regarding offset, amplitude difference and phase error.

Later, the correction values, as well as all other component settings, are supplied by the external EEPROM, which was optionally reconfigured via BiSS.

The output of special test signals supports the calibration process.

The sine-to-digital conversion with guaranteed no missing codes operates with a scan frequency of 50MHz, virtually in real time, with a maximum signal transmission delay of 250ns.

Depending on the selected resolution input frequencies up to 250kHz are permitted.

An alarm message is generated if an exceeding of frequency, an invalid signal level, eg caused by wire breakage, or an EEPROM data error occurs.

The conversion appears with an adjustable resolution, which can be between 4 to 8192 angular steps for binary or between 25 to 2000 angular steps for decadic resolutions.

An output is only performed for A/B quadrature signals that can be reliably counted and processed by the subsequent system.

This is guaranteed by patented signal processing for glitch-free digital signals with hysteresis, which always keep a preselectable minimum spacing between the signal edges.

Matching the application and the environmental noise conditions, input interferences are filtered without increasing the signal transmission delay.

The incremental encoder signals are supplemented with a generated zero signal if the release is enabled at the zero input.

Position and width are selectable.

Being the first BiSS slave on the market the iC-NQ provides absolute position data with 24bit multiturn count within a few microseconds.

This is enabled by the fast bidirectional, synchronous-serial interface, BiSS, which allows up to eight axes to return triggered position data.

As an option, it is possible to configure for SSI compatibility.

With a single side +5V supply the component operates in the industrial operating temperature range of -20 to +85C.

For automotive applications a component qualification test can be performed according to the requirements.

The convertor chip is already designed for operating temperatures from -40 to +125C.

The ADS8372 is billed as the industry’s most linear 16bit successive-approximation ADC.
Texas Instruments reckons it has developed the industry’s most linear 16bit successive-approximation (SAR) analogue-to-digital convertor (ADC). The ADS8372 features 16bit no missing code performance, 600Ksample/s datarate, less than 0.75LSB (maximum) integral nonlinearity (INL) and less than 0.5 LSB (maximum) differential nonlinearity (DNL) over the entire industrial temperature range (-40 to +85C). The device is ideal for advanced, real-time applications in automated test equipment, medical imaging, optical networking, high-speed control loops and high-resolution data acquisition systems.

The ADS8372 also provides a complete solution with an on-chip reference and reference buffer, allowing breakthrough signal linearity without requiring active external components.

‘By ensuring linearity performance below 1LSB, the ADS8372 is the most accurate 16bit SAR ADC in the industry’, said Tony Chang, Strategic Marketing Engineer for TI’s precision data convertor products.

‘The ADS8372 also enables significant board space savings by integrating an internal reference and reference buffer and packaging it in a tiny 6 x 6mm QFN’.

The ADS8372 features a fully differential, pseudo-bipolar input range and includes a 16bit capacitor-based SAR ADC architecture with inherent sample and hold, 4.096V internal reference, internal reference buffer and conversion clock.

The device offers a high-speed CMOS, SPI-compatible serial interface with clock speeds up to 40MHz.

A pseudo-differential, unipolar version, the ADS8370, is also available.

Other key specifications include +/-0.2 ppm/C offset drift, 94dB signal-to-noise ratio (SNR), 120dB spurious-free dynamic range (SFDR) and 110mW power consumption at 600Ksample/s (only 15mW during nap mode and 10uW during power down).

The ADS8372 is optimised to work with TI’s high-performance TMS320 DSP platforms.

TI also has a range of high-speed amplifiers, such as the THS4031 and THS4131, which are well suited to drive the analogue inputs of the ADS8372.

The ADS8372 is a RoHS-compatible ‘green’ device in a 6 x 6mm QFN-28 package, and is available now from TI and its authorised distributors.

The ADS8372 is priced starting at $13.00 in 1000-piece quantities (suggested resale pricing).

Evaluation modules are available.

The ADS8372 is billed as the industry’s most linear 16bit successive-approximation ADC.
Texas Instruments reckons it has developed the industry’s most linear 16bit successive-approximation (SAR) analogue-to-digital convertor (ADC). The ADS8372 features 16bit no missing code performance, 600Ksample/s datarate, less than 0.75LSB (maximum) integral nonlinearity (INL) and less than 0.5 LSB (maximum) differential nonlinearity (DNL) over the entire industrial temperature range (-40 to +85C). The device is ideal for advanced, real-time applications in automated test equipment, medical imaging, optical networking, high-speed control loops and high-resolution data acquisition systems.

The ADS8372 also provides a complete solution with an on-chip reference and reference buffer, allowing breakthrough signal linearity without requiring active external components.

‘By ensuring linearity performance below 1LSB, the ADS8372 is the most accurate 16bit SAR ADC in the industry’, said Tony Chang, Strategic Marketing Engineer for TI’s precision data convertor products.

‘The ADS8372 also enables significant board space savings by integrating an internal reference and reference buffer and packaging it in a tiny 6 x 6mm QFN’.

The ADS8372 features a fully differential, pseudo-bipolar input range and includes a 16bit capacitor-based SAR ADC architecture with inherent sample and hold, 4.096V internal reference, internal reference buffer and conversion clock.

The device offers a high-speed CMOS, SPI-compatible serial interface with clock speeds up to 40MHz.

A pseudo-differential, unipolar version, the ADS8370, is also available.

Other key specifications include +/-0.2 ppm/C offset drift, 94dB signal-to-noise ratio (SNR), 120dB spurious-free dynamic range (SFDR) and 110mW power consumption at 600Ksample/s (only 15mW during nap mode and 10uW during power down).

The ADS8372 is optimised to work with TI’s high-performance TMS320 DSP platforms.

TI also has a range of high-speed amplifiers, such as the THS4031 and THS4131, which are well suited to drive the analogue inputs of the ADS8372.

The ADS8372 is a RoHS-compatible ‘green’ device in a 6 x 6mm QFN-28 package, and is available now from TI and its authorised distributors.

The ADS8372 is priced starting at $13.00 in 1000-piece quantities (suggested resale pricing).

Evaluation modules are available.

The ADS8372 is billed as the industry’s most linear 16bit successive-approximation ADC.
Texas Instruments reckons it has developed the industry’s most linear 16bit successive-approximation (SAR) analogue-to-digital convertor (ADC). The ADS8372 features 16bit no missing code performance, 600Ksample/s datarate, less than 0.75LSB (maximum) integral nonlinearity (INL) and less than 0.5 LSB (maximum) differential nonlinearity (DNL) over the entire industrial temperature range (-40 to +85C). The device is ideal for advanced, real-time applications in automated test equipment, medical imaging, optical networking, high-speed control loops and high-resolution data acquisition systems.

The ADS8372 also provides a complete solution with an on-chip reference and reference buffer, allowing breakthrough signal linearity without requiring active external components.

‘By ensuring linearity performance below 1LSB, the ADS8372 is the most accurate 16bit SAR ADC in the industry’, said Tony Chang, Strategic Marketing Engineer for TI’s precision data convertor products.

‘The ADS8372 also enables significant board space savings by integrating an internal reference and reference buffer and packaging it in a tiny 6 x 6mm QFN’.

The ADS8372 features a fully differential, pseudo-bipolar input range and includes a 16bit capacitor-based SAR ADC architecture with inherent sample and hold, 4.096V internal reference, internal reference buffer and conversion clock.

The device offers a high-speed CMOS, SPI-compatible serial interface with clock speeds up to 40MHz.

A pseudo-differential, unipolar version, the ADS8370, is also available.

Other key specifications include +/-0.2 ppm/C offset drift, 94dB signal-to-noise ratio (SNR), 120dB spurious-free dynamic range (SFDR) and 110mW power consumption at 600Ksample/s (only 15mW during nap mode and 10uW during power down).

The ADS8372 is optimised to work with TI’s high-performance TMS320 DSP platforms.

TI also has a range of high-speed amplifiers, such as the THS4031 and THS4131, which are well suited to drive the analogue inputs of the ADS8372.

The ADS8372 is a RoHS-compatible ‘green’ device in a 6 x 6mm QFN-28 package, and is available now from TI and its authorised distributors.

The ADS8372 is priced starting at $13.00 in 1000-piece quantities (suggested resale pricing).

Evaluation modules are available.

Arrow UK has been recognised by Texas Instruments (TI) for its outstanding performance.
Arrow UK has been recognised by Texas Instruments (TI) for its outstanding performance. The award was given in recognition of Arrow UK’s significant growth rate with TI’s analogue product portfolio. David Spragg, Technical Marketing Director, Arrow UK, commented: ‘This award reflects the hard work from both sales and engineering to identify analogue design-in opportunities and successfully bring them to fruition’.

‘We also appreciate the invaluable day-to-day support received from the professional team at TI’.

‘Arrow UK, with its design-in orientation and profound understanding of the local market, was especially successful in working with its customers to help them select the right high-performance analogue products for their applications’, said Dirk Rathsack, TI’s Director of European Distribution.

Arrow UK has been recognised by Texas Instruments (TI) for its outstanding performance.
Arrow UK has been recognised by Texas Instruments (TI) for its outstanding performance. The award was given in recognition of Arrow UK’s significant growth rate with TI’s analogue product portfolio. David Spragg, Technical Marketing Director, Arrow UK, commented: ‘This award reflects the hard work from both sales and engineering to identify analogue design-in opportunities and successfully bring them to fruition’.

‘We also appreciate the invaluable day-to-day support received from the professional team at TI’.

‘Arrow UK, with its design-in orientation and profound understanding of the local market, was especially successful in working with its customers to help them select the right high-performance analogue products for their applications’, said Dirk Rathsack, TI’s Director of European Distribution.

A new high-side monitor provides a compact solution for the commonproblem of interfacing an optical power monitoring device, such as a logarithmic amplifier, to the cathode-side of a PIN photodiode.
Analog Devices is introducing a high-side monitor that provides a compact solution for the common and often challenging problem of interfacing an optical power monitoring device, such as a logarithmic amplifier, to the cathode-side of a PIN photodiode. The ADL5315 mirrors the current at the cathode at a 1:1 ratio over a six-decade range, producing an output equal to the reference current monitored. Competing optical current monitors have much higher mirror ratios over narrower ranges, rendering proportional outputs at very low input currents impossible, whereas discrete solutions are more complex to implement and may require significantly more board space.

The ADL5315 offers unmatched accuracy, with linearity of only 1% over a wide operating range of six decades - from 3nA to 3mA.

In PIN photodiode modules where only the cathode is available for connection, the high linearity allows the device to be paired with linear or logarithmic amplifiers to form a complete, precise optical power measurement system.

Although the ADL5315 is optimised for use with PIN photodiodes, its operating range makes it suitable for a broad array of applications that require accurate current mirroring.

The ADL5315 features an adjustable input current limit to protect the photodiode from damage caused by excessive input current.

Its miniature single-chip 2 x 3mm package makes it well-suited for space-constrained applications, such as SFP (small form factor pluggable) optical transceiver modules.

The ADL5315 is fully specified from -40 to +85C, and operates from a single 2.7 to 8V supply.

The ADL5315 is sampling now, with volume production scheduled for October 2005.

The device is priced at $1.93 per unit in 1000-piece quantities and is packaged in a 2 x 3mm, 8-lead LFCSP.