US2470051A

US2470051A - Electrical impedance filament and the method of making same

Info

Publication number: US2470051A
Application number: US754749A
Authority: US
Inventors: Arthur C Ruge
Original assignee: Baldwin Locomotive Works
Current assignee: Baldwin Locomotive Works
Priority date: 1947-06-14
Filing date: 1947-06-14
Publication date: 1949-05-10
Anticipated expiration: 1966-05-10

Description

RUGE

METHOD OF MAKING SAME Filed June 14, 1947 m m M 6m P HR KW w/ m M w e m m R o 41% INVENTOR Arfhur C. Ruge A TTOR/Vfy menacin 10,19 9

ELECTRICAL 2,019,051. IMPEDANCEFILAMENT AND THE METHOD OF MAKING SAME Arthur C. Ruge, Cambridge, Mass., assignor to The Baldwin Locomotive Works, a corporation of Pennsylvania Application June 14, 1947, Serial N0. 754,749

"Claims. (CL 201-63) This invention relates to electrical impedance wire filaments and one object of my invention is to provide an improved wire filament having a desired temperature coeflicient of electrical resistance and an improved method of forming the filament so that the desired temperature coefiicient of electrical resistance may be obtained in a' convenient and controllable manner, it being understood that the term wire includes any other suitable form such, for example, as a ribbon.

In the past, it has been attempted to achieve a desired temperature coeflicient in a manner that required only the most painstaking effort, involving the use of materials of the very highest purity and the most exacting working and heat treatment of the product in all stages of its manufacture. Even so, it was only in the case of very pure platinum that a wire or ribbon could be produced in a practical manner so as to have a predetermined temperature coeflicient of resistance which could be regarded as accurate. It is for this reason that platinum was the accepted standard material for use in high-accuracy resistance thermometers. In contrast, nickel wire, which would otherwise be ideal as a resistance thermometer material for many applications, had to be obtained by selection if it was to fit a predetermined temperature-resistance curve. Slight impurities and variations in the cold working and annealing of the wire as it was manufactured serve to alter its electrical resistance response to temperature change.

In the field of precision resistors andwire strain gage manufacture, the problem of producing wire of predetermined temperature coefficient of resistance is even more acute. Here it is desired to have in the finished product a temperature coefficient of resistance as near to zero as possible, the practical limit being fixed by the end use of the product and the price that could be commercially justified for it. No pure metal has a temperature coefficient low enough for such applications, but in an effort to overcome this problem, alloys of two or more metals have been used which combine to give a low thermal coefiicient of resistance. Many such alloys and their composition are well-known, the most commonly used being the Well-known "Advance (or Constantan), Manganin, Ohmax, Nichrome, Kanthal, etc., the first two of which have exceptionally low coefiicients.

Inasmuch as one approach to true zero thermal coefficient of resistance as heretofore practiced involved a high degree of purity of material and precise control of manufacture in order to come within reasonable limits, it was only by a process of selection that the manufacturer could obtain wire which was suitable for the highest grade resistors, strain gages, etc. Selection was costly of time and material, and it required that adequate stocks be maintained well in advance of use. In another approach to this problem, as disclosed in my Patent 2,350,972, I employed combinations of two kinds of separate wires but strain measuring problems are complicated by the fact that the overall temperature coefiicient is a function of the coefllcient of thermal expansion of the material on which the gage is mounted, as well as the thermal coeflicient of resistance of the filaments themselves. The solutions of selection of material or of double selection in the latter instance are awkward, costly and imperfect.

. In my present invention I employ a wire, filament, or ribbon, the thermal coetficient of resistance of which preferably but not necessarily is already as close to the desired value as may be readily obtained commercially. On this conductor I place a coating of electrically conductive material so chosen that the desired overall coeflicient is obtained by regulating the thickness of the coating. Such a coating may be applied by any of the well-known methods such as electro-plating, 'flash coating, vacuum evaporating, sputtering, dip or spray coating with or without a binder, chemical transforming of a coating applied in any manner, etc. I make no claim to these general methods of coating per se but merely list a number of representative and well-known ones that can be used, among others, as one element in producing my invention. Hence it is seen that broadly my invention comprises the application of an electrically conductive coating, or a combination of conductive coatings, to an electrical conductor in order to achieve a desired overall temperature coefiicient of resistance for the resulting structure.

Other objects and advantages will be more apparent to those skilled in the art from the following description of the accompanying drawings in which:

Fig. 1 is an enlarged cross sectional view of a filament embodying my invention;

Fig. 2 is a diagrammatic circuit for illustrating the operation of my improved filament;

Fig. 3 is curves illustrating certain characteristics of my improved filament; and

Fig. 4 diagrammatically illustrates a plating arrangement that can be employed in producing my filament.

In Fig. 1 I show a conductor I on which a conductive coating 2 has been applied. For simplicity, I show a round wire I of diameter d with a single coating 2 of thickness t, whereas the conductor could be of any shape and a multiplicity of coatings could be applied solong as the desired overall thermal coefiicient of resistance is obtained. In Fig. 2 I show the equivalent circuit of this simple structure, 1'1 and 1': being the electrical resistances of a length t of conductor I and coating 2 respectively. The overall resistance is the parallel resistance of 11 and r: and the overall thermal coemcient of resistance is a function of the relative magnitudes of n and r: as well as of their individual thermal coefiicients of resistance.

If we assume for the sake of simplicity that the two conductive elements I and 2 possess linear thermal coeflicients of resistance on and as respectively, and that the coating thickness t is such that the resistance of length 4' is reduced from n before coating to (1-p) 11 after coating, then it may be shown that the theoretical change in overall thermal coefllcient is simply Mar-a1) to first order accuracy. As a first approximation this formula is very useful, but, as a practical matter, I have found that it is necessary to determine the proper value of p for a given coating by actual experiment where close results are required. As an example, Fig. 3 shows the results of gold plating applied -to a hooo" diameter constantan wire. The abscissae are plotted in terms of the quantity 3: multiplied by 100 (i. e., the percent change in resistance due to plating), while the ordinates represent the change in temperature coefilcient of resistance of the structure. The curve indicated as measured is taken from tests of six samples of the parent wire plated to various thicknesses. The curve marked Theoretical is calculated from the formula pun-a1) using the value of as for gold as given in Handbook of Chemistry and Physics, 26th edition, and the experimentally determined value of on for the constantan wire in question. It is seen that the disagreement is quite marked for low values of 12, whereas the two curves approach parallelism as 31 increases. This simply means that the temperature coemcient of .very thin films is considerably different from that of bulk metals. It should be realized that a 2% reduction in resistance in this instance corresponds to a gold plating thickness of a fraction oi a controlling the result by making use of the change of resistance produced by the coating makes it possible to obtain a high degree of accuracy and reproducibility. By contrast, such methods as are based upon change in weight or diameter are very crude indeed. In the case of a coating applied by electroplating I can also control the amount of coating by regulating the temperature of the bath, its concentration. the time, and the plating current density. Similarly, in evaporation or sputtering coating I can control the well-known conditions to regulate the thickness of the coating. However, in all of these methods I consider it preferable to use the change in resistance as my final control since it is susceptible of very high accuracy with the aid of relatively simple equipment. In all of these methods I could, if desired, measure the resistance change while the coating is being applied. I make electrical contact to the ends of the piece being coated (or a desired section of it) and merely observe the resistance change during the coating by means of a suitable resistance measur ing bridge.

In the case of electroplating there is of course a shunting effect due to the conductivity of the plating bath. However, I can either take this eflect into account in making my measurement of resistance and resistance change or I can arrange the plating bath so that its shunting eflect is very small or negligible. Fig. 4 shows a suitable arrangement, although many variations are of course possible. Filament 3 is introduced into a small-bore tube 6 which is filled with suitable plating solution. Tube 4 may be made of the metal to be plated on filament 3 or it may be lined on its inner surface (by plating or otherwise) with the desired plating metal. Insulating supports 5 hold the filament 3 so that it does not make direct contact with tube 4. The plating circuit is indicated generally at 6. Leads 7 connect filament 3 to a suitable resistance measuring bridge which may be an AC bridge so that measurement is not influenced by the DC plating current passing through the filament; or the plating cur= rent may be interrupted during measurement of resistance. It is only necessary to make the bore of tube 4 small so that the lengthwise conductivity of the plating bath is small relative to that of the filament being plated. Tubes 8 provide for introduction of fresh plating fluid as needed, it being understood that the holes in supports 5 through which the wire 3 passes are sumciently small so that leakage of plating fluid is negligible.

In using the system shown in Fig. 4 for a continuous process it requires only the provision of sliding contacts 9 and II) to the filament which will be pulled through at a uniform rate by winding on a rotating spool. The fluid supply means 8 allows either periodic or continuous replacement of the plating fluid introduced at one of the pipes 8 and discharged at the other at the required rate. The process may be made automatic by means of any well-known automatic current regulating controller which regulates the plating current as the wire moves through the bath so that the percent resistance change of the filament is held constant or at a desired value, it being understood that the speed of the filament is held substantially constant. Alternatively, the speed of the wire as it moves through the bath can be varied by such controller while the current remains substantially constant so as to obtain the above mentioned result of change of re sistance.

For most conductive coatings the temperature coeflicient of resistance will be found to be positive, but it will be seen that the change in coefiicient of the coated wire can be made positive or negative by choice of coating material such that (112-411) is positive or negative. In the case of such a resistance wire as that known as "0hmax, my coating method is especially advantageous. This wire has a negative thermal coefficient which is easily corrected for by a thin coating of a material such as gold, nickel, rhodium, or the like. The high resistance of this material is only slightly diminished by the coating required to bring it to a zero coeflicient so that it becomes an excellent wire for precision resistors, strain, gages, etc. Without the coating the coefficient is too high, to enable one to make use of it as a precision resistor material.

Advance and manganin are well-known materials in the low temperature-coefllcient range. Both of these materials can be made to have negative coefficients so that they are easily brought to zero coefiicient by very small coatings of positive coeflicient materials. In using the term zero coefficient I mean of course zero" coemcient under certain specified conditions.

' Since all conductors exhibit more or less sensitivity to strain, it is clear that the temperature coefiicient of a given wire or filament will be afiected by any thermal strains that are imposed upon it. If, for example,-a resistor is made by winding a wire about a plastic card or form, the final coefficient of the resistor is dependent among other things upon the strain set up in the wire as a result of thermal expansion of the plastic as well as upon the original coeflicient oi the free wire. One of the decided advantages oi my coated wire method is that I can adjust the wire by a simple means so that it will have the desired charactertics when in its final condition of use. .Thus I need not make the usual compromisesfor the sake of expediency but instead I can tailor my filament to suit the use it is put to.

Now in some instances it is desirable to supply a coating having a negative temperature coefilcient. There are a number of materials that are suitable for this purpose; among them are certain alloys of bismuth-lead, bismuth-tin, bismuth-lead-tin, and others known to the metallurgical art.

Also, there are a variety of so-called semi-conductors such as silver sulphide, copper oxide, uranium oxide, etc., which exhibit negative tem- 'perature coefficients of electrical resistance. The

alloys can be plated, evaporated, sputtered, or even in some instances applied by passing the wire through a molten bath as in tinning. The semi-conductors may for instance be coated from a frit suspension or they may be produced by chemical transformation. As an example. a silver sulphide coating may be formed by first coating the filament with metallic silver and then transforming the silver to silver sulphide by exposure to heat in a sulphur atmosphere. The numerous possibilities of chemical transformation are well-known and therefore need not be listed. It is to be noted that since my invention deals with relatively thin films of coating material I can employ chemical transformation successfully whereas it would not be practical in-the case of thick layers where the transformation process might not penetrate properly or where the transformed layer would lose its adhesion as a result of its relatively great thickness. Another way to form the semi-conductor film is by deposition from a chemical bath.

The use of multiple coatings is also of practical interest, especially where it is desired to protect a less stable coating such, for example, as silver with a stable coating like gold or platinum. It is only necessary that the coatings be so controlled that the desired overall coefilcient is obtained. An alloy of a coat with the filament body or of alloy coatings on the body may be formed by well-known heat-treating after deposition of a coat or of multiple coats of diflerent materials making up the alloy, and the term "coat" or "coating" herein includes the resulting material.

Another aspect of my invention is the possibility of drawing down or rolling out a wire or filament which has been coated when in a size large enough to permit easy handling and convenient control of the coating. Thus, I may start with a 5-mil wire which I have determined to be suitable if coated with i mil of platinum and then drawn down to the final desired size which may be 1 mil. This may in some instances be easier to do with existing commercial equipment than it would be to apply the correspondingly minute film of platinum to a wire initially drawn down to final size.

An interesting variation of my invention involves the removal of excess coating when necessary or desirable. The simplest and preferred method is by electro-chemical deplating, although chemical erosion or transformation can also be mployed.

A further and very important aspect of my invention concerns the use of my coating method to alter the strain-sensitivity of the coated wire, ribbon, or filament. Suppose, for example, I wish to produce a wire for resistance thermometry in which the resistance versus temperature is substantially unafiected by strain or dimensional changes of the material on which it is wound or mounted While being highly sensitive to temperature change. Dimensional changes of the material on which it is mounted whether due to temperature, shrinkage, aging or other causes, are thus made unimportant as regards the stability and accuracy of the finished product. To accomplish this I can, by way of example, use a platinum wire coated with nickel. Since nickel has a negative strain sensitivity while platinum has a ositive strain sensitivity, it will be seen that I can so proportion the nickel coating that the composite structure will be insensitive to strain. And since both materials have high positive temperature coefllcients of resistance, the combination is ideally suited to resistance thermometry. Also, I can for instance make use of bismuth and certain bismuth alloys as the negative strain-sensitive material.

Carrying the principles of my invention to the bonded or unbonded wire strain gage field, it is seen that I can control the strain-sensitivity within rather wide limits and thus arrive at a wire having a desired strain sensitivity without having to resort to selection as has been the case heretofore. Furthermore, by use of multiple coatings it is possible to control both the temperature coefiicient of resistance and the strainsensitivity of the composite structure. Having the means to perform these functions individually I as already described, it is merely a matter of combining the required proportions to perform the combined function.

An important application is to a bonded or unbonded wire resistance strain gage which is made of a coated wire of such characteristics that the gage when mounted on a given material is substantially unafiected by the unrestrained thermal expansion and contraction of said material. Such a gage will measure strains which are associated with stress in the material regardless of whether or not they are of thermal origin, but will not respond to thermal strains in the free or unrestrained material. The coated wire can be bonded direct to the material or it can be supported on a structure or device which transmits strain to the wire, as for example in the unbonded type of wire gage such as that shown in Carlson Patent 2,059,549. In the bonded wire type. it is of course well-known that the filament is attached to or supported'by the material subject to strain by a medium. In the case of the unbonded wire type strain gage, mechanical supporting means are used to attach the filament to the material whose strain is to be measured.

The same technique permits the construction of a wire for high-precision resistors such that the wire exhibits very low or negligible sensitivity to strain and. at the same time has a low or negligible temperature coeflicient of resistance. Such a wire does not require the careful aging for strain relief that is necessary in the case of wires having strain sensitivity, and its thermal coefflcient of resistance is not a function of the material which it is wound or mounted on. Dimensional changes of the material on which it is mounted whether due to temperature, shrinkage, aging or other causes, are thus made unimportant as regards the stability and accuracy of the finished product. As an example of a practical combination of materials for such a wire, I can use Advance wire coated first with gold and then with a bismuth-tin alloy or vice versa. The bismuth-tin alloy, which may be tin for example, will have negative strain sensitivity and a negative thermal coefficient of resistance, whereas the gold has positive strain sensitivity and positive resistance coefiicient. The advance wire will exhibit a moderate positive strain sensitivity and a very small thermal coeflicient of resistance which may happen to be either positive or negative. The coating of gold will serve to change the temperature coeflicient in a positive direction and to increase the positive strain sensitivity slightly. The bismuth-tin coating will offset these two efiects. If the proportions are proper, and ifthe characteristics of the materials are within limits which can be mathematically determined, then it is possible to produce a composite structure having truly negligible thermal coeflicient of resistance and at the same time truly negligible strain sensitivity.

It will be understood of course that in the above disclosure I use the term thermal coeflicient of electrical resistance" in the broad sense, realizing that the coefficient is not necessarily a constant over a wide range of temperature. Indeed, with most materials the resistance vs. temperature function is definitely non-linear. In the application of my invention I can control the coeflicient (or resistance-temperature rate) at or in the vicinity of a definite point of temperature. Such would be the case in a precision resistor or wire strain gage where the range'of useful operation can be as limited as the accuracy or operating requirements may dictate. 0n the other hand, I can control the resistance change over a wide temperature interval-or perhaps the average coefficient over an interval-in which case the expression thermal coeflicient" in the above disclosure broadly represents. the resistance-temperature function. Thus, in a resistance thermometer I may by my coating method not only control the total resistance change over a given interval, or the co-efficient at a given point, but I may at the same time control the form of the resistance-temperaturefunction over the interval. In other words, I can within limits fixed only by the available conductive materials construct a coated conductor having a desired resistance-temperature function. This serves to 'f further illustrate the broad usefulness of my invention.

Another application of my invention is to the manufacture of electrical impedance elements or devices in which the impedance is substantially unaffected by dimensional changes in the material on which the element is wound or mounted. For example, a precision resistor or other impedance" element such as a coil wound on a plastic card or spool would thus be made insensitive to dimensional changes of the card or spool by making it of a coated filament which is insensitive to strain.

From the disclosure herein it is seen that I have provided an improved filament whose enumerated qualities are of precision order and have also provided a very effective and novel method for producing such filaments.- The temperature-resistance function and/or the strain-sensitivity of such filaments can be deliberately made to have the desired characteristics within the limits set by available materials. It is to be understood that in the claims the words coat" and coating apply either to single or multiple coats.

It will of course be understood by those skilled in the art that various changes may be made in the construction and arrangement of parts without departing from the spirit of the invention as set forth in' the appended claims.

I claim:

1. An electrical resistance filament having a substantially zero change of electrical resistance in response to temperature change, comprising a continuous body portion of electrically conductive material substantially uniform in cross section along its length, and a coat of substantially uniform cross sectional area, on said body, of electrically conductive material having a change of electrical resistance in response to temperature change diflerent and of opposite sign from that of said body and of a thickness so that the filament has an overall substantially zero resistance change in response to temperature change.

2. A strain gage electrical resistance filament having a predetermined change of electrical resistance in response to a given change of strain, comprising a body portion of electrically conductive material and a coating, on said body, of electrically conductive material having a change of electrical resistance due to strain different from that of said body and of a thickness so that the filament has an overall change of resistance in response to strain of said predetermined value.

3. The combination set forth in claim 2, further characterized in that the coating of material is of such thickness and thermal coefiicient of electrical resistance and of such change of resistance in response to strain that the filament has a substantially ,zero change of resistance in response to temperature and a predetermined change of resistancein response to strain.

4. The method of forming an electrical resistance filament having a predetermined thermal coefiicient of electrical resistance consisting in taking a body portion of electrically conductive material and then coating the same with an electrically conductive material of such thickness and of such thermal coeflicient of electrical resistance opposite in sign to that of the body portion that the filament has an overall coeficient of said predetermined value.

5. An electrical resistance strain gage adapted to be mounted on a material whose strain is to be measured and which material has a given coeflicient of thermal expansion, comprising a filament, means attached to said material for supporting said filament so that the resistance of said filament changes in response to strain of said material, said filament having a body portion of electrically conductive material and a coat of electrically conductive material different from that of said body portion and of such thickness and of such thermal coefficient of resistance that the change of resistance versus strain of said filament is substantially unaffected by unrestrained thermal expansion and contraction of said material whose strain is to be measured, whereby said filament is responsive to strains resulting from stress in said material.

ARTHUR C. RUGE. 15 2,335,707

Number Name Date 10,944 Weston July 17, 1888 1,180,614 Simpson Apr. 25, 1916 1,223,322 Gebaur Apr. 17,- 1917 10 1,273,506 Lederer July 23, 1918 1,441,686 Jones Jan. 9, 1923 2,023,603 Lodge Dec. 10, 1935 2,138,938 Plensler Dec, 6, 1938 2,318,102 Ruge May 4, 1943 Streicher Nov. 30, 1943 10 REFERENCES crrEn The following references areof record in the file of this patent:

UNITED STATES PATENTS