Automation and Control Systems--Design and Construction of Instruments (part 1)



Home | Forum | DAQ Fundamentals | DAQ Hardware | DAQ Software

Input Devices
| Data Loggers + Recorders | Books | Links + Resources


AMAZON multi-meters discounts AMAZON oscilloscope discounts


1. Introduction

The purpose of this section is to give an insight into the types of components and construction used in commercial instrumentation.

To the designer, the technology in instrument Technology depends on the availability of components and processes appropriate to his task. Being aware of what is possible is an important function of the designer, especially with the rapidity with which techniques now change. New materials such as ceramics and polymers have become increasingly important, as have semicustom (ASiCs) and large-scale integrated (LSI and VLSI) circuits. The need for low-cost automatic manufacture is having the greatest impact on design techniques, demanding fewer components and suit able geometries. Low volume instruments and one-offs are now commonly constructed in software, using "virtual instrumentation" graphical user interfaces such as Labview and Labtech, or "industrial strength" HMi programs such as Wonderware or intellution.

The distinction between computer and instrument has become blurred, with many instruments offering a wide range of facilities and great flexibility. Smart sensors, which interface directly to a computer, and Fieldbus interconnectivity have shifted the emphasis to software and mechanical aspects.

Historical practice, convention, and the emergence of standards also contribute significantly to the subject. Standards, especially, benefit the designer and the user, and have made the task of the author and the reader somewhat simpler.

Commercial instruments exist because there is a market, and so details of their design and construction can only be understood in terms of a combination of commercial as well as technical reasons. A short section describes these trade offs as a backdrop to the more technical information.

2. Instrument design

2.1 The Designer's Viewpoint

Many of the design features found in instruments are not obviously of direct benefit to the user. These can best be understood by also considering the designer's viewpoint.

The instrument designer's task is to find the best com promise between cost and benefit to the users, especially when competition is fierce. For a typical medium-volume instrument, its cost as a percentage of selling price is distributed as follows:

Purchase cost [ Design cost, Manufacturing cost, Selling cos, Other overheads, Profit Z]

Operating/maintenance cost may amount to 10 percent per annum. Benefits to the user can come from many features, for example:

1. Accuracy

2. Speed

3. Multipurpose

4. Flexibility

5. Reliability

6. Integrity

7. Maintainability

8. Convenience

Fashion, as well as function, is very important, since a smart, pleasing, and professional appearance is often essential when selling instruments on a commercial basis.

For a particular product the unit cost can be reduced with higher volume production, and greater sales can be achieved with a lower selling price. The latter is called its "market elasticity." Since the manufacturer's objective is to maximize return on investment, the combination of selling price and volume which yields the greatest profit is chosen.

2.2 Marketing

The designer is generally subordinate to marketing considerations, and consequently, these play a major role in deter mining the design of an instrument and the method of its manufacture. A project will only go ahead if the anticipated return on investment is sufficiently high and commensurate with the perceived level of risk. It is interesting to note that design accounts for a significant proportion of the total costs and, by its nature, involves a high degree of risk.

With the rapid developments in technology, product life times are being reduced to less than three years, and market elasticity is difficult to judge. In some markets, especially in test and measurement systems, product lifetime has been reduced to one single production run. In this way design is becoming more of a process of evolution, continuously responding to changes in market conditions. Initially, there fore, the designer will tend to err on the cautious side, and make cost reductions as volumes increase.

The anticipated volume is a major consideration and calls for widely differing techniques in manufacture, depending on whether it is a low-, medium-, or high-volume product.

2.3 Special Instruments

Most instrumentation users configure systems from low cost standard components with simple interfaces, occasionally the need arises for a special component or system, it is preferable to modify a standard component wherever possible, since complete redesigns often take longer than anticipated, with an uncertain outcome.

Special systems also have to be tested and understood in order to achieve the necessary level of confidence. Maintenance adds to the extra cost, with the need for documentation and test equipment making specials very expensive.

This principle extends to software as well as hardware.


TABLE 1 Electronic components


FIG. 1 Printed electronic circuits. (a) Printed circuit board (PCB); (b) traditional axial component; (c) through-hole plating (close-up); (d) surface-mounted assemblies.

3. Elements of construction

3.1 Electronic components and Printed circuits

Electronic circuitry now forms the basis of most modern measuring instruments. A wide range of electronic components are now available, from the simple resistor to complete data acquisition subsystems. TABLE 1 lists some of the more commonly used types.

Computer-aided design makes it possible to design complete systems on silicon using standard cells or by providing the interconnection pattern for a standard chip which has an array of basic components. This offers reductions in size and cost and improved design security.

The most common method of mounting and interconnecting electronic components at one time was by double sided, through-hole plated fiberglass printed circuit board (PCB) (FIG. 1 (a)). Component leads or pins are pushed through holes and soldered to tinned copper pads (FIG. 1 (b)). This secures the component and provides the connections. The solder joint is thus the most commonly used component and probably the most troublesome. In the past eight years, new techniques for surface-mount assembly have made through-hole circuit boards obsolete for any thing but the most low-volume production.

Tinning the copper pads with solder stops corrosion and makes soldering easier. Fluxes reduce oxidation and surface tension, but a temperature-controlled soldering iron is indispensable. Large-volume soldering can be done by a wave soldering machine, where the circuit board is passed over a standing wave of molten solder. Components have to be on one side only, although this is also usually the case with manual soldering.

The often complicated routing of connections between components is made easier by having two layers of printed "tracks," one on each surface, permitting crossovers. Connections between the top and bottom conductor layers are provided by plated-through "via" holes. it is generally considered bad practice to use the holes in which components are mounted as via holes, because of the possibility of damaging the connection when components are replaced. The through hole plating (see FIG. 1(c)) provides extra support for small pads, reducing the risk of them peeling off during soldering. They do, however, make component removal more difficult, often requiring the destruction of the component so that its leads can be removed individually. The more expensive components are therefore generally provided with sockets, which also makes testing and servicing much simpler.

The PCB is completed by the addition of a solder mask and a printed (silk-screened) component identification layer.

The solder mask is to prevent solder bridges between adjacent tracks and pads, especially when using an automatic soldering technique such as wave soldering. The component identification layer helps assembly, testing and servicing.

For very simple low-density circuits a single-sided PCB is often used, since manufacturing cost is lower. The pads and tracks must be larger, however, since without through-hole plating they tend to lift off more easily when soldering.

For very high-density circuits, especially digital, multilayer PCBs are used, as many as nine layers of printed circuits being laminated together with through-hole plating providing the interconnections.

Most electronic components have evolved to be suit able for this type of construction, along with machines for automatic handling and insertion. The humble resistor is an interesting example; this was originally designed for wiring between posts or tag-strips in valve circuits. Currently, they are supplied on long ribbons, and machines or hand tools are used for bending and cropping the leads ready for insertion (see FIG. 1(b)).

Important principles relevant to the layout of circuits are also discussed in section 35.

3.2 Surface-mounted Assemblies

As noted earlier, surface-mounting took over as the design of choice in the 1990s. Sometimes, the traditional fiberglass rigid board itself has been replaced by a flexible sheet of plastic with the circuits printed on it. Semiconductors, chip resistors and chip capacitors are available in very small out line packages, and are easier to handle with automatic placement machines.

Surface mounting eliminates the difficult problem of automatic insertion and, in most cases, the costly drilling process as well. Slightly higher densities can be achieved by using a ceramic substrate instead of fiberglass (FIG. 1(d)). Conductors of palladium silver, insulators, and resistive inks are silk-screened and baked onto the substrate to pro vide connections, crossovers, and some of the components.

These techniques have been developed from the older "chip and wire" hybrid thick film integrated circuit technique, used mainly in high-density military applications. In both cases, reflow soldering techniques are used due to the small size.

Here, the solder is applied as a paste and silk-screened onto the surface bonding pads. The component is then placed on its pads and the solder made to reflow by application of a short burst of heat which is not enough to damage the component. The heat can be applied by placing the substrate onto a hot vapor which then condenses at a precise temperature above the melting point of the solder. More simply, the substrate can be placed on a temperature-controlled hot plate or passed under a strip of hot air or radiant heat.

The technique is therefore very cost effective in high volumes, and with the increasing sophistication of silicon circuits results in "smart sensors" where the circuitry may be printed onto any flat surface.


FIG. 2 Ribbon cable interconnection. (a) Ribbon cable assembly; (b) ribbon cable cross-section; (c) insulation displacement terminator; (d) dual in-line header.

3.2.1 Circuit Board Replacement

When deciding servicing policy it should be realized that replacing a whole circuit board is often more cost effective than trying to trace a faulty component or connection. To this end, PCBs can be mounted for easy access and provided with a connector or connectors for rapid removal. The faulty circuit board can then be thrown away or returned to the supplier for repair.

3.3 Interconnections

There are many ways to provide the interconnection between circuit boards and the rest of the instrument, of which the most common are described below.

Connectors are used to facilitate rapid making and breaking of these connections and simplify assembly test and servicing. Conventional wiring looms are still used because of their flexibility and because they can be designed for complicated routing and branching requirements. Termination of the wires can be by soldering, crimp, or wire wrap onto connector or circuit board pins. This, however, is a labor-intensive technique and is prone to wiring errors.

Looms are given mechanical strength by lacing or sleeving wires as a tight bunch and anchoring to the chassis with cable ties.

Ribbon cable and insulation displacement connectors are now replacing conventional looms in many applications. As many as sixty connections can be made with one simple loom with very low labor costs. Wiring errors are eliminated since the routing is fixed at the design stage (see FIG. 2).

Connectors are very useful for isolating or removing a subassembly conveniently. They are, however, somewhat expensive and a common source of unreliability.

Another technique, which is used in demanding applications where space is at a premium, is the "flexy" circuit.

Printed circuit boards are laminated with a thin, flexible sheet of Kapton which carries conductors. The connections are permanent, but the whole assembly can be folded up to fit into a limited space.

It is inappropriate to list here the many types of connectors. The connector manufacturers issue catalogs full of different types, and these are readily available.

3.4 Materials

A considerable variety of materials are available to the instrument designer, and new ones are being developed with special or improved characteristics, including polymers and superstrong ceramics. These materials can be bought in various forms, including sheet, block, rod, and tube, and processed in a variety of ways.

3.4.1 Metals

Metals are usually used for strength and low cost as structural members. Aluminum for low weight and steel are the most common. Metals are also suitable for machining precise shapes to tight tolerances.

Stainless steels are used to resist corrosion, and precious metal in thin layers helps to maintain clean electrical contacts. Metals are good conductors and provide electrical screening as well as support. Mumetal and radiometal have high permeabilities and are used as very effective magnetic screens or in magnetic components. Some alloys-notably beryllium-copper-have very good spring characteristics, improved by annealing, and this is used to convert force into displacement in load cells and pressure transducers. Springs made of mnemonic keep their properties at high temperatures, which is important in some transducer applications.

The precise thermal coefficient of the expansion of metals makes it possible to produce compensating designs, using different metals or alloys, and so maintain critical distances independent of temperature. Invar has the lowest coefficient of expansion at less than 1 ppm per K over a useful range, but it is difficult to machine precisely.

Metals can be processed to change their characteristics as well as their shape; some can be hardened after machining and ground or honed to a very accurate and smooth finish, as found in bearings.

Metal components can be annealed, i.e., taken to a high temperature, in order to reduce internal stresses caused in the manufacture of the material and machining. Heat treatments can also improve stability, strength, spring quality, magnetic permeability, or hardness.

3.4.2 Ceramics

For very high temperatures, ceramics are used as electrical and heat insulators or conductors (e.g., silicon carbide). The latest ceramics (e.g., zirconia, sialon, silicon nitride, and silicon carbide) exhibit very high strength, hardness, and stability even at temperatures over 1,000ºC. Processes for shaping them include slip casting, hot isostatic pressing (HiP), green machining, flame spraying, and grinding to finished size.

Being hard, their grinding is best done by diamond or cubic boron nitride (CBN) tools. Alumina is widely used, despite being brittle, and many standard mechanical or electrical components are available.

Glass-loaded machinable ceramics are very convenient, having very similar properties to alumina, but are restricted to lower temperatures (less than 500 ºC). Special components can be made to accurate final size with conventional machining and tungsten tools.

Other compounds based on silicon include sapphires, quartz, glasses, artificial granite, and the pure crystalline or amorphous substance. These have well behaved and known properties (e.g., thermal expansion coefficient, conductivity and refractive index), which can be finely adjusted by adding different ingredients. The manufacture of electronic circuitry, with photolithography, chemical doping, and milling, rep resents the ultimate in materials technology. Many of these techniques are applicable to instrument manufacture, and the gap between sensor and circuitry is narrowing-for example, in chemfets, in which a reversible chemical reaction produces a chemical potential that is coupled to one or more field-effect transistors. These transistors give amplification and possibly conversion to digital form before transmission to an indicator instrument with resulting higher integrity.

3.4.3 Plastics and Polymers

Low-cost, lightweight, and good insulating properties make plastics and polymers popular choices for standard mechanical components and enclosures. They can be molded into elaborate shapes and given a pleasing appearance at very low cost in high volumes. PVC, PTFE, polyethylene, poly propylene, polycarbonates, and nylon are widely used and available as a range of composites, strengthened with fibers or other ingredients to achieve the desired properties. More recently, carbon composites and Kevlar have exhibited very high strength-to-weight ratio, useful for structural members. Carbon fiber is also very stable, making it suitable for dimensional calibration standards. Kapton and polyamides are used at higher temperatures and radiation levels.

A biodegradable plastic, poly 3-hydroxy-buty-rate, or PHB, is also available which can be controlled for operating life. Manufactured by cloned bacteria, this material represents one of many new materials emerging from advances in biotechnology.

More exotic materials are used for special applications, and a few examples are:

1. Mumetal: very high magnetic permeability.

2. PVdF: polyvinylidene fluoride, piezoelectric effect.

3. Samarium/cobalt: very high magnetic remanence (fixed magnet).

4. Sapphire: very high thermal conductivity.

5. Ferrites: very stable magnetic permeability, wide range available.

3.4.4 Epoxy Resins

Two-part epoxy resins can be used as adhesives, as potting material, and as paint. Parameters such as viscosity, setting time, set hardness, and color can be controlled. Most have good insulating properties, although conducting epoxies exist, and all are mechanically strong, some up to 300ºC. The resin features in the important structure material: epoxy bonded fiberglass. Delicate assemblies can be ruggedized or passivated by a prophylactic layer of resin, which also improves design security.

Epoxy resin can be applied to a component and machined to size when cured. It can allow construction of an insulating joint with precisely controlled dimensions. Generally speaking, the thinner the glue layer, the stronger and more stable the joint.

3.4.5 Paints and Finishes

The appearance of an instrument is enhanced by the judicious use of texture and color in combination with its controls and displays. A wide range of British Standard coordinated colors are available, allowing consistent results. In the United States, the Pantone color chart is usually used, and colors are generally matched to a PMS (Pantone Matching System) color. For example, PMS 720 is a commonly used front panel color, in a royal blue shade.

Anodized or brushed aluminum panels have been popular for many years, although the trend is now back toward painted or plastic panels with more exotic colors. Nearly all materials, including plastic, can be spray-painted by using suitable preparation and curing. Matte, gloss, and a variety of textures are available.

Despite its age, silk-screen printing is used widely for lettering, diagrams, and logos, especially on front panels.

Photosensitive plastic films, in one or a mixture of colors, are used for stick-on labels or as complete front panels with an LED display filter. The latter are often used in conjunction with laminated pressure pad-switches to provide a rugged, easy-to-clean, splash-proof control panel.

3.5 Mechanical manufacturing Processes

Materials can be processed in many ways to produce the required component. The methods chosen depend on the type of material, the volume required, and the type of shape and dimensional accuracy.

3.5.1 Bending and Punching

Low-cost sheet metal or plastic can be bent or pressed into the required shape and holes punched with standard or special tools (FIG. 3). Simple bending machines and a fly press cover most requirements, although hard tooling is more cost effective in large volumes. Most plastics are thermosetting and require heating, but metals are normally worked cold. Dimensional accuracy is typically not better than 0.5 mm.

3.5.2 Drilling and Milling

Most materials can be machined, although glass (including fiberglass), ceramics, and some metals require specially hardened tools. The hand or pillar drill is the simplest tool, and high accuracy can be achieved by using a jig to hold the work-piece and guide the rotating bit.

A milling machine is more complex, where the workpiece can be moved precisely relative to the rotating tool. Drills, reamers, cutters, and slotting saws are used to create complex and accurate shapes. Tolerances of 1 nm can be achieved.

3.5.3 Turning

Rotating the workpiece against a tool is a method for turning long bars of material into large numbers of components at low unit cost. High accuracies can be achieved for internal and external diameters and length, typically 1 nm, making cylindrical components a popular choice in all branches of engineering.

A fully automatic machining center and tool changer and component handler can produce vast numbers of precise components of many different types under computer control.

3.5.4 Grinding and Honing

With grinding, a hard stone is used to remove a small but controlled amount of material. When grinding, both tool and component are moved to produce accurate geometries including relative concentricity and straightness (e.g., parallel) but with a poor surface finish. Precise flats, cylinders, cones, and spherics are possible. The material must be fairly hard to get the best results and is usually metal or ceramic.

Honing requires a finer stone and produces a much better surface finish and potentially very high accuracy (0.1 nm). Relative accuracy (e.g., concentricity between outside and inside diameters) is not controllable, and so honing is usually preceded by grinding or precise turning.


FIG. 5 Examples of casting and molding. (a) Molded enclosure; (b) cast fixing.


FIG. 6 Mechanical components.

3.5.5 Lapping

A fine sludge of abrasive is rubbed onto the work-piece surface to achieve ultra-high accuracy, better than 10 nm if the metal is very hard. In principle, any shape can be lapped, but optical surfaces such as flats and spherics are most common, since these can be checked by sensitive optical methods.

3.5.6 Chemical and Electrochemical Milling

Metal can be removed or deposited by chemical and electrochemical reactions. Surfaces can be selectively treated through masks. Complex shapes can be etched from sheet material of most kinds of metal using photolithographic techniques. FIG. 3(b) shows an example where accuracies of 0.1 mm are achieved.

Gold, tin, copper, and chromium can be deposited for printed circuit board manufacture or servicing of bearing components. Chemical etching of mechanical structures into silicon in combination with electronic circuitry is a process currently under development.

3.5.7 Extruding

In extruding, the material, in a plastic state and usually at a high temperature, is pushed through an orifice with the desired shape. Complex cross-sections can be achieved, and a wide range of standard items are available, cut to length.

Extruded components are used for structural members, heat sinks, and enclosures (FIG. 4). Initial tooling is, however, expensive for non-standard sections.


FIG. 3 Sheet metal. (a) Bent and drilled or punched; (b) chemical milling.


FIG. 4 Extrusion. (a) Structural member; (b) heat sink.

3.5.8 Casting and Molding

Casting, like molding, makes the component from the liquid or plastic phase but results in the destruction of the mold.

It usually refers to components made of metals such as aluminum alloys and sand casts made from a pattern. Very elaborate forms can be made, but further machining is required for accuracies better than 0.5 mm. Examples are shown in FIG. 5. Plastics are molded in a variety of ways, and the mold can be used many times. Vacuum forming and injection molding are used to achieve very low unit cost, but tooling costs are high. Rotational low-pressure molding (rotomolding) is often used for low-volume enclosure runs.

3.5.9 Adhesives

Adhesive technology is advancing at a considerable rate, finding increasing use in instrument construction. Thin layers of adhesive can be stable and strong and provide electrical conduction or insulation. Almost any material can be glued, although high-temperature curing is still required for some applications. Metal components can be recovered by disintegration of the adhesive at high temperatures. Two part adhesives are usually best for increased shelf life.

Jigs can be used for high dimensional accuracies, and automatic dispensing for high volume and low-cost assembly.

3.6 Functional components

A wide range of devices is available, including bearings, couplings, gears, and springs. FIG. 6 shows the main types of components used and their principal characteristics.

3.6.1 Bearings

Bearings are used when a controlled movement, either linear or rotary, is required. The simplest bearing consists of rubbing surfaces, prismatic for linear, cylindrical for rotation, and spherical for universal movement. Soft materials such as copper and bronze and PTFE are used for reduced friction, and high precision can be achieved. Liquid or solid lubricants are sometimes used, including thin deposits of PTFE, graphite, and organic and mineral oils. Friction can be further reduced by introducing a gap and rolling elements between the surfaces. The hardened steel balls or cylinders are held in cages or a recirculating mechanism. Roller bearings can be precise, low friction, relatively immune to contamination, and capable of taking large loads.

The most precise bearing is the air bearing. A thin cushion of pressurized air is maintained between the bearing surfaces, considerably reducing the friction and giving a position governed by the average surface geometry. Accuracies of 0.01 nm are possible, but a source of clean, dry, pressurized air is required. Magnetic bearings maintain an air gap and have low friction but cannot tolerate side loads.

With bearings have evolved seals to eliminate contamination. For limited movement, elastic balloons of rubber or metal provide complete and possibly hermetic sealing.

Seals made of low-friction polymer composites exclude larger particles, and magnetic liquid lubricant can be trapped between magnets, providing an excellent low-friction seal for unlimited linear or rotary movement.

3.6.2 Couplings

It is occasionally necessary to couple the movement of two bearings, which creates problems of clashing. This can be overcome by using a flexible coupling which is stiff in the required direction and compliant to misalignment of the bearings. Couplings commonly used include:

1. Spring wire or filaments.

2. Bellows.

3. Double hinges.

Each type is suitable for different combinations of side load, misalignment, and torsional stiffness.

3.6.3 Springs

Springs are used to produce a controlled amount of force (e.g., for preloaded bearings, force/pressure transducers or fixings). They can take the form of a diaphragm, helix, crinkled washer, or shaped flat sheet leaf spring. A thin circular disc with chemically milled Archimedes spinal slots is an increasingly used example of the latter. A pair of these can produce an accurate linear motion with good sideways stiffness and controllable spring constant.

cont. to part 2 >>

PREV. | NEXT

Related Articles -- Top of Page -- Home

Updated: Saturday, February 4, 2017 14:04 PST