Temposonics. Magnetostrictive Linear-Position Sensors. Multi-Position Sensing with Magnetostrictive Sensors. Technical Paper. Sensing Electronics

Temposonics® R Magnetostrictive Linear-Position Sensors SENSORS Multi-Position Sensing with Magnetostrictive Sensors 551077 D Technical Paper Nul...
Author: Calvin Thornton
0 downloads 0 Views 425KB Size
Temposonics® R

Magnetostrictive Linear-Position Sensors

SENSORS

Multi-Position Sensing with Magnetostrictive Sensors

551077 D

Technical Paper Null Zone

Electrical Stroke Length

Dead Zone

Sensing Electronics Pressure Isolation Tube (aka Sensing “Rod”)

Sensing Magnet Interrogation Signal T

Return Signal

~ 1 microsecond

Time Clock Trigger Level

Today the need for factory automation is greater than ever and position sensor suppliers are seeing a higher demand for continuous linear position feedback. In the past, either discrete or no sensing (open loop control) was sufficient for the control process. Now, with the demand for more process flexibility and faster cycle performance coupled with cost reductions in servo actuation and continuous feedback sensors, the demand for such systems has never been greater. Still the cost of these new systems is a concern for designers and integrators who have traditionally used discrete feedback devices and lower cost actuation and control hardware. Therefore, it is prudent to take every opportunity to reduce the overall system and per axis cost wherever possible through the selection of the optimal combination of components. One example of this is the use of an inherent capability of magnetostrictive linear-position sensors to provide multi-axis feedback along the same plane of motion. One industry’s evolution in feedback Fifteen years ago, many injection molding machines still used proximity sensors to detect and help control motion of the injection, carriage, mold, and ejector motion during an injection cycle. Adjustment of the cycle was accomplished by adjusting the position of the proximity sensors, which required a lot of setup time and often resulted in marginal performance. Today, where more precision, faster cycle times and quicker setups are critical to the machine’s value proposition, linear-position sensors are used to replace the old “array” of proximity sensors. Magnetostrictive sensors are particularly well suited for this application – especially All specifications are subject to change. Please contact MTS for specifications that are critical to your needs.

for larger, longer stroking machines – because they are capable of stroke ranges greater than 5000 mm, they can also provide resolution as small as 1 micron.

Plastic Hopper

Mold

Another important, but little known characteristic of magentostrictive sensors is their ability to generate “simultaneous” (explained in the next section) multi-position outputs along a given sensing element. Machine builders that have recognized the benefits of this capability have also realized that many of their motion control axes operate in parallel and in reasonably close proximity. For example, in most cases, the injection carriage and injection ram centerlines are parallel and overlap one another. By developing a scheme to bring the motion of these two axes to a single sensor, manufacturers can essentially cut their per-axis sensor cost in half. A graphical representation of this is illustrated in Figure 1. The technological “secrets” behind this capability are discussed in the next two sections.

Positio Sensor

Position Magnets

Carriage Unit

Machine Base

Magnetostrictive position sensing basics Magnetostrictive position sensors - invented in 1975 by Jack Tellerman - are essentially sonic (as opposed to ultrasonic) wave sensing devices. By using a high resolution clock, accurate absolute position between a fixed reference point and a moving magnet can be determined by the time it takes for a sonic wave to travel that distance. Another major benefit of magnetostrictive position sensing is that the position magnet does not touch the waveguide, meaning there are no parts to wear out and therefore long sensor life. This is accomplished using the internal sensing scheme illustrated in Figure 2. Waveguide Magnetoelastic material

Position Magnet

Sonic Wave Pickup Converts wave motion to electrical signal

Figure 1: A single magnetostrictive position sensor measuring two parallel axes of motion - the injection carriage and injection screw position

waveguide. In this instance, the so-called” Wiedemann-Effect” produces a torsional strain wave in the wire, explaining the name “waveguide”. • The strain wave travels at the speed of sound in the wavevguide (~ 2850 meters/second) in both directions away from the position magnet. One wave is absorbed at the far end of the waveguide by a damping mechanism (so no reflection waves interfere). The other wave travels to the pickup.

Magnet Position

• At the pickup, the torsional wave is converted to a longitudinal wave along another magnetostrictive strip of material know as Wiedemann-Effect the “tape”. Taking advantage of the Vallari-effect, which is Generates sonic wave essentially inverse magnetostriction, the changing strain on the tape in the presence of the constant magnetic field (from a Current Pulse Waveguide interrogation biasing magnet) generates an electronic signal in the conduc(Starts timer) tive pickup coil. The “return signal” is in the form of an impulse response whose shape is a highly predicable characteristic of this sensing configuration. The complete Return Signal waveguide, damp and pickup assembly is commonly referred Used to stop timer to as a Sensing Element (SE).

Figure 2: Fundamental physics of magnetostrictive position sensing

• The magnet position is determined by measuring the time duration from the initiation of the SE interrogation pulse until the return signal is detected by the pickup. The clock is

At their core, magnetostrictive position sensors are comprised of four basic components; Position magnet, waveguide, pickup (also known as “sonic wave converter”), and driver and signal conditioning electronics. The sensing process is as follows:

Null Zone

• A momentary current “interrogation” pulse is imposed on the conductive “waveguide” wire made from a nickel based magnetostrictive alloy. This creates the magnetic field concentric to the waveguide axis as shown.

Electrical Stroke Length

Dead Zone

Sensing Electronics Sensing Magnet

Pressure Isolation Tube (aka Sensing “Rod”)

Interrogation Signal T

• When the waveguide magnetic field interacts with the permanent magnet field from the position magnet, the magnetostrictive effect produces a resulting strain on the MTS Sensors

Injection Screw

Return Signal

~ 1 microsecond

Time Clock Trigger Level

Figure 3: Physical realization of a position sensor for hydraulic cylinder and associated signal timing response 2

551077

stopped when triggered by the time characteristic as shown in Figure 3. Here we have illustrated the internal sensing characteristic relative to a physical sensor realization for one particular application where the SE assembly is housed inside a high pressure-rated stainless steel isolation tube.

One limitation of this method shown in Figure 4, is two magnets that are in close enough proximity, their return signals will literally overlap and cause distortion from constructive wave interference. Therefore, multi-position magnetostrictive sensor applications are limited to minimum spacing between magnets of 75 mm. This also means that motion which would cause magnets to overlap one another during operation would result in a loss of valid magnet signal. Fortunately, in most applications of this technology, there is room to space the magnets, if not the measured moving surfaces are oriented so that the 75 mm spacing is achieved. Use of clever design and logical process steps will typically mitigate this physical sensing technology limitation.

• The drive and signal conditioning electronics serve to generate the precisely timed interrogation signal and convert the internal timing measurement to the desired output – analog voltage or current, digital pulse or serial data, or even high speed industrial network bus communication. The position (X) is proportional to the time between the two pulses (T) by the speed of sound (S), or

The primary challenge presented by this method of measurement is how to convey the individual magnet positions to the instrumentation or control interface. It is conceivable to use separate signal channels for each, however the wiring and connector interface would become complex and costly. MTS offers one of it’s analog output sensors from the Temposonics R-Series model family with two separate channels corresponding to two magnet positions. Otherwise, industrial fieldbus networks are the ideal medium for translating the necessary data for more than two magnet positions from a single sensor.

X=TxS The accuracy of the output is ensured with a precise scaling factor calibration – the speed of sound in the waveguide – which is found using a laser interferometer for each sensor at the final stage of production. While all of this may be much more than anyone ever wanted or needed to know in order to use a magnetostrictive position sensor, it is important to understand when considering applying the technology to multi-magnet positions.

In most instances where multi-magnet feedback with magnetostrictive sensors has been implemented, the means to communicate the position data has been Profibus DP (CANbus is another option). This is due to increased availability, popularity, a robust network structure and relatively high baud rate for fast data communication. Applying multiple-magnet sensing to the Profibus DP industrial network will illustrate one possible data structure and how controls engineers can easily take advantage of this capability.

Multiple-magnet position sensing We see from the above description that for a given interrogation signal, one return signal is created for one magnet. Therefore, if we add additional sensing magnets, we get one additional return signal at a later time for each subsequent magnet. This characteristic is illustrated in Figure 4. In order to determine these positions, it is necessary to continue capturing the elapsed time corresponding the entire length of the SE. It turns out that this is also the worst case situation for a single magnet sensor (imagine the magnet being at the position furthest from the sensing electronics), so there is no additional sensing time penalty when multi-magnets are used. Since this is a fundamental characteristic of magnetostrictive sensing technology, it comes at no additional cost to the sensor. MTS Temposonics sensors are capable of sensing 15 magnet positions simultaneously using a single sensor.

Without getting into detail about industrial fieldbuses and their implementation, the information along the network is communicated in serial “data frames” with standardized properties prescribed by the particular fieldbus creator and/or user community. Exactly how the Profibus DP bus data frame1 is structured and what this 1Standards maintained by the Profibus Trade Organization (PTO)

Sensing Electronics Magnet 1

Magnet 2

Magnet 3

T3 T2

Interrogation Signal

T1

Time

Return Signal 1

Return Signal 2

Return Signal 3

Figure 4: Magnetostrictive position sensor raw signal characteristic with multiple magnets applied to a single sensing element MTS Sensors

3

551077

data means is indicated by a data file supplied with the sensor (or from the company website) called the General Station Description, (GSD) file. Using the GSD file, a software configuration tool allows the user to select the necessary parameters and operational modes for a given system. As an example, for the MTS Temponsonic RSeries Profibus DP multi-magnet version P101, the sensor position data structure can be selected as one of the following three for each magnet position2:

magnet positions, the standard Intel data format is used, and all of the magnets are producing a valid return signal, each magnet position data will take the form shown in Table 2 below.

• Intel® standard data format with status - Status / Low / Med / High byte • Motorola® data format with status (Siemens®) - Status / High / Med / Low byte

0001

0000

1

Active

0010

0000

2

Active

0100

1000

4

Missing

1111

0000

15

Active

Med byte

High byte

1

0001 0000

LLLL LLLL

MMMM MMMM

HHHH HHHH

2

0010 0000

LLLL LLLL

MMMM MMMM

HHHH HHHH

3

0011 0000

LLLL LLLL

MMMM MMMM

HHHH HHHH

4

0100 0000

LLLL LLLL

MMMM MMMM

HHHH HHHH

Another advantage of using the fieldbus structure is the ability to alert and respond to errors or events along the network. For example, in the event that a missing or inactive magnet error is generated, the user may select one of the following options:

For example, once the magnet is moved far enough from the SE or from the normal operating stroke of the sensor, the SE will be unable to generate a return signal and therefore no position data. Because this may happen by design, the R-Series Profibus DP sensor has a number of “missing magnet” diagnostic options (discussed later in this section). Examples of valid magnet status bytes are shown in Table 1 below. Magnet no. operation

Low byte

One bit of the position data corresponds to the sensor resolution setting selected by the configuration program or default value in the parameter file. The entire data packet for this example would begin with the first magnet status followed by the data and the status and data for each subsequent magnet appended to the previous. This illustrates a primary disadvantage of the serial data communication structure of industrial fieldbus systems. An increased amount of data requires more communication time. For example, a 15 magnet system will require 60 bytes of data. Fortunately, at a 12 Mbaud communication rate, this only represents 40 microseconds of communication time across the bus.

The status byte is structured to indicate the magnet number (first four bits) and operational status (next four bits). The remaining three bytes (24 bits) are used for the position of the particular magnet number. The magnet status is one of two modes – enabled (on or active) = 0 or disabled (off, missing or inactive) = 8. It is important to understand that there may be instances where the magnet appears to be “off” or undetected by the SE for predictable or unknown reasons.

Magnet no.

Status

Table 2: Example of multi-magnet position data

• Inverse Motorola data format (Allen-Bradley®) - Low / Med / High / Status byte

Status byte

Magnet no.

• Generate zero position value for the particular magnet position. • Generate previous position value (hold) for the particular magnet position. • Generate zero position value for the all magnet positions. • Generate previous position value (hold) for the all magnet positions. These options could enable the controller to either continue normal operation until the next event or shut down the system safely. Other similar standard and custom options may be available upon request. Multi-magnet application examples Returning to the injection molding application, we can see from the expanded illustration in Figure 5, (on page 5), that injection molding machines provide an ideal application setting for multi-magnet sensors. Here, we see how two sensors are applied to four different (but parallel) position axes – injection carriage, injection ram, mold platen and ejector. In theory, this could be accomplished with a single sensor, but physical constraints can often limit realistic access to the sensor housing.

Table 1: Examples of multi-magnet status bytes for R-Series Profibus DP – P101 sensor

The example in Table 2 will help to illustrate how multiple-magnet position data is conveyed. Once the sensor data (GSD) file is installed on the Profibus “master” controller, the user can configure the sensor for the number of active magnets, position resolution (per bit), and other settings detailed later in this section including the diagnostic error handling. Assuming an application requires 4

Since these are considered motion control axes, it is a challenge to use an industrial fieldbus architecture and achieve the necessary closed-loop performance. For the application illustrated in

2 Referred to as “reduced data structures” to facilitate high speed data transfer.

MTS Sensors

4

551077

Figure 5, analog output sensors could be used, but another method developed specifically for this industry was the use of sensor outputs precisely synchronized to an external controller clock signal. By ensuring that this timing is essentially synchronous, machine builders can achieve quasi-deterministic results that help them produce their performance requirements, reduce the system cost associated with wiring and junction boxes, and achieve the distributed architecture advantages associated with industrial fieldbuses.

flexible sensing elements. This option is preferred for long stroke applications because these “flex” sensors can be coiled in 1 meter diameters thereby simplifying and greatly reducing the cost of shipping. With the practical limit of 15 magnet positions per sensor, more than one sensor would be required for such an application. The intention of this position system is to change the setup of the machine for the next slitting run. Originally, this was accomplished

Figure 5: Multi-magnet sensors applied to injection molding pocess

Another application where multiple-magnet sensors are often applied are on paper and film slitter and winder machinery. These applications (the slitter version is illustrated in Figure 6) are nearly ideal for these types of sensors. On some of these machines, measuring as much as 10 meters across, there can be as many as 60 slitters along the same axis. For these types of applications with strokes > 5 meters, MTS Temposonics offers sensors with

by manually moving and locking the slitter cartridges in place. Positioning was accomplished with a tape measure or gauge blocks. In order to reduce downtime and increase the number of setups possible in a given time, these positioning systems have been automated. Now, using the multi-magnet feedback approach, the next setup can be accomplished in seconds rather than hours.

Paper, Film or Foil Roll

Slitter Blades

...

Screw Drive

Positioning Screw

Position Sensor

Position Magnets

Cartridge Clutch

Figure 6: Multi-magnet sensors applied to paper slitting process MTS Sensors

5

551077

In this example, slitter positioning is accomplished by turning the positioning screw at a known fixed RPM in the appropriate direction. In order to move a cartridge, the controller simply engages the clutch until the appropriate position is reached and the clutch is released. Once released from the positioning screw, the slitter cartridge is locked in place by an automatic braking system. The machine controller can either position the cartridges in series or optionally with some cartridges in tandem, depending on the required final positions. This multi-position sensing capability has also been applied to multi-platen presses, printing rollers, and machine tools with multiple tools along the same axis to name a few. There are even seismic motion detection systems where the motion of an array of magnets is measured by a single sensor in order to sense geological properties of the earth’s surface as well as earthquake and avalanche activity.

• Use of smart programmable R-Series fieldbus sensors mean that setups can be stored and recalled directly at the PLC / HMI for even faster setup times. • Superior resolution and accuracy of the magnetostrictive position feedback mean a more precise “cut”, better process quality and less wasted product. • Advanced diagnostics available as a standard feature in smart fieldbus sensors means less downtime due to maintenance and troubleshooting. • The availability of multi-magnet sensing with longer, flexible sensing element sensors can help simplify shipping & installation and eliminate need to use overlapping shorter rigid sensors, both lowering overall cost.

Conclusions To meet the demand for lower overall cost, and more productive machinery, machine designers and controls engineers must continually develop innovative methods to control their products and processes. Historically, by using linear feedback, the factory automation industry has made significant advances moving from discrete sensing devices to continuous linear feedback. Now, as it has become a more understood and accepted, multi-position linear sensors are being used to further reduce per axis feedback costs while potentially increasing the in-use benefits gained from application of smart fieldbus-based sensors. Examining some of the applications of this technology, the benefits of using fieldbus based multi-position magnetostrictive sensors include: • Multiple tool, platen or cartridge positions from a single sensor means lower per tool cost of feedback. • Tool or cartridge positioning automation significantly reduces changeover time allowing for a higher number of setups and therefore higher machine productivity.

Part Number: 09-06 551077 Revision D Temposonics and MTS are registered trademarks of MTS Systems Corporation. All other trademarks are the property of their respective owners. All Temposonics sensors are covered by US patent number 5,545,984. Additional patents are pending. Printed in USA. Copyright © 2006 MTS Systems Corporation. All Rights Reserved.

UNITED STATES

GERMANY

JAPAN

MTS Systems Corporation Sensors Division 3001 Sheldon Drive Cary, NC 27513 Tel: (800) 633-7609 Fax: (919) 677-0200 (800) 498-4442 www.mtssensors.com [email protected]

MTS Sensor Technologie Gmbh & Co. KG Auf dem Schuffel 9 D-58513 Lüdenscheid Tel: +49 / 23 51 / 95 87-0 Fax: +49 / 23 51 / 56 491 www.mtssensor.de [email protected]

MTS Sensors Technology Corporation Ushikubo Bldg. 737 Aihara-cho, Machida-shi Tokyo 194-0211, Japan Tel: + 81 (42) 775 / 3838 Fax:+ 81 (42) 775 / 5512 www.mtssensor.co.jp [email protected]

Suggest Documents