US 20060186878 A1
A linear variable transformer (LVDT) for use in a transducer. The LVDT has a non-ferromagnetic core which may eliminate Barkhausen noise and thereby improve the sensitivity of the resulting measurements. In one aspect, this system may be used in an atomic force microscope.
1. A position sensor, comprising:
a moving coil part, having a first coil form formed of a non ferromagnetic material, and a coil element having an electrical connection part, formed around said first coil form, said moving coil part constrained to move in a linear direction, and said moving coil part including a connection element adapted for connection to a moving object of interest;
a stationary coil part, having a second coil form also formed of a non ferromagnetic material, and a second coil element wound on said second coil form, with said second coil element having at least first and second electrical connections which produce an output signal indicative of a moving relationship between said moving coil part and said stationary coil part, said stationary coil part sufficiently close to said moving coil part such that magnetic flux from said moving coil part is induced into said stationary coil part; and
a shell material, surrounding at least a part of said moving coil part and stationary coil part, said shell material supporting an external magnetic field therein.
2. A position sensor assembly, comprising:
a primary non ferromagnetic coil assembly including all non ferromagnetic material, coupled to a movable object of interest, and being movable according to movement of the movable object of interest;
at least first and second stationary coil assemblies, each including all non ferromagnetic materials, said first and second stationary coil assemblies being located adjacent to said primary non ferromagnetic coil assembly; and
an electronics portion, producing electrical current to one of said coil assemblies, and receiving an induced signal from the other of said coil assemblies, and determining a position of said primary non ferromagnetic coil assembly relative to said first and second stationary coil assemblies from said induced signal, said electronics portion including a synchronizing element which synchronizes a detection of a signal from said receiver circuit with a desired time of detection based on said drive signal.
3. A method, comprising:
driving a primary coil with a voltage;
sensing voltages induced into a plurality of secondary coils, from movement of the primary coil relative to the secondary coils, in a way that avoids parts of the signal being effected by Barkhausen noise.
4. A method, comprising:
using a molecular force probe to cause deflection of a cantilever; and
using movements of said cantilever to drive a plurality of coils which detects said movements substantially without being effected by Barkhausen noise.
5. A method, comprising:
using a surface profiling instrument to cause a deflection; and
moving using movements caused by the deflection to drive a plurality of coils to detect said movements substantially without being effected by Barkhausen noise.
6. A method, comprising:
using an atomic force microscope to cause a deflection; and
moving using movements caused by the deflection to drive a plurality of coils to detect said movements substantially without being effected by Barkhausen noise.
7. A system, comprising:
an atomic force microscope having an optical detection system, said atomic force microscope defining a frame of reference, said atomic force microscope having a cantilever, with a flexure that constrains the cantilever to move in a specified axis relative to said frame of reference, and with a moving coil part which detects movement of the cantilever in the direction of the specified axis, said moving coil being formed without ferromagnetic materials.
8. A system as in
9. A system as in
10. A system as in
This application is a divisional of U.S. patent application Ser. No. 10/683,592, filed Oct. 10, 2003, which is a divisional of U.S. patent application Ser. No. 10/016,475, filed Nov. 30, 2001, which claims the benefit of U.S. provisional application Ser. No. 60/250,313, filed Nov. 30, 2000, and U.S. provisional application No. 60/332,243, filed on Nov. 16, 2001. The disclosures of the prior applications are considered part of (and are incorporated by reference in) the disclosure of this application.
U.S. Patent Documents
U.S. Pat. No. 2,364,237 12/1944 Neff 340/199
U.S. Pat. No. 2,452,862 11/1948 Neff 340/xxx
U.S. Pat. No. 2,503,851 4/1950 Snow 340/196
U.S. Pat. No. 4,030,085 6/1977 Ellis et al. 340/199
U.S. Pat. No. 4,634,126 1/1987 Kimura 273/129
U.S. Pat. No. 4,669,300 6/1987 Hall et al. 73/105
U.S. Pat. No. 4,705,971 11/1987 Nagasaka 310/12
U.S. Pat. No. 5,414,939 5/1995 Waugaman 33/522
U.S. Pat. No. 5,461,319 10/1995 Peters 324/660
U.S. Pat. No. 5,465,046 11/1995 Campbell et al. 324/244
U.S. Pat. No. 5,469,053 11/1995 Laughlin 324/207.18
U.S. Pat. No. 5,477,473 12/1995 Mandl et al. 364/576
U.S. Pat. No. 5,513,518 5/1996 Lindsay 73/105
U.S. Pat. No. 5,705,741 1/1998 Eaton et al. 73/105
U.S. Pat. No. 5,739,686 5/1998 Naughton et al. 324/259
U.S. Pat. No. 5,767,670 6/1998 Mahler et al. 324/207.12
U.S. Pat. No. 5,777,468 7/1998 Mahler 324/207.18
U.S. Pat. No. 5,948,972 9/1999 Samsavar et al. 73/105
U.S. Pat. No. 6,267,005 7/2001 Samsavar et al. 73/105
Bertram, H. N. (1994) Theory of Magnetic Recording
Bozorth, R. M. (1951) Ferromagnetism
Crommie, M. F.; Lutz, C. P.; Eigler, D. M. Confinement of electrons to quantum corrals on a metal surface. (1993) Science, vol. 262, p. 218-20.
Drexler, K. E. (1991) Nanotechnology 2, 113-118.
Hristoforou, E.; Chiriac, H.; Neagu, M. A low core mass Linear Variable Differential Transformers sensor using amorphous wires. Romanian Journal of Physics, vol. 41, (no. 9-10), Editura Academiei Romane, 1996. p. 765-9. 4
Kano, Y.; Hasebe, S.; Huang, C. (Edited by: Suematsu, Y.) New type LVDT position detector. CPEM '88 Digest. 1988 Conference on Precision Electromagnetic Measurements, (CPEM '88 Digest. 1988 Conference on Precision Electromagnetic Measurements, Tsukuba, Japan, 7-10 Jun. 1988.) Tokyo, Japan: Soc. Instrum. & Control Eng, 1988. p. 95-6. xxi+430
Midgley, G. W.; Howe, D.; Mellor, P. H. (Edited by: Nicolet, A.; Belmans, R.) Improved linearity linear variable differential transformers (LVDTS) through the use of alternative magnetic materials. Electric and Magnetic Fields. From Numerical Models to Industrial Applications. Proceedings of the Second International Workshop, Leuvan, Belgium, 17-20 May 1994.) New York, N.Y., USA: Plenum, 1995. p. 311-14. xii+376 pp.
Meydan, T.; Healey, G. W. Linear variable differential transformer (LVDT): linear displacement transducer utilizing ferromagnetic amorphous metallic glass ribbons. Sensors and Actuators A (Physical), vol. A32, (no. 1-3), (EUROSENSORS V Conference, Rome, Italy, Sep. 30-Oct. 2, 1991.) April 1992. p. 582-7. 5 references.
Piner, R. D.; Jin Zhu; Feng Xu; Seunghun Hong; Mirkin, C. A. (1999) “Dip-pen” nanolithography. Science, vol. 283, p. 661-3.
Young Tae Park; Han Jun Kim; Kwang Min Yu; Rae Duk Lee Study on a linear variable differential transformer for precision measurements. Korean Applied Physics, vol. 2, (no. 4), November 1989. p. 347-51. 9 references.
Saxena, S. C.; Seksena, S. B. L. A self-compensated smart LVDT transducer. IEEE Transactions on Instrumentation and Measurement, vol. 38, (no.3), June 1989. p. 748-53. 6 references.
The present invention relates generally to devices that convert very small mechanical displacements, as small as the sub-nanometer level, into differential voltages, and vice versa.
One position sensor that has been available since early in the last century to convert mechanical displacements into differential voltages, and vice versa, is the linear variable differential transformer (LVDT). In the conventional commercially available LVDT (Part Number 50-00-005XA, Sentech Inc, North Hills, Pa.) depicted in
Another position sensor that can convert mechanical displacements into differential voltages is one employing capacitive technology. Commercially available capacitive sensors operating in circumstances analogous to commercially available LVDTs have surpassed LVDT performance by nearly an order of magnitude despite relying on signal conditioning circuitry of the same type as that employed for LVDTs (see, for example, the Physik Instrumente Catalog, 2001 edition). Consequently, even though they are difficult to work with and substantially more costly (because of more demanding manufacturing requirements), capacitive sensors are often the components of choice in applications that require the highest measurement resolution and bandwidth.
Because of their simplicity and low cost, it is of great interest to determine the sources of the limits in LVDT resolution and how those limitations might be overcome. As set forth more fully below, we have concluded that the limit on resolution in conventional LVDT resolution is Barkhausen noise in the ferromagnetic core. Barkhausen noise the name given to sudden jumps in the magnetic state of a ferromagnetic material (Bozorth). In ferromagnetic material, defects can lead to special sites where domain walls are preferentially pinned. The domain wall can then be depinned by thermal energy or an external magnetic field. When this happens, the domain wall will jump to another metastable pinning site, causing a sudden change in the overall magnetic state of the material. In general, when LVDT cores are formed from ferromagnetic material the flux changes due to Barkhausen noise will not cancel out in a differential measurement of the two secondaries. This uncanceled noise leads directly to noise in the core position signal. Furthermore, sudden changes in the magnetic state of the core can cause changes in the sensitivity of the sensor, again leading to positional noise.
There are a number of schemes to reduce Barkhausen (and electrical) noise in conventional LVDTs, including increasing the primary coil drive current and using amorphous magnetic materials in the core (Meydan, T., et al.; Hristoforou, E., et al.; and Midgley, G. W., et al.). While increasing the drive current will increase the signal to noise in conventional electronics, it is ineffective when dealing with Barkhausen noise because it creates a larger oscillating magnetic field that can in turn dislodge pinned domain walls more easily leading to increased positional noise. Using amorphous magnetic materials has been shown to reduce Barkhausen noise to a small degree but is ultimately ineffective because Barkhausen noise is a fundamental property of ferromagnetic materials.
Although without an understanding of Barkhausen noise issues, some non-conventional LVDT designs have eliminated its effects by substituting an air core for the ferromagnetic core of the conventional LVDT. These include a sensor designed for use in a very high magnetic field described by Ellis and Walstrom (U.S. Pat. No. 4,030,085), a pin-ball machine described by Kimura (U.S. Pat. No. 4,634,126) and a variety of mechanical gauging applications described by Neff (U.S. Pat. Nos. 2,364,237 and 2,452,862) and Snow (U.S. Pat. No. 2,503,851). The gauge described by Snow used an excitation scheme where two primary coils rather than one were excited. One coil was driven at 180 degrees from the other, resulting in oscillating magnetic fields from the two primaries that tend to cancel each other out. A single air core in the center was used as a detector. This is a different excitation scheme from the one usually used by us and others. Essentially, the roles of the primaries and secondaries are reversed. One might expect that from the reciprocity theorem (see, for example, Bertram, H. N., Theory of Magnetic Recording, Cambridge Press, 1994) that the electromagnetics of the two situations are identical. However, there are also some necessary differences in the noise performance associated with the signal conditioning. In general, the response of the sensors based on air core LVDTs in the prior art was significantly less sensitive than the improved sensor described here and would not be suitable for the sub-nanometer, high speed positioning performance we have obtained. Furthermore, the sensors described in the above prior art did not make use of any of the improvements we have incorporated into our excitation and signal conditioning electronics. The best performance claimed in these air-core LVDTs is comparable to current commercially available LVDTs.
Finally conventional LVDTs are also severely limited in the presence of an external magnetic field. As the external magnetic field increases, the core saturates and the LVDT becomes ineffectual. This limitation has been addressed by a non-conventional design in which the LVDT is fabricated entirely from non-ferromagnetic material and operated near the resonance of the primary and secondary coils (U.S. Pat. No. 4,030,085). This design however shares other limitations of conventional LVDTs in its large length scale and, for this and other reasons, is even less sensitive than the conventional design.
Present and future nanotechnology depends on the ability to rapidly and accurately position small objects and tools. Current length scales in many manufacturing technologies are in the 100 nm range and shrinking. The recording head in a commercial hard disk drive, for example, has a write gap routinely less than 100 nm and has pole-tip recession controlled to tens of nanometers. The current generation of semiconductor integrated circuits uses 180 nm wide traces, with the move to 130 nm expected within two years and to 100 nm within five years (International Technology Roadmap for Semiconductors, 1999 edition). Current sensors employed to control and verify the lithographic processes used in disk drive and IC manufacturing is already only barely providing sufficient resolution for critical dimension measurements.
Recent experimental work has gone beyond these relatively large sub-micron scales. Examples include the controlled manipulation of clusters of molecules (Piner et al.) and even individual atoms (see for example, Crommie et al.). One goal of nanotechnology is to build molecular machines (Drexler, K. E.). If such devices are constructed, they will require precise positioning of individual atoms and molecules. This will require three-dimensional positional information with sub-A precision and rapid response as even modest molecular machines in biological organisms can contain thousands of atoms.
An object of the present invention is to provide a simple, low cost and high resolution sensor that does not require the precision machining of other high-resolution position sensors or the careful selection, machining or treatment of the magnetic material used in conventional LVDT cores.
A second object is to provide a high-gain LVDT that does not suffer from Barkhausen noise.
Another objective is to provide a high-resolution sensor that is insensitive to temperature variations and does not itself cause temperature variations via eddy current heating of a high-permeability core.
Another objective is to provide a very high-resolution sensor that is suitable for integration with instruments requiring or benefiting from such resolution, including profilometers and scanning probe microscopes, such as atomic force microscopes and molecular force probes.
These and other objects are achieved according to the present invention by (i) replacing the high permeability core with a low permeability core that reduces or eliminates Barkhausen noise, (ii) reducing the length scale of the LVDT sensor to boost the spatial sensitivity and (iii) improvement in the signal conditioning circuitry.
The basic idea behind any LVDT is that the mutual inductances between a moving primary and two secondaries change as a function of position. The upper panel of
It is instructive to identify the limits on LVDT resolution. At least one manufacturer claims that its conventional LVDTs have “infinite resolution” and that any deviation from perfection is due to shortcomings in the signal conditioning or display (Lion Precision Model AB-01, Lion Precision, St. Paul, Minn.). This claim is puffery for at least two reasons. First, coils wound with wire will be subject to Johnson noise (Herceg, Edward E., An LVDT Primer, Sensors, p. 27-30 Jun. (1996). noise. As a result, Johnson noise is present in both conventional LVDTs and in the LVDT described here. Johnson noise will translate into an electric signal at the output of the signal conditioning electronics that will be indistinguishable from actual motion of the sensor. Second, and probably most important for conventional LVDTs, the use of ferromagnetic materials introduces the phenomenon known as Barkhausen noise. If the ferromagnetic core were perfect, meaning that the permeability is constant and the magnetization changes in a perfectly linear and smooth manner, this phenomenon could be ignored. However, ferromagnetic materials are never perfect. Barkhausen noise is the name given to sudden jumps in the magnetic state of ferromagnetic material. Defects in ferromagnetic material can lead to special sites where domain walls are preferentially pinned. The domain wall can then be de-pinned by thermal energy or an external magnetic field. When this happens, the domain wall will jump to another metastable pinning site, causing a sudden change in the overall magnetic state of the material.
We have concluded that the current limit on resolution in LVDT sensors is not signal conditioning or display errors but rather Barkhausen noise in the ferromagnetic core. Because the Barkhausen noise often originates from switching of small volumes of magnetic material, the flux changes in each of the two secondaries will not cancel out in the differential measurement. This in turn leads to noise in the core position signal.
We tested the hypothesis that Barkhausen noise is the limiting source of noise for high sensitivity position measurements in LVDTs. For this purpose, we measured the positional noise of (i) a commercially available LVDT using the as-shipped primary coil and ferromagnetic core (
A temperature stabilized, HeNe laser interferometer 20 and NIST-traceable calibration gratings of the sort commonly used in atomic force microscopy (Calibration standard, NT-MDT, Moscow, Russia) were used to calibrate the sensitivity and linearity of the different LVDTs. The beam 21 from the laser interferometer was reflected off of the moving core 18 which for this purpose was fitted with a retroreflector 22. We calibrated the different LVDTs by fitting the LVDT response to the interferometer response.
Because they originate from the microscopic distribution of pinning sites in ferromagnetic material, the details of Barkhausen jumps will differ from core to core, even when the cores are constructed from the same material.1 Furthermore, because the ferromagnetic core is close to a remanent state for a typical first LVDT measurement, the Barkhausen noise will depend on details of the magnetic history of an individual core.
Ferromagnets are frequently in metastable micromagnetic states. These metastable states are typically relatively far from the ground state. There are numerous paths for the ferromagnet to relax towards the ground state. This results in phenomena of slow relaxation where the magnetic state will gradually change over long time scales, typically from hours to days, weeks and even months (see, for example, Bozorth). As the magnetization relaxes towards an ever more stable state, the permeability and stability of the ferromagnet relative to perturbations by an external magnetic field will also change. In general, the domain walls will gradually relax toward a more stable configuration leading to fewer Barkhausen jumps.
One method of increasing the signal to noise is to simply increase the drive current. If the noise is at some constant value, increasing the drive current should simply increase the signal to noise.
It should be noted that the described LVDT can not only convert motion to voltage, but can, in reverse fashion, convert voltage to motion. The following applications of the described LVDT make use of one or both these properties.
Precision Force Measurements.
Surface Profiling. Conventional LVDT sensors have been used in a variety of surface profiling instruments, including a number of commercially available instruments (for example the Dektak 8 from Veeco Instruments, Plainview, N.Y.).
Use of the LVDT described here permits substantially improved profilometry relative to instruments employing a conventional LVDT. This includes surface profiling at higher speeds and lower forces. Moreover, since it is possible to both sense position and apply a force with our improved LVDT, the mass of the instrument can be reduced and its speed increased while allowing for a simplified design.
Atomic Force Microscopy.
Research, development and manufacturing of devices with micron and smaller length scales has begun and is accelerating. Recently, IBM has obtained some results where individual atoms are positioned on a surface with a local probe (in this case, the probe of a scanning tunneling microscope). Precision placement of atoms, a goal of molecular manufacturing will require sub-angstrom positioning of local probes in three dimensions.
The non-ferromagnetic sensors described here lend themselves to the length scale reductions required by these new techniques and devices. The sensitivity of the LVDT is proportional to the geometric sensitivity of the mutual inductance dM/dx. The mutual inductance is proportional to the magnetic energy and inversely proportional to the square of current,
As the length scale of sensors and devices decrease, magnetic sensors have another advantage. Magnetic forces scale as 1−4. This implies that the forces resulting from simply operating the sensor will scale favorably in the case of a MEMs or smaller sized non-ferromagnetic LVDT.
The described embodiments of the invention are only considered to be preferred and illustrative of the inventive concept; the scope of the invention is not to be restricted to such embodiments. Various and numerous other arrangements may be devised by one skilled in the art without departing from the spirit and scope of the invention.