Linear encoders can do a lot to improve factory automation. When used as secondary feedback they can greatly enhance the precision of motion control systems. They can act as a feedback device for automatic size change, and they can be used in gauging applications.
However, they can be troublesome to maintain. Most linear encoders are made from a glass strip or rod that is etched with index marks and read optically. These kinds of encoders can achieve very high accuracy…with high price points to match. However, a consistent problem in many factory automation environments is the mechanical fragility of the glass scale encoder. They can be easily broken by shock, vibration, or impact. The presence of dirt and liquids can also interfere with proper operation. Repair costs can become a problem, not to mention the cost of carrying the spare parts needed to cope with long lead times for replacements.
Depending on the resolution and accuracy class required, one alternative to these issues is the magnetic linear encoder. Today’s magnetic encoders can achieve resolution to 1 μm and accuracy to ±5 μm. Rather than index marks on glass, the scale consists of magnetic poles precisely located on a ferromagnetic strip of tape. A magnetic read head glides over the tape and outputs digital position signals. The magnetic system is much more tolerant of shock and vibration, and can tolerate most kinds of liquids and dirt. The main caveat is ferrous particles or chips; these can accumulate on the magnetic strip and cause position deviations.
Most magnetic linear encoders offer incremental signals, but a new option is absolute position over an SSI or BiSS-C serial interface. This allows the encoder to report position upon power-up, without the need for a time-wasting homing or reference run. This can be helpful in situations like a power outage, where it may not be possible to re-home the machine without damaging work in process and/or breaking tooling.
Most industrial processes do not take place in a climate-controlled laboratory or clean room environment. Real-world industrial activity generates or takes place under harsh conditions that can damage or shorten the life expectancy of equipment, especially electronic sensors.
A cross-section of industrial users was surveyed about operating conditions in their facilities. The responses revealed that plant operators are challenged by a variety of difficult environmental factors, the biggest being heat, dust/dirt/water contamination, vibration, and extreme temperature swings.
Over one-third of the industrial users surveyed reported that premature sensor failure is a problem in their operations. That is a surprisingly high percentage and something that needs to be addressed to restore lost productivity and maintain long-term competitiveness.
Many heavy industries are dependent on automated hydraulic cylinders to move and control large loads precisely. The cylinder position sensors are often subjected to damaging environmental conditions that shorten their life expectancy, leading to premature failure.
When looking at a data sheet for an inductive proximity sensor, there are usually several different specifications listed with regard to the switching distance (or operating distance). Which of these various specifications really matter to someone trying to use a prox sensor in a real-world application? How can a specifier or user decide which sensor is going to work best in their situation?
Fortunately, there is an international standard that defines sensor switching distances and spells out test methods to assure that sensor specifications from product to product and even manufacturer to manufacturer can be directly compared “apples to apples.”
This standard is IEC 60947-5-2Low voltage switchgear and controlgear – Part 5-2: Control circuit devices and switching elements – Proximity switches.
Operating (switching) distance s
In the diagram shown here, the letter “s” refers to a given sensor specimen’s actual switching distance when tested. It is defined as the distance (between the standard target and the sensing face of the proximity switch) at which a signal change is generated. For a normally open sensor, the target approaches the sensor axially, that is, the sensor approaches the active surface from the front (not the side). There are several subscripts used to describe different aspects of a sensor’s switching behavior.
Rated operating distance sn
… is the nominal switching distance of the sensor. It is simply used as a standard reference value. The rated operating distance is the best figure to use when comparing different sensor models to get an idea of their essential sensing distance capabilities.
Effective operating distance sr
…is the range of actual switching distances that any given proximity sensor will fall into when measured under specified conditions of mounting, temperature, and supply voltage. For well-designed and manufactured sensors, the sensor will be triggered between 90% and 110% of the rated operating distance. For example, various samples of a proximity sensor model with a rated operating distance (sn) of 8mm may deliver switch-on points anywhere between 7.2mm and 8.8mm.
Usable operating distance su
…takes into account the effects of the sensor’s full ambient temperature range (low to high) and variation of the supply voltage from 85% to 110% of the nominal voltage rating. The IEC standard requires the usable operating distance (su) to be between 90% and 110% of the effective operating distance (sr). For our example of a sensor with a rated operating distance (sn) of 8mm, the usable operating distance would fall between 6.5mm and 8.8mm. Pop quiz: why is the max of usable operating distance not 9.7mm (sr of 8.8mm * 110%)? Answer: the usable operating distance can always be less than but can never be greater than the maximum effective operating distance.
Assured operating distance sa
This is the distance of the target to the sensor where the sensor can be guaranteed to have turned on. If a target approaches within the assured operating distance, you can be confident that the sensor will detect it. It is 90% of sr which is in turn 90% of sn, which is in effect 81% of sn. Going back to our example of a sensor with a rated operating distance (sn) of 8mm, sa would be 81% * 8mm = 6.5mm. So in essence, sa = su(min).
Differential travel H
Now when the target recedes, at what distance will the sensor switch off? All good-quality sensors have a built-in property called hysteresis, which means that the sensor will turn off when the target is further away from the sensor than the point where it turns on. This is necessary to prevent chattering and instability when the target approaches the sensor. We want the sensor to turn on and stay on, even if the target might be vibrating as it crosses the threshold of detection. For most sensors, it is defined as ≤ 20% of the effective operating distance sr. The differential travel is added to the value of sr to define the switch-off point.
In practice, for any group of sensors, the minimum value of H would be zero and the maximum value would be sr(max) + 20% of sr(max). For our example of a sensor with a rated operating distance (sn) of 8mm, 7.2mm ≤ sr ≤ 8.8mm. So, the range of switch-off points would be 7.2mm ≤ sr+H ≤ 10.6mm. It might sound like a large range, but for any given sensor specimen the switch-off point is never greater than 20% of that particular sensor’s switch-on point.
The good news is that you don’t have to conduct sensor tests yourself or go through all of these calculations manually to determine a sensor’s performance envelope. The sensor manufacturer provides all of these useful figures pre-calculated for you in the sensor data sheet.
Learn more about the basics of the most popular automation sensor here.
Tabletop automation is a trend that is gaining momentum, especially in the fields of medical laboratory automation and 3D printing. Both of these applications demand a level of linear positioning accuracy and speed that might suggest a servomotor as a solution, but market-driven cost constraints put most servos out of financial consideration. New advances in stepper motor design, including higher torque, higher power ratings, and the availability of closed-loop operation via integrated motor encoder feedback are enabling steppers to expand their application envelope to include many tasks that formerly demanded a servo system.
Meeting the Demand for Even More Accurate, More Reliable Positioning
As tabletop automation development progresses, performance demands are increasing to the point that stepper systems may struggle to meet requirements. Fortunately, the addition of an external linear encoder for direct position feedback can enhance a stepper system to enable the expected level of reliable accuracy. An external linear encoder puts drive-mechanism non-linearity inside the control loop, meaning any deviations caused by drive component inaccuracy are automatically corrected and compensated by the overall closed-loop positioning system. In addition, the external linear encoder provides another level of assurance that the driven element has actually moved to the position indicated by the number of stepper pulses and/or the movement reported by the motor encoder. This prevents position errors due to stepper motor stalling, lost counts on the motor encoder, someone manually moving the mechanism against motor torque, or drive mechanism malfunction, i.e. broken drive belt or sheared/skipped gearing.
Incremental, Absolute, or Hybrid Encoder Signals
The position signals from the external encoder are typically incremental, meaning a digital quadrature square wave train of pulses that are counted by the controller. To find a position, the system must be “homed” to a reference position and then moved the required number of counts to reach the command position. The next move requires starting with the position at the last move and computing the differential move to the next command position. Absolute position signals, typically SSI (synchronous serial interface) provide a unique data value for each position. This position is available upon power-up…no homing movement is required and there is no need for a pulse counter. A recent innovation is the hybrid encoder, where the encoder reads absolute position from the scale, but outputs a quadrature incremental pulse train in response to position moves. The hybrid encoder (sometimes referred to as “absolute quadrature”) can be programmed to deliver a continuous burst of pulses corresponding to absolute position at power up, upon request from the controller, or both.
For more information about magnetic linear encoder systems, visit www.balluff.us.
In many types of metals production, pickling is a process that is essential to removing impurities and contaminants from the surface of the material prior to further processing, such as the application of anti-corrosion coatings.
In steel production, two common pickling solutions or pickle liquors are hydrochloric acid (HCl) and sulfuric acid (H2SO4). Both of these acids are very effective at removing rust and iron oxide scale from the steel prior to additional processing, for example galvanizing or rolling. The choice of acid depends on the processing temperature, the type of steel being processed, and environmental containment and recovery considerations. Hydrochloric acid creates corrosive fumes when heated, so it typically must be used at lower temperatures where processing times are longer. It is also more expensive to recover when spent. Sulfuric acid can be used at higher temperatures for faster processing, but it can attack the base metal more aggressively and create embrittlement due to hydrogen diffusion into the metal.
Acids can be just as tough on all of the equipment involved in the pickling lines, including sensors. When selecting sensors for use in areas involving liquid acid solutions and gaseous fumes and vapors, care must be given to the types of acids involved and to the materials used in the construction of the sensor, particularly the materials that may be in direct contact with the media.
A manufacturer of silicon steel was having issues with frequent failure of mechanical pressure sensors on the pickling line, due to the effects of severe corrosion from hydrochloric acid at 25% concentration. After determination of the root cause of these failures and evaluation of alternatives, the maintenance team selected an electronic pressure sensor with a process connection custom-made from PVDF (polyvinylidene fluoride), a VitonTM O-ring, and a ceramic (rather than standard stainless steel) pressure diaphragm. This changeover eliminated the corroded mechanical pressure sensors as an ongoing maintenance problem, increasing equipment availability and freeing up maintenance personnel to address other issues on the line.
In any continuous manufacturing process such as steel production, increased throughput is the path to higher profits through maximum utilization of fixed capital investments. In order to achieve increased throughput, more sophisticated control systems are being deployed. These systems enable ever-higher levels of automation but can present new challenges in terms of managing system reliability. Maintenance of profit margins depends on the line remaining in production with minimal unexpected downtime.
It is essential that control components, such as sensors, be selected in accordance with the rigorous demands of steel industry applications. Standard sensors intended for use in more benign manufacturing environments are often not suitable for the steel industry and may not deliver dependable service life.
When specifying sensors for steel production applications, some environmental conditions to consider include:
High temperatures exist in many areas of the steel-making process, such as the coke oven battery, blast furnace, electric arc furnace, oxygen converter, continuous casting line, and hot rolling line. Electronic components are stressed by elevated temperatures and can fail at much higher rates than they would at room temperature. Heat can affect sensors through conduction (direct transfer from the mounting), convection (circulating hot air), or radiation (line-of-sight infrared heating at a distance). The first strategy is to install sensors in ways that minimize exposure to these three thermal mechanisms. The second line of defense is to select sensors with extended temperature ratings. Many standard sensors can operate up to 185° F (85° C) but high temperature versions can operate to 212° F (100° C) or higher. Extreme temperature sensors can operate to 320° F (160° C) or even 356° F (180° C).
Don’t forget to consider the temperature rating of any quick-disconnect cables that will be used with the sensors. Many standard cable materials will melt or break down quickly at higher temperatures. Fiberglass-jacketed cables, for example, are rated to 752° F (400° C).
Shock and Vibration
Steel making involves large forces and heavy loads that generate substantial amounts of shock under normal and/or abnormal conditions. Vibration is also ever-present from motors, rollers, and moving materials. As with heat, look for sensors with enhanced specifications for shock and vibration. For sensors with fixed mountings, look for shock ratings of at least 30 G. For sensors mounted to equipment that is moving (for example, position sensors on hydraulic cylinders), consider sensors with shock ratings of 100 to 150 G. For vibration, the statement of specifications can vary. For example, it may be stated as a frequency and amplitude, such as 55 Hz @ 1 mm or as acceleration over a frequency range, such as 20 G from 10…2000 Hz.
Don’t forget that the quick-disconnect connector can sometimes be a vulnerability under severe shock. Combat broken connectors with so-called “pigtail” or “inline” connectors that have a flexible cable coming out of the sensor that goes to a quick-disconnect a few inches or feet away.
The best way to protect sensors from mechanical impact is to install them in protective mounting brackets (a.k.a. “bunker blocks”) or to provide heavy-duty covers over them. When direct contact with the sensor cannot be avoided, choose sensors specifically designed to handle impact.
Another strategy is to use remote sensor actuation to detect objects without making physical contact with the sensor itself.
Corrosion and Liquid Ingress
In areas with water spray and steam, such as the scale cracker on a hot strip line, corrosion and liquid ingress can lead to sensor failure. Look for stainless steel construction (aluminum can corrode) and enhanced ingress protection ratings such as IP68 or IP69K.
When All Else Fails…Rapid Replacement
If and when a sensor failure inevitably occurs, choose products and accessories that can minimize the downtime by speeding up the time required for replacement.
Strategies include quick-change sensor mounts, rapid-replacement sensor modules, and redundant sensor outputs.
In the case of redundant sensor outputs, if the primary output fails, the system can continue to operate from the secondary or even tertiary output.
Magnetostriction is a property of ferromagnetic (iron-based, magnetizeable) materials that causes them to change their shape or dimensions in the presence of a magnetic field. In addition to numerous other practical uses, this magnetostrictive effect is ideally suited for use in industrial linear position measurement sensors. Magnetostrictive linear position sensors use an iron-alloy sensing element, typically called a waveguide. Referring to the diagram at right, the waveguide (1) is housed inside a pressure-rated stainless steel tube or in an aluminum extrusion. The position magnet (2) is attached to the moving part of the machine, or the piston of a hydraulic or pneumatic cylinder. Measurements are initiated by applying a short-duration electrical pulse to a conductor (3) attached to the waveguide. The current creates a magnetic field (4) along the waveguide.
The magnetic field from the position magnet interacts with the generated magnetic field, inducing a torsional mechanical strain on the waveguide. When the current pulse stops, the strain is released, causing a mechanical pulse to propagate along the waveguide. This mechanical pulse travels at a constant speed, and is detected at the signal converter (5).
The time between the initial electrical pulse and the received mechanical pulse accurately represents the absolute position of the position magnet and, ultimately position of the machine or hydraulic cylinder. The position of the magnet along the waveguide is calculated by very accurately timing the interval between the initial current pulse, also known as the Interrogation Pulse, and the detection of the mechanical return pulse.
Our modern technological society owes a lot to the scientific work and inspiration of a 17th-century French mathematician and physicist, Blaise Pascal. Pascal was a pioneer in the fields of hydrostatics and hydrodynamics, which deal with the subject of fluid mechanics under pressure.
One of the most important physical principles he defined is known today as Pascal’s Law:
“A change in pressure at any point in an enclosed fluid at rest is transmitted undiminished to all points in the fluid.”
It is this characteristic of fluids held in containment that allows force applied to a fluid in one location to be delivered to another remote location. A well-known example would be the hydraulic brake system in a car. Mechanical pressure from the driver’s foot is transferred to the brake fluid through a master cylinder. This pressure is then instantly communicated to braking cylinders located at each wheel, causing them to apply mechanical force to press friction pads against a brake drum or rotor, thus slowing or stopping the vehicle.
In the industrial world, the compact yet incredible power of hydraulic cylinders is a constant source of awe and amazement. Through the magic of fluid power leverage via Pascal’s Law, hydraulic cylinders are capable of generating tremendous lifting forces to move massively heavy structures.
In order for such great force to be harnessed to do useful work, it must be kept fully under control. Force that is out of control is either useless or destructive. When it comes to controlling the movement of a powerful hydraulic cylinder, the piston/ram position must be continually monitored in near-real-time.
The most popular device for measuring cylinder position is called a Magnetostrictive Linear Position Sensor. Sometimes these position sensors are called LDTs (Linear Displacement Transducer) or MDTs (Magnetostrictive Displacement Transducer). All of these terms refer to the same type of devices.
To get an idea of the power and control that is feasible with modern hydraulic cylinders and integrated cylinder position sensors, have a look at this amazing video from ALE Heavylift. The topside of a giant offshore oil platform was jacked up 131 ft (40 m) and then skidded horizontally a distance of 295 ft (90 m) to place it on top of its supports. Imagine the incredible synchronization of speed, position, and operational sequencing needed to safely lift and place such a massive structure.
For more information about magnetostrictive linear position sensors for hydraulic cylinders, visit the Balluff website at www.balluff.us.
When it comes to selecting the most appropriate position detection technology for an absolute rotary encoder application, it’s helpful to consider the general advantages and potential disadvantages of the two most common approaches: optical and magnetic.
Internally, absolute optical rotary encoders are comprised of:
An LED light source
A rotating coded disk to modulate the light beam from the LED
An array of photodetectors to convert the impulses of light into electrical signals.
The spinning code disk contains a series of concentric tracks that each represent one bit of resolution, and each track is associated with a separate photodetector.
Among optical encoders, there are two main variations: optical mask and optical phased-array. Optical mask encoders are the more straightforward implementation. A grated mask featuring slots of the same size as the slots in the optical disk is placed on top of the photodetectors to prevent the light spilling over from one channel to another. The chief advantage of the optical mask encoder rests in the ability of the encoder manufacturer to offer a variety of resolutions with the same photodetector array simply by changing the optical code disk and associated mask. On the downside, very high-resolution optical mask encoders require a very small air gap between the mask and the disk of about 0.001…0.003″ (25…75μm). Reliably maintaining such a tight gap requires tight manufacturing and assembly tolerances, and can lead to problems in severe shock and vibration environments.
As a result of the limitations of optical mask encoders, phased-array encoders were developed. Rather than relying on only a single detector for each channel, an ASIC (application-specific integrated circuit) provides an array of very small photodetectors for each channel. The responses of these multiple detectors are averaged, producing a more robust detection signal that is less susceptible to variation than a single detector. This additional signal robustness can be used to relax mechanical construction and assembly constraints such as disk flatness, eccentricity, and misalignment. The end result is a wider air gap tolerance for phased-array encoders compared to the optical mask types.
Both optical mask and phased-array detection schemes offer similar application advantages and disadvantages. They are immune to intense magnetic fields found around MRI machines or DC injection braking of AC induction motors. Due to the wider gap between disk and detectors, phased-array encoders are more tolerant of shock and vibration.
Regardless whether optical mask or phased-array detection is employed, both variations are rather susceptible to environmental contamination. Particulates such as dirt, dust, or powders and liquids like water or oil can block or attenuate the optical signals, leading to output errors. Another environmental consideration is that elevated temperatures and temperature variations can accelerate LED aging, leading to reduced light output and less reliable signal detection over time.
Absolute optical encoders are typically available with resolutions ranging from 10-bit (1024 pulses / 360°) to 22-bit (4,194,304 pulses / 360°).
Optical Encoder Disks
There are three popular construction methods for optical disks, each having certain advantages or disadvantages:
Glass + Metal Film
Very flat, allowing for tighter air gap and higher resolution
Fragile; can shatter when exposed to high shock or severe vibration
More tolerant of high shock and vibration
Higher resolutions not feasible due to weakening of the disk caused by necessarily large number of slots
More robust than glass + metal
Susceptible to sag and flutter, requiring a higher air gap that limits resolution
Absolute magnetic rotary encoders are comprised of only two components:
An ASIC (application-specific integrated circuit) with integrated precision magnetic sensors
A magnetically-coded rotating disk made of rubber ferrite on a metal carrier substrate
The magnetic disk employs a coding scheme called the Nonius principle, consisting of two concentric, adjacent tracks of alternating north and south magnetic poles. The number of poles on each track differs, typically by one pole. For example, the outer track may have 32 poles and the inner track 31. Going around the disk, there is a continuing shift of pole alignment between the inner and outer track. At any given position around the disk, the offset angle between inner and outer poles is unique.
Two magnetic field sensors inside the ASIC each produce a sinusoidal signal in response to the north and south poles as they traverse over them. The phase shift between these two signals is unique for every position around the disk. Digital electronics convert this analog phase shift into a serial digital data value corresponding to the absolute rotary position of the disk around 360° of rotation. A great advantage of magnetic encoders is that the maximum gap between the sensing ASIC and the magnetic disk surface is larger than for optical mask encoders. A typical specification for a magnetic encoder gap would be 0.012″ ±0.008″ (0.33mm ±0.2mm), compared to an optical mask encoder requiring a gap of about 0.002″ ±0.001″ (50 μm ±25μm).
Magnetic encoders are extremely robust. Virtually immune to shock and vibration, they are also impervious to many kinds of particulates and liquid contaminants, including non-magnetic (non-ferrous) metal shavings and powders. This ability to tolerate contamination largely reduces or eliminates the need for costly sealed enclosures. The primary caveats when applying a magnetic encoder are the presence of very strong magnetic fields that could disrupt the encoder’s operation and the presence of ferrous particles or dust that would be attracted to the magnetic surface, where they would potentially cause distortion of the magnetic poles.
Although magnetic encoders don’t currently offer the highest levels of resolution available with some optical encoders, they do offer more than enough resolution for a wide range of applications. Absolute magnetic encoders are available with resolutions ranging from 12-bit (4096 pulses / 360°) to 17-bit (131,072 pulses / 360°).
Comparison of Optical and Magnetic Absolute Encoder Operating Principles
We have all gotten that dreaded phone call or email…the customer received their order, but there was a significant problem:
Too lose…or too tight
Incomplete processing, e.g. missing threads
Something is damaged
Missing fluids or fluids at wrong level
…and so on
Assuming that we have reliable suppliers delivering quality parts that meet the required specifications…everything else that can (and often does) go wrong happens inside our own facilities. That means that solving the issues is our responsibility, but it also means that the solutions are completely under our control.
During the initial quality response meetings, at some point the subjects of “better worker training” and “more attention to detail and self-inspection” may come up. They are valid subjects that need to be addressed, but let’s face it: not every manufacturing and assembly problem can be solved by increased worker vigilance and dedication to workmanship. Nor, for that matter, is there the luxury of time or capacity for each worker to spend the extra time needed to ensure zero defects through inspection.
It is often more effective to eliminate errors at their source before they occur, so that further human intervention isn’t required or expected.
Some things to look for when searching for manufacturing trouble spots:
Are all fasteners present and properly tightened, in the proper torque sequence
Correct machine setup: is the right tool or fixture in place for the product being produced?
Manual data entry: does the process rely on human accuracy to input machine or product data?
Incorrect part: is it simply too hard to determine small differences by visual means alone?
Sequencing error: were the parts correct but came together in the wrong sequence?
Mislabeled component: would the operator realize that part is wrong if it was labeled incorrectly in the first place? Sometimes where the error has impact and where it actually occurred are in two different places.
Part not seated correctly: is everything is correct, but sometimes the part doesn’t sit properly in the assembly fixture?
Critical fluids: is the right fluid installed? Is it filled to the proper level?
Once the trouble spots have been identified, the next step is to implement a detection and/or prevention strategy. More information on the error proofing process is available on the Balluff website at www.balluff.us/errorproofing