USRE23615E - Compensated thermopile - Google Patents

Description

W. G. FASTIE COMP Jan. 13, 1953 ZNSATED THERMOPILE Original Filed Jan. 18, 1950 4 Sheets-Sheet l INVENTOR. WILLIAM G. FASTIE Ldd e a ,0. EM

ATTORNEYS 4 Sheets-Sheet 5 Original Filed Jan. 18, 1950 1N VEN TOR. WILLIAM G. FASTIE wwaiwg MM ATTOR N EYS Jan. 13, 1953 w. G. FASTIE 23,615

COMPENSATED THERMOPILE Original Filed Jan. 18, 1950 4 Sheets-Sheet 4 INVEN TOR. WILLIAM G. FASTIE 1) .2 a J M ATTORNEYS Reissuecl Jan. 13, 1 953 23,615 COMPENSATED THE-BMOPILE William GfFastie, Owings Mills,

Leeds and Northrup Company,

Md., assig'nor tb Philadelphia,

Ba a corporation of Pennsylvania I dated June 24,1952, Serial aegis-m1 in). 2.6011508;

No.139,308, vJa' uary reissue August 14,

18, 1950. Application for, 1952, Serial No. 304%: 'j

. 35 Claims. (01. 136-4) Matter enclosed in lieavy brackets I: II appears in the original patent but terms no part oi this reissue specification; matter printed in italicsindicates the additions made theme.

This invention relates to electrical measuring apparatus which includes a thermocouple, a thermopile or analogous structure for measuring the intensity of radiant heat, such for example, as received from a radiant body, and has for an object the provision of a structure whose electrical output is substantially unaffected by ambierrt temperature. Structuresof the character to which my invention is applicable are not only especially useful in the pyrometry field Where it is desired to measure temperatures ranging from 300 F. to 4500" F., but also have utility generally in devices which involve measurement of radiant energy, as for example, the measurement of quantities or properties of material interposed between the structure and a source of radiant energy. There are many types of such equipment used for determination of density, material composition, etc.,'iutilizing the transmission characteristics of the material to change the radiation received by the responsive element from a source.

In Quereau Reissue Patent 19,564, there is disclosed a compensated thermopile in which the thermocouples are constructed of iron and constantan. In accordance with the teachings of the Quereau patent, the tendency oi the temperature-diiference to decrease with an increase of ambient temperature is the rising voltage-temperature characteristic of the elements of the thermopile. In order that the tendency of the temperature-difference to decrease may be compensated by the rising voltagetemperature characteristic, Quereau proposed, among other things, the resistance between the .hot and cold junctions, as by increasing the diameter of the thermocouple wires, so that heat could be rapidly removed by conduction from the hot'junctions and the target. By this construction, radiation loss from the receiver, with resultant lowering of its temperature with respect to its surroundings, was minimized. In practice, the thermocouple wires of the Quereau thermopile were one-hundredth of an inch in diameter with a target size of one-half 'inchdiameter. One difficulty with the thermopile of the Quereau type has been the relatively heavy .mass of the thermocouple wires and the target resulting in undesirably low sensitivity and a relatively long response time. Such. thermopiles have been found. inapplicable to temperature measurements in the high-temperature pyrometric field where the requirements areshort response time and high sensitivity.

compensated for by lowering of the thermal effort to-overcome some-ofthe undesirable features of the Quereau type of thermopile, ithas been proposed to make thermopiles of low heat capacity, low heat conductivity and of thermopile materials having a rising voltagetemperature characterist-ic. In most thermopiles of the last-mentioned type, the tendency of the temperature-difference to decrease with an increase of ambient temperature is not entirely compensated for by the rising voltage-temperature characteristic of the thermopile. Accordingly, it has beenproposed toconnect in parallel with the output terminals of the thermopile a compensating coil comprisinga. resistor having a suitable temperature coefficient of resistivity to provide the additional compensation needed to correct the thermopile output for ambient temperature efiect. .7

In accordance with the present invention, it has been recognizedv that a number of materials ordinarily used to .form thermocouples-have positive temperature coefficients of heat conductivity, that is, the heat conduction increases with rise in temperature vlevel for the same temperaturedifierence, andlit hasnbeen further recognized that other materials have negative temperature cofiicients. of conductivity, i..;e.,. the heat. conduction decreases .withjazrise .in temperature level for the same temperature-difference. More particularly, Chromel constantan and goldnickel alloys have .pos'itive temperature coefficients. of conductivity, andiron, nickel, copper andjothers have negative coefiicients or conductivity. In the compensated thermopile of Quereau, the iron and constantan components of each thermocouple actv oppositely because of the oppositev sign. of their'icoefficients of conductivity. The values of thermal conductivity of constantan and iron'and'the values of their temperature coeflicients otthermal conductivity are such that] the. .netv thermal conductivity, .f or given equal length and equal lO'mil diameter wires of. an, iron constantanthermocouple, does not appearto change with change in ambient temperature level ov'era rangeas from 1B.C. to ?.0. vvIn thermopiles including elements of Chromel. and constantan alloys, both. elements have an, increasing conductivity-temperature characteristic which tends 'toidecrease the output voltage with increasing ambient temperature.

Furthermore, the loss of .::energy .from the thermopile receiver through gaseous conduction also;inc reases with increasing; ambient temperature, ;further tending; to ldecrease the output voltage.: -.Furthermore;.-radiation-; loss vfrom the 3 thermopile receiver increases with increasing ambient temperature, further reducing the output voltage. The rising voltage-temperature characteristic of Chromel-constantan is insufficient completely to offset all of these effects, no matter what wire dimensions or receiver dimensions, or gas is used, so that some form of compensator must be used to correct for the ambient temperature error of a Chromel-constantan thermocouple. It is desirable to use Chromel-constantan for of its strength, large thermoelectric effect low thermal conductivity and thermoelectric stability.

In accordance with the present invention it has been found that a Chromel-constantan thermopile whose output is independent of ambient temperature maybe produced in several ways. In one preferred form of the invention, a thermal shunt is provided between the hot junctions and the cold junctions to establish a ratio between the total heat losses and the losses along the solid conductors from thehot junctions to the cold junctions to provide a thermopile with an invariant voltage output notwithstanding change of ambient temperature. Further in accordance with the invention, the proper control of the ratio of the heat losses along the solid conductors to the total losses from the hot junctions to thecold junctions will provide ambient temperature-compensated thermopiles.

For further objects and advantages of the invention and for a more detailed understanding of the invention, reference is to be had to the following description taken in conjunction with the accompanying drawings in which:

Fig. 1 is a sectional view of a radiation pyrometer embodying the invention;

Fig. 2 diagrammatically illustrates the optical system of Fig. 1;

Fig. 3 is a plan view of the part M of Fig. 1;

Fig. 4 is a perspective view of the sensitive elementand associated parts of Fig. 1, greatly enlarged;

Fig. 5 is an enlarged sectional view of the sensitive element and including its housing, with member M of Fig. 1 securedthereto;

Fig. 6 is a perspective view of an adjustable light trimmer 40 secured to the shank of a screw 4| in Fig. 1 which can be locked in fixed position by a locking screw 42;

Fig. 7 is a plan view of a of a thermopile embodying Figs. 8, 9 and 10 are plan views respectively of further modifications of thermopile assemblies embodying the invention.

Referring to the drawings, the invention in one form has been shown applied to the sensitive element or thermopile ID of aradiation pyrometer II which includes a suitable optical system of the type disclosed and claimed in an application Serial No. 154,690, filed April 7, 1950 by Raymond C. Machler for Impro ed Optical Systems for Radiation Pyrometers, and assigned to the same assignee as the present invention. The optical system of the radiation pyrometer H thermopile assembly the invention; and

limits the energy received by the thermopile to that emanating from a sharply defined area of the surface of.[the] a hot body or source S and produces a radiant energy flux of increased density at the sensitive element or detector In by employing a two-mirror optical system which includes the concave mirror l2 and the concave mirror l3. There are combined the energy-concentrating property of a wide angle optical ele-' ment with the sharp image-forming property of thermocouples because 7 i are thermally intimately connected thereto as by a narrow angle optical element. In Fig. 1, the narrow angle primary mirror I2 produces on a diaphragm I4 an image of the source S, illustrated as a section of a refractory wall or other heatradiating body. In Fig. 2 the source S is represented by an arrow S, and the image thereof by the arrow s. Radiation from only the sharply defined portion of the source represented by that part of the image s in front of opening 14a of the diaphragm passes through that opening Ma to the wide angle secondary mirror l3 which redirects and concentrates the received radiation upon the thermopile or detector H1. The mirror l3 forms an image of the opening l4a on the target of the thermopile, which image is smaller than the target 25, Fig. [3] 2. Due to the mirror l3, the density of the radiant energy on the target of the detector is greater than it is at the opening Ma.

It is to be further observed that the mirror I2, because not silvered at its central portion, provides a transparent opening I 2a through which there may be viewed through the eyepiece IS the image s, Fig. 2, projected on the plate l4. It is a feature of the present invention to provide a coating of magnesium oxide on the surface of plate I4 facing opening or aperture Ila. The coating has the following properties: It does not scatter or reflect appreciable radiation in the infra-red region where the greater amount of radiant energy from hot bodies is located. Hence, that surface does not result in errors due to the presence of scattered radiation. Radiation entering pyrometer H from other than the optical line of sight is not reflected within the housing since all interior surfaces are coated with a dull black paint except the window, mirrors and the surface coated with magnesium oxide. The latter coating provides a visual image or makes visible an image of the sharply defined area of the source and, hence, provides a means by which the pyrometer may be sharply focused. A window It 'is'provide'd to protect the interior of the instrument from ingress of foreign material. The eyepiece IS, the windowlfi and the housing as a whole may be sealed for operation of the thermopile under subatmospheric or superatmospheric pressure. The window IE may be made of quartz or other material which will pass energy' in' the infra-red region and which will withstand the heat to which it may be subjected in use.

When the thermopile is thermally insulated from its surroundingsthe transient effects due to sudden changes in ambient temperature may be reduced. However, the insulated system does not permit absorbed radiant energy readily to leave the thermoelectric system, with the result that the thermopile can be heated to undesirably high operating temperatures if subjected to intense radiation. In accordance with the present invention, the thermopile is not thermally isolated, but nevertheless, its output is not to any substantial extent changed by a rapidly varying ambient temperature. This is accomplished by adjusting the physical dimensions of the parts associated with the cold junctions and the parts associated with the hot junctions, so that they respond in like manner to changes in ambient temperature. vMore particularly, it will be observed, Fig. 4, that the cold junctions I8 of the thermocouples l9 are in intimate heat-conductive relation with a metal mounting ring 20. Though, electrically insulated therefrom, they eaers g a suitable ceramic cement- ,Ihe mountin ,rl s zflmay be 'o a y s ita le metal such nicke r. copp r- T ring stherma yi nne ed to an outer housing 2|, Figs, 1 and 5, through three metallic legs 25a, b and 20c which may be integral with the ring 20, orthey may be riveted or welded inserts. The copper housing 2| including end-plate or closure 21a and plate N form a thermal shield completely surrounding the thermopile except for the opening Ma which allows radiation to pass to the mirror l3. The copper housing 2| is in thermal andmetallic contact with the diaphragm |4, which ls itself preferably made of heavy copper, and isalong a shoulder of housing 22 in metalli and thermal contact with it. Thediaphragm' ULF g, 3, has three legs .|4b spaced 120?apartinterconnectin a central supporting disc I40 and the rim of the diaphragm. To increasethe flow of heat from the thermopile housing 2| to the pyronieter housing 22, the legs Mb; may be made larger, particularly in cross-sectional area The size of the extensions 20a, 20b and 200, Fig. 4, and the thickness of the copper legs Mb are so proportioned as to maintain the supporting, disc and housing 2| at substantially ambient temperature, The thermal conduction thus provided is effective to achieve the foregoing objective when [measuring the energy] making measurements with the device as a whole in an intense radiation field.

Referring to Fig. 5, it is to be observed the extensions 20a, 20b and 200 are formed by pins having enlarged end-portions which abut against the inturned ends of the housing 2| which receive the threaded ends of screws extending through the plate M. The pins have reduced portions which are pressed into openings of the ring support 20 of the thermopile, and each pin is provided with further reduced end-portions bent slightly toward the axis of the housing 2| for engaging at points spaced approximately 120 apart the concave face of the mirror l3, a spring extension 2|e from member 2l=b engaging the back of the mirror l3 to press it against the ends of the three pins. The spring-carrying member 2| b is secured to an [end-screw] end-plate or closure 2|a of housing 2| by a hollow rivet 2|c which is preferably provided with a block 2 Id to prevent entry of radiationinto the housing 2|. It is to be observed that the openings in plate M- through which the fastening screws extend are made somewhat larger than the shank of the screws to provide for optical adjustment of housing 2| relative to the opening Main plate I4, thus providing for optical adjustment of the illustrated assembly with reference to plate I4.

With the above understanding of the elements of the radiation pyrometer thus far described, attention will now be directedto the thermopile itself, the materials of which it is made, the proportioning of the parts, and the thermal and electrical functions of thepartswhich are best shown in the enlarged isometric view ofFig. 4. The several thermocouples l9, four of them extending [diametrically] radially to the right and four to the left respectively of the target 25, have their hot junctions intimately in thermal contact with the target but electrically insulated there-' from. More particularly, the target or radiation receiver 25 includes a platinum disc to which is secured, as by a ceramic glaze, a disc of mica. The hot junctions of thermocouples l9 are secured to the mica disc in like manner Preferably a small quantity of powdered black'ceramlc material, know as a frit, placed on themlca disc. llpoii application ofhheatthe frit ismelted. to form'a fused'coatingto ,secure theihot junc-p tions to the mica,disc.l.jand,;to"provideja black radiation-absorbing surface facing the mirrorv 13..

v It is well understood byithose skilled inlthe part.

that the electromotive force; generated ,between the output conductors, andZ'l. by the eighttheb. mocouples connectedin series .,aiding relation will;

vary in magnitude with chan'ge'in therdifference between the temperature of the hot measuring junctions at the target 25 and, the temperature of the cold or reference, junctions, J8. ,When no radiation from an external source isdirectedto.

the hot junction target '25, itis' deslredthatlno electromotive force shall appear between output. conductors 26 and 21. sucha, result will always obtain when the hotv junctions and the .cold junc tions are at the ,samfl. ,l fimperatu fl- .However .1 those skilled intheaart have 0 18 bflenconcerrred with the problem of cmrfictingorxeliminathig voltage output whenthe thermopileis subjected, to a suddent change, in ambient temperature,

which produces a difference in temperaturefbetween the hot junctions and-thecoldjunctionsa More particularly, if the temperatureof thecold, junctions should suddenlyriseabove or fallbelow. the temperature of the hot junctions, there will. be produced a corresponding or appreciable. change in electromotive force, which change mayv disappear when conditionsequalize. However,.if

the temperature of the instrument is maintained at a level different from that of otherobjectsin its vicinity, for example, if it is partiallyin-cone. tact with a cold surface and partially in contact with a hot surface, there will be a. continuous flow of heat through the instrument, which can.

produce a sustained outputsignal, if the .coldzand hot junctions are not on the same isothermal. From the foregoing it will be understoodthataa transient error as well as asustained error,- may be introduced even when no, radiationqfroma hot body is incident on the hot junction targetor receiver 25. 'I'heforegoing factors do not ad-. versely affect operation of ,the' pyrometers disclosed herein by reason of the structural provisions including ring 20, itspins 2,|la-.-2|lc, housing [22] 21 and plate [24]14;.

When a thermopile isreceiVmg radiation.at constant rate and has reached .acondition. of

equilibrium, its hot junction target will be losing heat at exactly the same rate, asit is receiv-- ing it. Accordingly,-the manner iniwhichthe.

hot junction target loses heat is of prime importance. There are three modes or channels of heat-loss: (1) gaseous, by conduction and con.- vection; (2) conduction, through solid materials,

such as the thermocoupleelements themselves;

and (3) reradiation, fromthe target to the walls of the enclosing cavity and through the'window or opening Ila. The foregoing may be for a condition of equilibrium as follows:

' Er=Lg+Ls+Lr l J (1) where j Er=radiant energy input to the target i 1 s=gaseous losses L5=solid conduction losses Lr=radiation losses Both the gaseous losses and the solldconduction losses follow Newton's law of cooling, that is, the rate of each loss is'proportlonal to the temperaturedifference; in the one case proportional to mathematically Sam the temperature-difference between the target and the atmosphere within thethermopilehous' ing 2| which is substantially at the temperature of the cold junctions and at ambient temperature in general because of the action of the copper enclosure 2| and its heat-conducting mounting; and in' the other case proportional to the temperature-difference between'the hot junction target 25 and the ring which likewise is maintained substantially at ambient temperature by means of the restricted heat-conducting mounting including projections or legs 20a, 20b and 200 and'legs 14b. While the radiation loss is in accordance with the Stefan-Boltzman law, the usual range of temperature difference between the-hot junctions and the cold junctions. is generally' below 60, and the fraction of loss-by radiation is small. According1y','radiation loss canbe approximated by Newton's law of cooling instead of taking into account the difference between the fourth powers of the respective hot and cold junction temperatures (absolute).

Two other factors are'of importance in thermopiles particularly adapted to pyrometric tern;- perature measurement." First, the sensitivity should be high. 'For' greater sensitivity, there should be minimized the rate of transfer or loss of heat from the hot junction target Secondly, the speed of response should be high. Since this will depend upon the time required for the .hot junction target 25 to attain its final temperature, it is desirable to reduce its heat capacity. This is done by reducing its thermal The thermocouple wires are selected for high thermoelectric power, resistance to oxidation, uniformity of drawing and the constancy and reproducibility of their thermoelectric power characteristic. Thermocouples of Chromel and constantan alloys meet the foregoing requirements, .and'eve'n for small sizes, have adequate strength. Theislope of their temperature-electromotive vforce curve increases with increasing temperature. Heretofore, it has been necessary with Chromel-constantan thermocouples to utilize a compensating coil connected across the output wires of the thermopile in order completely to .compensate for change in ambient temperature. iWith a compensating coil arrangement,

the operation ofthe thermopile is' converted from an electromotive force measurement to a current measurement, that is, the drop of potential throughthe compensating coil is mea'sured'rather than the electromotive force of the thermopile. In contrast, in accordance with the present invention,.the electromotive force of the thermopile is measured; and when there is zero' temperature-difference between the hot junctions and the cold junctions, there is zero output across the output conductors; and when theoutput electromotive force is balanced b an equal electromotiveiforce from a potentiometer'circuit, there is zero current-flow in the circuit.

In accordance Withthepresent invention, the thermopile is completely compensated for change in ambient temperature without the use of a compensating coil by control of the flow of heat between the hot and cold junctions. .In accordance with one form of the present invention, there is utilized one or more Chromel-constantan thermocouples. Ihe Chromel alloy andtheconstantan alloy have positive temperature ,coeificients of thermal conductivity, thatis, as the temperature of each increases, it becomes a better conductor of heat? whereas, iron, nickel and,

in fact, most pure metals have negative tempercouple, a rise in ambient temperature causes a decreased rise in the temperature of the hot junctions over that of the cold junctions because of an increase in the conduction of heat by both the Chromel and the constantan. This decreased rise in the hot junction temperature relative to the cold junction temperaturewill develop with rising ambient temperatures, Whereas in the case of thermocouples including elements of' copper, iron or other relatively high heat-conducting elements in combination with antimony, the effect willbeto' decrease the heat-flow from the hot junction, and the hot junction temperature will rise faster than the cold junction temperature.

In one form of the present invention the fully compensated thermopile is provided'by the thermal shunt 28 which modifies and controls the flow of heat in such'manner as to make the operation of the thermopile substantially independent of ambient temperature change; Thus, there is avoided the need'of a compensating coil connected across the output conductors of the thermopile with its consequent disadvantages;

The thermal shunt functions independently of the electrical components and yet by its flowcontrol of heat modifies the output of theelectrical components in compensation for changes in the ambient temperature an'd'the like. In the absence of a' compensating coil and of a thermal shunt, the effect of the positive temperature cc'efiicient of heat conductivity, both for air'and for the solid conductors, will be to increase the relative amount of flow of heat from the hot junction region'25 to the cold junction region l8 with rise of ambient temperature. Such a disproportionate rise in thermal conductivity means that the actual temperature of the target 25 will be somewhat less than it should be in order to maintain the [necessary] temperature difference between the hot and cold junctions necessary to produce the desired electromotive force. However, by adding the thermal shunt formed by the nickel wire 28, the thermal impedanceas the ambient temperature rises does notdec'rea'se as much as before its addition. Accordingly, the temperature rise ofthe'hot junction area formed by the target 25 will be more nearly in keeping with ambient temperature rise lian it would otherwise be with only positive coefiicient paths, because of the proportionately less loss of heat through the nickel at the higher temperatures Accordingly, the electromotive force output will be higher. It is in this way that compensation is provided since it is only necessary relatively to proportion'the sizes of the Chromel and c'onstantan wires, with respect to the size of the wire '28 forming the thermal'shunt. While dimensions and detail design specifications will hereinafter bejpresented for typical modifications of the invention, an explanation will now be given of underlying theory 'and'a procedure by means of which the invention may beapplied to'a wide range'of materials with either positive or negative coefiicients of heat conductivityjand by means of which fully compensated thermopilesmay be produced in the absence of a separate element forming the thermal shunt, such thermocouples including equivalent heat conduction paths formed'by'elements of the" thermocouples themselves.

.9 Ijleequilibrium equation for the thermcpile has alreaqy been expressed'as.fellows? It is. well known that within the. temperature. range. of C. to 200.? C. the. condubtion'of.-heat over. thesolid paths, as through thethermocouple. elements, and. the conductionoL-heatthroiighthe. atmosphere can be represented byNewton's law. of cooling which states thatthe magnitude of the conducted heat. energy between any twoipointslin. a. thermal system is. dependent upon. the. ternperature-difference. between. saiditwo. points. It is. also well known that the thermal conductivity over the foregoing. temperature, range. or; most gases or. solids varies. in a'linear... fashion with change. in temperature. level. It. is. also well established that radiation. loss follows. Newton's law. and the linearvariation of thermal, heat ex.- change. for small. temperaturezdiiferenoes,. pro: vided the magnitude of the radiation'loss is small with respect to. the. solid. conduction losses and with respect to. the. gaseous conduction losses. Accordingly, any. term on the righthand. side. of: Equation 1 can be. written in the. form:

=M(1+mT) At (2) where 1 L=component loss M=thermal conduction of the [material of the particular path] heat-loss path at the. arbitrary reference level mr temperature coificient tivity T==ambient temperature above an arbitrary ref.-

erence level At:temperature-difference points of conduction of thermal conducbetween the two and substitutin equival t turns in Equa ion it may, Wltitten; N v

Where G=gas loss factor involving receiver area, di-

mensions of housing and physical properties of the gas in the housing g;temper ature cofficient of gas loss S=solid conduction factor involving length,

cross-sectional area and specific heat conduc- F tivity of solid supports of the receiver s: temperature coflicient of solid conductivity. R=radiation loss factor rb=temperature coefiicient of radiation loss eT=temperature difference between the hot and Y cold junctions The constants A and C may be determined from thethermo-electfic power characteristics of the thermocouple materials used, in the ex- 1-0 ample Chromel and con stantan, as. expressed. by

where.

AV=the voltage change for a small change in temperature, AT, and is generally expressed in microvolts, and.

Tzthe ambient temperature above the reference level hereinbefoije r eferredto.

More particularly, atthe reference level of 60 F., which for purposes. of.- calculation may be taken as T0, or T 0, the output AV may be taken as a zero reading, For. a change, a rise in temperature ofonedeg ree, theoutputwill increase by thirty-tln'eemi'cr cits; For' temperatures of one hundred degrees. angle; one hundred and one degrees, above the reference level, the outputs AV will be 3450 and, 3436, a difierence of thirtysix: microvolts. Hence, using the, data. at the reference level A will be found to'have a value 9f: 33- Using. the. data at the. high r. e el in Equa ion 5.;

The. ratio. of, AV- to, AT, with. the units, as, ab ve. given, represents. the change in, voltage, in micro; oolts per degree of change. intemperatwfe in degrees F, The constant A, accordingly, is defined in terms ofmicrovol ts per degfee'E; Since in the determination of C the term. (BE-33), is divided by temperature; 1003. 117., the constant, C in terms of microvolts per deg'i'ee per degree commonly written as- 'per degrees E1 Because later used, it is how woe noted that the BY b b n ne WY PPS 3 am; i: x res o qxy a om i gq AV: EIRL' A condition for AV to be independent of changes in ambient temperature ue, thefvalueofT) can be mathematically egip'ressed by'the equaclAV a Performing the mathematical operation upon Equation "fi'indic'ated by Equation 7, there "results,

ig+$:kRt

I o+s+e 5-? we r m 1. 1 a d. ie e'i e it it i (8 ease-. 5;

It is s n fr m Equation 21 that a th rmq le will b ambient temperature-indep ndent if the phy ical dimensions err/or physical cons entspi the materials of the"t ermopilehave'the proper values to satisfy the conditions set forth by Equa tion' '8. 1 In accordance "with the present invention, there will be hereafter disclosed the physical 11. dimensions and/or physical constants which will satisfy the conditions set forth by Equation 8.

The six unknown constants may be determined by solving six simultaneous equations of the general type of Equation 6 already set forth, and which are as follows:

NAEIRL For the several measurements of the voltage output of the thermopile a radiant input signal of known and constant intensity will be applied to the target of the thermopile. The measurement for Equation 6a will be made with the ambient temperature at a reference level of 60 F. It will be recalled that only the radiationreceiving face of the receiver or target 25 is blackened by the fused ceramic frit and, thus, the first measurement for Equation 6a will be made without change in structure.

The voltage output of the thermopile will then determine the magnitude of the term AV; of Equation 6a.

For Equation 6b the ambient temperature will be increased 100 F. to a new temperature of 160 Rand a new voltage output obtained which will determine the value of AVb.

The radiation receiver or target 25 will then be evacuated in conventional manner, the housing for the pyrometer already having been described as capable of being sealed for subatmospheric operation. For this measurement, the pressure should be less than one ten-millionths of an atmosphere. There will then be obtained AVc of Equation 60, for an ambient temperature at the reference level of 60 F. (which makes T= in Equation 6 and also makes G=0).

For Equation 6d the same measurement will be repeated with the thermopile under the same vacuum, but with an ambient temperature 100 F. above the reference level for determination 01' Ave.

The other face of the target or radiation receiver 25 will then be blackened, and with both faces thereof black, measurements are made at the reference level of 60 F. and 100 F. thereabove, as at 160 F., for determination respectively ofAV= and AV: 01' Equations 6e and 6f.

By blackening both faces of the radiation receiver 25, the radiation loss is made twice the value (2R, Equation 6c) of that with but one face of the receiver blackened. It may be ob served that while only one face of radiation receiver 25 received radiant energy, in all of the foregoing operations of known and constant intensity, the blackening of the surface of the target which does not receive radiant energy only afiects its emissivity and, hence, the blackening of the back face thereof doubles the radiation loss. r

. There will now have been determined 4W 1 each of Equations 6a to 6f.

Since [ER] Er in each of the six Equations 6a to St is known, the six simultaneous equations now contain only six unknowns, namely G, S, R, g, s, r, which can be determined by simultaneous solution. Because of selection of the units, each of the foregoing unknowns G, S and R will be ea:- pressed in terms of microwatts per degree F., while the unknowns g, s and r will have the dimensions of per degree F.

It will now be assumed that a thermopile of suitable construction has been tested and the values of G, S, R, E, s and r determined. It will be further assumed that the thermopile is not compensated, i. e., its output varies with change in ambient temperature. It will be recalled that for ambient temperature-independence, Equation 8 must be satisfied. In accordance with the present invention, Equation 8 may be satisfied by adding to the thermopile, a thermal shunt. Further assuming that C and A have been determined for the thermopile in question, the addition of the thermal shunt of conductance S and coefiicient of thermal conductivity s of proper sign and known value will supply terms to Equation 8 which will meet its requirements, for example,

Since-the only unknown in Equation 9 is S (s' being known for the selected material), it can be ascertained. The resultant conductance S is related to the dimensions of the thermal shunt of one or more paths and to its conductivity, that is to say:

where N=number of shunt paths P=conductivity of the material thermal shunt A=cross-sectiona1 area of each shunt path L=length of one shunt path Accordingly, the dimensions of the thermal shunt may be readily ascertained.

Pursuant to the foregoing, the thermopile in Figs. l-5 was provided with a thermal shunt formed by a nickel wire of .005 inch in diameter, the Chromel-constantan wires each being made of wire 2 mils in diameter flattened to 1 mil thickness. The larger diameter of the pins 20a, 20b and 200 extending from the ring 20 was .049 inch, the length thereof from the ring to the, plate l4 bein .106 inch. It is to be observed there is an air space between the ring 20 and the heatconductive housing 2|, and between the pins 20a20c of that housing. Thus, substantially all of the heat flow is limited from the ring 20 through" the solid heat-conduction paths provided by the "enlarged ends of pins 20a20c. Though the diameter or cross-sectional area of pins 20a20c can be varied to some extent, it is desirable to have its dimensions of the same order as those herein set forth in relation to the other dimensioning of the associated parts. Thus, the provision of the restricted heat-conducting paths provided by pins 20a-20c, together with the high heat-conductivity housing 2| including its high heat-conductivity enclosure Zla, has been found to minimize and to overcome substantially entirely rapid or transient effects of the change in ambient temperature, more particularly to prevent-the appearance of transient voltages at-the output of the thermopile, due to. such rapid changes.

In this connection it is to be observed that the housing 2| is conductively related to the outer housing 22 through its heat-conductive association with the plate I4. Reference to the dimensioning of the spokes thereof has already been referred to, and it will be recalled that they provide heat-conductive paths between the housin 2| and the outer housing 22, and thereby limit or make lower the rise of temperature of the housing 2| and the associated assembly when the pyrometer is subjected to an intense radiation field. Because of the physical construction referred to, the spokes of the plate l4 may be made larger than would otherwise'be the case to provide better heat-conducting paths. The thickness of the housing 2| is determined in relation to the diameter of the .pins 20a-2llc and the dimensioning of the ring 20. Since the effect upon the thermopile and the function thereof in avoidance of transient voltages due to rapidly varying changes in ambient temperature is dependent upon the interrelation of these parts, it is to be understood that they maybe varied so that different dimensions can be utilized for each, other than those in a typical embodiment where the housing 2| was made of copper of .025 inch thickness, the ring 20 of .049 inch radially and .094 inch axially. More particularly, if the thickness of housing 2! beincreased, the diameter of the legs Mia-20c will be increased, or they may remain the same diameter, in which case the thermal capacity of the ring 20 will be decreased as by axially shortening it or making it of less radial thickness, or the materials may be selected in terms of their relative heat conductivity and heat capacity.

It is believed it will now be helpfulto present a numerical example which will be done in terms of the modification of Fig. 8, constructional details of which will be later presented and which comprises eight Chromel-constantan thermocouples of 3-mil diameter wire size. The constants G, S, R, g, s, r, C and A, determined as above indicated are as follows:

(3:129 microwatts per degree F.

8:129 microwatts per degree F.

- =53 microwatts per degree F.

g=.GGO9 [microwatt per degree] per degree F.

s=.002 [microwatt per degree] per degree F.

r=.005 llmicrowatt per degree] per degree F.

%=.0009 [microwatt] per degree F.

Inasmuch as the Chromel-constantan thermopile will not be compensated, it will be assumed that copper will be selected for the thermal shunt.

Accordingly.

e: .0000? [microwatt rwr degree] per degree The foregoing values may be inserted in Equation 9 since that equation expresses theconditions which must be met to provide a thermopile whose cutput will. not vary with change in ambient temperature. The only unknown from Equation 9 will be S. Solving Equation 9 for the value of S, there will be obtained:

S' lhi microwatts per degree F.

It is now only necessary to substitute the value of S in Equation 10 since it can be arbitrarily decided that four thermal shunt paths. will. be v lp FQfihB qnductivityof t e c pper from which it hasbeendecided-to construct thermal shunt being known, i. e., P=4.8 watts 14 per square inch per inch per degree F., andthe length of each shunt path being known, since it will be the distance betweenthe hot and cold junctions of each thermocouple, specifically .25 inch. Accordingly, solving Equation 10 for A,

the area, there will be obtained:

, A.=2.3 l0 square inches this corresponding with a wire'size of 1.7 mils diameter. In practice, it has been found that thermal shunt of 2 mil diameter is-highly satisfactory and avoiding any change upon the output of the thermocouple due to change in ambient temperature between the range of 60 F. and )5. Thus, there is close correlation between the mathematical explanation of the underlying theory and the experimental verification thereof. In terms of Fig. 1, and further substantiating the underlying theory of the present invention, the constants applicable to the" thermopile' of- Figs. 1-5 comprising the eight thermocouples, each of Chromel-constantan flattened 2-mil wires, the constants G, R, g, s, r and j will be the same as set-forth above. Constant S will be equal to 57 microwattsper .degree F.

In accordance with the invention, the use of a nickel shunt will be preferred inasmuch as it has a more negative temperature coeflicient of conduction than thecopper and, hence, will adequately perform its corrective function w'ith a smaller conducting factor than copperjand, thus, there will be less loss ofheat due to the provision of the thermal shunt of nickel instead of copper. Accordingly, the constant s for. the nickel shunt will be equal to -.O00 2 [microwatt per degree] per degree F) In accordance with Equation 9, S may bedetermi n'edg to be.7.7.3 microwatts per degree F. p

A indicated in Figs. 1-5,, there will be two shunt paths each having a length of .25 inch, and the specific conductivity of' the nickel will be equal to .746 watt per sq. inch por -inch per degree F. Solving Equation 10, for the area A, there will be obtained a value of l 2. 2 10- square inches, this corresponding with the nickel shunt of approximately 4 mils diameter-[s]. Inpractice, a 5 mil diameter nickel shunt was found fully to compensate the thermopile or to provide a voltage output independent of change in. ambient temperature.

Applying the foregoing analysis to the modifi= cation of Fig. 7, the calculations indicated an area for the thermal shunt equal to 4=.2 1(J square inches. In practice this area was provided by using two 2-mil nickel shunts andtwo l-mil nickel shunts, the total cross-sectional area of-the nickel shunts being 3.9 10- square inches, thus showing a further confirmation-0'1 the theoretical basis for the present invention. It will be recalled that a greater'sensitivity of the thermopile is obtained by reducing the transfer or loss of heat from the hot junction area or target 25. A thermopile which has high sensitivity will ordinarily comprise conductors having small cross-sectional areas, and later exam-' ples will be given of thermopiles constructed of wires of l-mil diameter and less. After selection of the small diameter wires for high sensitivity and'after decision as to the materials toyiel d high thermoelectric power for maximum-output of the thermopile, the'condition'sset forth by Equation 8 provide for the further dimensioning iro oi the. c nduc ors included. in h th ilo. to P o ide the amb en e em raharaeteristics. It will be observed that E uat on 8 shows the ratio f t m o the spective heat loss factor (G+S-|-R) to the sum of the respective product of the heat loss factors and the respective temperature coefficients of thermal conductivity (Gg+Ss+Rr) is equal to, or shall at least approximate equality with, the ratio or the temperature-voltage constant A to the temperature-voltage constant C, which constan s are. o course. d fined by Eq n For some applications, the speed of response of the thermopile of the modification of Figs. 1P5 may be higher than desired. In such event, additional discs may be attached to the target 25 to add thermal mass. However, for thermopiles of greater speeds of response than of the type show in. Eise- ..1,6, the m ficat on of Fi 7 may be utilized where; the speed of response will be greatly increased ever that of the earlier modmel alloy one mil in [thickness] diameter fiattened to one-half mil [in diameter] thickness, is brazed or spot welded to a, small target or radiation receiver 25a of nickel, one-tenth mil in thickness and squalf i. twenty mils on a side.

The opposite end of the wire I9 is spot welded to: the end of conductor 26 which is itself secured to the face of ring by a ceramic frit in manner described in connection with Fig. l. The other element oi the thermocouple comprises the, wire I9a oi constantan likewise comprising a wire flattened to one-half mil from an original diameter of one mil. One end is spot welded to the target a and the other end spot welded to the conductor 21 which is similarly secured to the face-oi ring 20 by a ceramic frit.

I In accordance with the invention as already explained, there is provided a thermal shunt which achieves operation of the radiationeresponsiye device unaffected by change in ambient temperature. With a ring 20 of .2567 inch [in diameter] radius, the same as for the modification of Figs. i=5, the thermal shunt comprises two con ductive paths 28, 28 formed by a single nickel wire two rails in diameter and extending diametrically aQIQBs ring 211 with the central portion thereof spotwwelded to target 25a. The thermal shunt also includes conductive paths 28a, 23a formed by a second wire of nickel one mil in diameter and similarly extending diametrically across ring 20 with the mid-portion spot welded totarget 25a. Since there is provided welded construction at the target 25a, the ends of the wires forming the heateconductive paths are secured to the upper face of ring 20 in intimate and good heat-exehangingrelation therewith by the ceramic f-rit wh ch l ke ise prov de e e tri al nsu a n botwoen e w res, an he ring 2 Though the ends of the wires 2 and 28a could be secured in electrically conductive relation to ring 2!), the nsula d ar n e out ref rr n order o electrically to ground any part of the sensitive element of the heat-responsive device. Such a heat-responsive or radiation-responsive device has :been iound satisifactorily to function with an output independent of an ambient temperauro wh c cha ges within he ran e of from 60 a n o 1 ification. As shown in Fig. 7, a wire iii of Chroapplied to the target 25a the ture of the hot junctions will be reater due to. loss c ndu ti n of heat there rom Since both the Chromel and constantan wires have large positive temperature coefficients of heat conductivity, any rise in their temperature as by change in the ambient, increases their heat conductivity and tends to increase the effectiveness of the heat path provided by them. However, because of the small diameter of each wire the effect is of a low order, much less than it would be if the wires were of three mil diameter since there is a high thermal impedance established by the small cross-sectional area of each path. Accordingly, it will .be seen that the degree of compensation needed will be corresponding-ly less and, therefore, the thermal shunts provided 'by the wires 28 and 28a, which are need-ed, are of Small cross-sectional area and there is a relatively small transfer of heat through the thermal paths which means that there is a relatively small loss of sensitivity by reason of the inclusion of the thermal shunts.

A heat-responsivedevice of the type shown in Fig. '7 has also been constructed utilizing Chromel-constantan wires of one-half mil di'. ameter flattened to approximately one-quarter mil thickness. In accordance with the present invention, a thermal shunt of a single wire, such as the wire 28 of Fig. '7, of nickel, of two mil diameter has been found satisfactory in producing an output which is independent of change in ambient temperature through the aforesaid range.

Referring to Fig. 8, the ring 20 is made some.- what larger than the corresponding ring of earlier modifications and in Fig. 8 the ring 20 is made of constantan.

Extending radially from a central target 25 are a plurality of constantan wires 53 which are attached at their opposite ends to the ring and to the target as by a ceramic frit in manner already described. There is welded to the inner end of each constantan wire 50 an end of Chromel wires 5| which are respectively soldered at their opposite ends to the constantan wires 552 at points spaced inwardly from the ring 28, each Chromel wire being of course connected at one end to one radial wire and at the opposite end to the adjacent radial wire of constantan. The soldered connections at the outer ends of constantan wires 5! are made relatively massive as by using a fairly large drop of solder to assure good heat' mocouple assembly. of the leads 52 and 53, of course, form a cold Junction.

Pursuant to the present invention, with Chromel-constantan wires of three mil diameter the output of the thermopile may be made independent of change in ambient temperature by providing in manner already described four thermal shunts 54, 55, 5B and 5! each formed by copper wire of two mil diameter. These thermal shunts form heat-conduction paths between the receiver 25 and alternate cold junctions of the thermopile.

Referring to Fig. 9, the ring 20 may correspond with the similarly numbered ring of the modi fication of Figs. 1-5. In this modification of the invention, radially extending thermocouple elements 50 are formed by constantan wires of three mil diameter. They are secured to the target 25 in the same manner as described in connection with Fig. 8 with their opposit ends secured to the ring 20 by ceramic frits.

In Fig. 9 there are four wires 5! of Chromel each three mils in diameter, and there are four wires 55 of copper of two mils diameter. Hence, it will be seen that there are heat-conduction paths between the hot junction of the target 25 and the cold junction of ring 20 through three differing materials; that is through the Chromel wires, the constantan wires and the copper wires. In accordance with the invention, there has been combined in the modification of Fig. 9 the function of the thermal shunt with wires forming a part of the electrical system of the thermopile. That is to say, the addition of the copper wire to the Chromel-constantan combination provides the heat-conducting paths which result in an output of the thermopile independent of change of ambient temperature over the aforesaid range of from 60 F. to 160 F. without the need to provide thermal paths separate and independent from the electrical network. The underlying theory" applicable to Fig. [9] 4 which has already been set forth at length is equally applicable to the modification of Fig. 9, the copper wires then being considered as the paths providing the thermal shunt referred to in said theory.

Reference has already been made to the fact that by decreasing the diameter of the wires forming each thermocouple, the impedance of the heat-conducting path therethrough is greatly increased. If the wires are made quite small, the

impedance will limit to a negligible degree the amount of heat which may be conducted therethrough. Stated differently, the heat losses due to heat conduction through the wires will then be of minor importance compared with the gaseous losses. If the thermopile be constructed of fine wires such as eight two-mil diameter copper-constantan thermocouples, there will be over-compensation becausemost of the heat will then flow through the copper, having a temperature coefiicient of conductivty of negative sign. That is to say, the output, with change in ambient temperature of such a thermopile, will increase with increase in ambient temperature. In accordance with the present invention, the thermopile of Fig. 9 will be fully compensated by utilizing constantan wires 50, each of three-tenths mil diameter, and using all associated thermocouple wires of copper of three-tenths mil diameter, the copper providing the thermal-shunt corrective action. If finer wires than the three-tenth mil diameter were used the output voltage will decrease with increase of ambient temperature.

The thermal shunt or equivalent heat-conducting path will be equally effective for thermopiles where the gaseous losses are of major magnitude. The gaseous losses increase with rising ambient temperature and, hence, the conductive path by the thermal shunt will be in the opposite direction.

Referring to the modification of Fig. 10, it will be observed there is provided a ring 00 and that there extend across the ring a plurality of thermocouple elements 6i which, in conjunction with the diagonally extending thermocouple elements 62, form differentially connected thermocouples. This is to say, the diagonally extending elements 62 are secured to the elements 0! and as by a ceramic frit to'an underlying target or radiation receiver 63 Whereas the opposite ends of said diagonally extending wires are likewise secured to adjacent wires GI and by a ceramic frit to a second target or radiation receiver 64. Lead wires 55 and 66 are connected to the outermost Wires 6|, a voltage appearing across the output of the thermopile whenever the temperature of the junctions at target 53 differs from the temperature of the junctions at the target 64. In order to make the output voltage across the leads 65 and 56 independent of change in ambient temperature there can be provided in accordance with the present invention thermal shunts S1 and 88, thermally interconnecting the targets 63 and E4, the thermal conduction factor of the thermal shunting paths 61 and 08 being determined in generally the same manner as above set forth.

It is also to be understood that individual thermocouples in the differential thermopile may be selected of materials which will of themselves perform the functions of the thermal shunts 61' and 68 in manner already set forth in connection with Fig. 9.

Reference has already been made to the desirability of selecting a thermal shunt with the proper sign for the coefficient of thermal conductivity. For example, where the solid heatconduction paths formed by the thermocouple elements both have temperature coefficients of conductivity which are positive in sign, a thermal shunt having a coefficient of negative sign will ordinarily be indicated. Conversely, in a thermocouple such as copper-constantan, copperantimony or bismuthantimony, the thermal shunt will preferably be made of a material having a positive coefficient of thermal conductivity and thus may be made of constantan, Chromel, or the like.

What is claimed is:

l. A. temperature-responsive device comprising at least one thermocouple, means supporting said thermocouple with its hot junction disposed for application of heat thereto, and a thermal shunt electrically insulated from said hot junction but thermally intimately connected thereto and forming at least one heat-conduction path from said hot junction to the cold junction of said thermocouple.

Z. The combination set forth in claim 1 in which said thermocouple is of Chromel and constantan elements and the thermal shunt is of nickel.

3. The combination set forth in claim 2 in which the Chromel-constantan thermocouples are formed of Wire approximately .002 inch in diameter and flattened, and the thermal shunt comprises an element of nickel having a cross section equal to that of nickel wire .005 inch in diameter.

4. The combination set forth in claim 2 in which the Chromel-constantan wires are relatively small about one mil diameter and the thermal shunt consists of nickel the cross-sectional area being approximately four times the cross-sectional area of the conductors of the thermocouple.

5. A thermopile comprising a central target, supporting structure spaced therefrom, at least three thermal conductivity paths extending between said target and said structure, two of said paths including a thermocouple having its hot junction thermally connected to said target and its cold junction thermally connected to said supporting structure, a. solid heat conductor forming a third path and having a heat conductivity cor 'efiicient of sign opposite to that of one of the paths formed-by one of the wires of the thermocouple for compensation due to changes in ambient temperature.

6. A radiation pyrometer including a dia phragm having an opening and a magnesium oxide coating on one face of the diaphragm to render more visible an image of the area from which radiation is directed to the pyrometer.

7. In an instrument of the radiation pyrometer type, a diaphragm having an opening and a mag nesium oxide coating on a face of diaphragm to render more visible from an end of the instrument opposite that receiving radiant energy in the infra-red portion of the spectrum an image of the area from which radiation is directed to the instrument.

8. A thermopile comprising thermocouple clements joined to form hot junctions at predetermined distances from cold junctions, the crosssectional area andlength of at east one or said elements being selected so as to provide low heat loss and high sensitivity of said thermocouple and which forms a first heat conduction path whose thermal conductivity changes ambient temperature, at least another element of the thermopile being dimensioned as to area and length for change in its thermal conductivity in a direction and to such an extent that the resultant not change in the thermal conduction between said hot junctions and said cold junctions is controlled in magnitude and in sense to produce within a wide range of ambient temperature variation a voltage output from said thermo- .pile which does not vary with such changes in said ambient temperature regardless of nonlinearity of the curve of electromotive force vs.

temperature for the thermoelectric comprising said thermocouple.

9. A thermopile comprising thermocouple elements joined to form hot junctions at prede termined distances from the cold junctions, the cross-sectional area and length of said elements being so selected as to provide the desired sensitivity of said thermopile, the heat conduction paths of said elements changing in thermal conductivity with change in ambient temperature, a thermal hunt extending between said hot junctions and said cold junctions, said shunt being of material for change in thermal conductivity in a direction opposite to that of said thermocouple elements and dimensioned as to area and length for change in thermal conduction to such an extent to compensate for said change in said thermal conduction of said thermocouple elements to produce substantially constant voltage output from said thermopile with constant input thereto for wide changes in said ambient temperature.

10. A heat-sensitive element for measuring ap paratue, said element comprising two regions both varying in temperature in response to variations in ambient temperature, a plurality of paths by means of which heat flows from one to the other of said regions, two of said paths being of solid heat-conducting material having a temperature coefficient such that both of said paths tend to conduct more heat with a rise in ambient temperature, and a third path for the interchange of heat between said two regions, the material of said third path having a temperature coefficient such that said till :i path conducts less heat with a rise in ambient temperature, the dimensions of said third path being such that the net conduction of heat by all of said plurality of paths is of such value that an electrical characteristic materials with change in tially'constant over a substantial range of am"- bient temperature variation.

11. A heat-sensitive element for measuring apparatus, said element comprising two regions both varying in temperature in response to variations in, ambient temperature, a plurality of paths by means of which heat flows from one to the other of said regions, two of said paths beingofi solid heat-conducting material having, a temperature coefficient such that both of said paths tend to conduct less heat with'a rise in ambient temperature, and a third path for the interchange of. heat between said two regions, the material of said third path having a temperature coefiici ent such that said third path conducts more heat witha rise in ambient temperature, the dimensions of said third path being such that the netv conduction of heat by all of said plurality of pathsis of such value that'an electrical characteristic of said sensitive-elementis maintained substantially constant over a substantial range of. ambient temperature variation.

1,2. A heat-sensitive element for measuring apparatus, said element comprising a heat-receiving region and a cooler region displaced from said heat-receiving region, both regions varying in temperature in response to variations in ambient temperature, a plurality of paths by means of which heat flows between said heat-receiving region and said cooler region, two of said paths being of solid heat-conducting material having a temperature coefficient such that both of said paths tend to conduct more heat with arise in ambient temperature, and a third path for the interchange of heat between'said two regions, the-material of said third path having a temperatureicoeificient such that said third path. conducts less heat with a rise'in ambient temperature, the dimensions of said third path being such that the net conduction of heat by all of said plurality of paths is of such value that an electrical characteristic of said sensitive element is maintained substantially constant over a substantial range of ambient temperature variation.

13. A heat-sensitive element for measuring appara-toe, said element comprising two radiant energy receiving areas varying in temperature with variations in ambient temperature, a plurality ofv paths for heat to flow from one area to the otherand to a source of reference temperature common to all said paths including wires coasting with said receiving areas, some of said wires being of a material, the thermal conductivity of which increases with a rise in ambient temperature, and other of said wires being of a material, the thermal conductivity of which decreases with a rise in ambient temperature, the dimensions of the heat path formed by said other of said wires being such that the net conduction of heat by all of said plurality of paths is of such values that an electrical characteristic of said sensitive element is maintained substantially constant over a substantial range of ambient temperature variation.

14. A heat-sensitive element for measuring apparatus, said element comprising two regions varying in temperature with change of ambient temperature, a plurality of solid conductive paths for heat to flow from one region to the other and to a source of reference temperature common to all, some of said paths being of a solid material size less than 5 nil diameter wire, the thermal conductivity of which increases with a rise in ambient temperature, and other of said paths being of a material of size differing from said firstnamed paths, the thermal conductivity of which decreases with a rise in ambient temperature, the dimensions of the heat path formed by said other of said paths being such that the net conduction of heat by all of said plurality of paths is of such value that an electrical characteristic of said sensitive element is maintained substantially constant over a substantial range of ambient temperature variation. i

15. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising conductors extending between said area and said support to form heat-conduction paths including at least two paths formed of materials having differing thermoelectric power and connected to form a hot junction at said heat-receiving area and a cold junction at said support, said conductors having diameters small to minimize loss of heat through their solid heat conduction paths, heat also being lost other than through said solid conduction paths as through a radiation path, said paths being effectively dimensioned to anproximate equality between a first ratio of the sum of the respective heat loss factors of all heat loss paths to the sum of the respective products of each of said heat loss factors and the respective temperature coefficients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C, said constants A and C having values which satisfy the equation AV:(A+CT) AT where AV: the voltage change for a small change in temperature, AT, and T:the ambient temperature above a reference level.

16. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising conductors extending between said area and said support to form heat-conduction paths including at least two paths formed of materials having differing thermoelectric power and connected to form a hot junction at said heat-receiving area and a cold junction at said support,

said conductors having diameters small to minimize loss of heat through their solid heat conduction paths, heat also being lost other than through said solid conduction paths as through a radiation path, said conductors being dimensioned to approximate equality between a first ratio of the sum of the respective heat loss fac tors of all heat loss paths to the sum of the respective products of each of said heat loss factors and the respective temperature coefficients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C. said constants A and C having values which satisfy theequation AV: (A-i-CT) AT where AV:the voltage change for a small change in temperature, AT, and

T:the ambient temperature above a reference level.

17. A heatsensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising conductors extending between said area and said support toform heat-conduction paths including at least two paths formed of materials having differing thermoelectric power and connected to form a hot junction at said heatreceiving area and a cold junction at said support, said conductors having diameters small to minimize loss of heat through their solid heat conduction paths, heat also being lost through a gaseous path and a radiation path, said paths being effectively dimensioned to approximate equality between a first ratio of the sum of the respective heat loss factors of all of said heat loss paths to the sum of the respective products of each of said heat loss factors and the respective temperature coeflicients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C, said constants A and C having values which satisfy the equation AV: (A-l-C'I') AT where 18. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising conductors extending between said area and said support to form heat-conduction paths including at least two paths formed of materials having temperature coefficients of thermal conductivity of opposite sign and having differing thermoelectric power and connected to form a hot junction at said heat-receiving area and a cold junction at said support, said conductors having diameters small to minimize loss of heat through their solid heat-conduction paths, heat also being lost other than through said solid conduction paths as through a radiation path, said paths being effectively dimensioned to approximate equality between a first ratio of the sum of the respective heat loss factors of all heat loss paths to the sum of the respective products of each of said heat loss factors and the respective temperature coeiiicients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C, said constants A and C having values which satisfy the equation AV: (A-l-CT) AT where AV:the Voltage change for a small change in temperature, AT, and

T:the ambient temperature above a reference level,

' 19. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising conductors extending between said area and said support to form heat-conduction paths including at least two paths formed of materials having differing thermoelectric power and connected to form a hot junction at said heatreceiving area and a cold junction at said support-and at leastone additional heat-conduction path electrically insulated from said junctions, the coeflicient of at least one of said paths having a sign opposite to the sign of the temperature coefficient of at least one of the remaining paths of heat-conduction, said conductors having diameters small to minimize loss of heat mi? through their solid heat conduction paths, heat also being lost other than through said solid conduction paths as through a radiation path, said conductors being dimensioned to approxi mate equality between a first ratio of the sum of the heat loss factors of each heat loss path to the sum of the respective products of each of said heat loss factors and the temperature coefiicients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C, said constants A. and C having values which satisfy the equation AV: (Al-CT) AT .where AV=the voltage change for a small change in temperature, AT, and

T=the ambient temperature above a reference level.

20. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprisingconductors of differing thermoelectric power connected to form a hot junction at said heatreceiving area and a cold junction at said support, said conductors having diameters small to minimize loss of heat through their solid heat conduction paths, heat also being lost through a gaseous path and a radiation path, a conductor extending between said area and said support forming a heat-conduction path having a temperature coeiiicient of thermal conductivity opposite to at least one of said first-named conductors, said paths being eifectively dimensioned to approximate equality between a first ratio of the sum of the respective heat loss factors'of all of said paths with respect to the sum of the respective products of heat loss factors and the respective temperature coefficients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C, said constants A and (3 having values which satisfy the equation AV: (A+CT) AT where AV=the voltage change for a small change in w temperature, AT, and

T=the ambient temperature above a reference level.

21. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heat receiving area to an adjacent support, comprising electrical conductors of difiering thermoelectric power connected to form a hot junction at said heat-receiving area and a cold junction at said support, said conductors having diameters small to minimize loss of heat through their solidheat conduction paths, heat also being lost through a gaseous path and a radiation path, a thermal'shunt comprising a heat conductor forming a heat-conduction path extending between said area and said support formed of a. material having a coemcient of thermal conductivity of sign opposite to that of at least one of said conductors, said conductors and said shunt being effectively dimensioned to approximate equality between a first ratio of the sum of the respective heat loss factors of all said paths with respect to the sum of the respective products of the heat loss factors and the respective temperature coefficients of thermal conductivity of each of said paths'anda second ratio of a temperature-voltage Z14 constantA to a temperature-voltage constant C, said constants A and C having values which satisfy the'equation AV: (A CT) AT Where AV=the voltage change for a small change in temperature, AT, and

T=the ambient temperature above a reference level.

22. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising non-ferrous conductors extending between said area and said. support to form heat-conduction paths including at least two paths formed of materials having differing thermoelectric power and connected to form a hot junction at said heat-receiving area and a cold junction at said support, said conductors having diameters small to minimize loss of heat through their solid heat conduction paths, heat also being lost other than through said solid conduction paths as. through. a radiation path, said paths being effectively dimensioned to approximate equality between a first ratio of the sum of the respective heat loss factors of all heat loss paths to the sum of the respective products of each of said heat loss factors and the respective. temperature coefiicients of thermal conductivity of each of said paths and a second ratio of a temperature-voltage constant A to a temperature-voltage constant C, said constants and C having values which satisfy the equation AV: (A-l-CT) AT where AV=the voltage change for a small change in temperature, AT, and

T=the ambient temperature above a reference level.

23. A heat-sensitive device of high sensitivity characterized by low loss of heat from a heatreceiving area to an adjacent support, comprising conductors extending between said area and said support to form heat-conduction paths including at least two paths formed of materials having differing thermoelectric power and connected to form a hot junction at said heat-receiving area and. a cold junctionat said support, at least one of said conductors being a ferrous metal, a thermal shunt extending between said hot junction target area and said cold. junction support, said conductors having diameters small to minimize loss of heat through their solid heat-conduction paths and through said thermal shunt, heat also being lost other than through said solid conduction paths as through a radiation path, said paths and said thermal shunt being effectively dimensioned to approximate equality be tween a first ratio of the sum of the respective heat loss factors of all heat loss paths to the sum of the respective products of each of said heat loss factors and the respective temperature AV=(A+CT) AT where AV=the voltage change for a small change in temperature, AT, and

25 T=the ambient temperature above a reference level.

electrically insulated from said hot junction butv thermally intimately connected thereto and forming at least one heat-conduction path from said hot junction to the cold junction of said thermocouple, a sub-housing forming an enclosure for said at least one thermocouple and said thermal shunt, said sub-housing being formed of a material having high thermal conductivity, heat-conductive structure supporting the cold junctions of said thermocouple, said heat-conductive structure having projections of limited cross-sectional area engaging said sub-housing for support therefrom to limit the heat-conduction path therebetween, a main housing enclosing said sub-housing, and supporting structure carried by said main housing and having extensions of' limited cross-sectional area disposed in heat-conductive relation with said sub-housing, the cross-sectional areas of said heat-conductive paths, the mass of said heat-conductive structure and the thickness of said sub-housing being selected for [the same] change in tempera-- ture [as the hot junctions and of the cold junc- 25. A temperature-responsive device comprising at least one thermocouple, means supporting said thermocouple with its hot junction disposed for application of heat thereto, a thermal shunt electrically insulated from said hot junction butthermally intimately connected there to and forming at least one heat-conduction path from said hot junction to the cold junction of said thermocouple, a sub-housing forming an enclosure for said at least one thermocouple and said thermal shunt, said sub-housing being formed of a material having high thermal conductivity, heat-conductive structure supporting the cold junctions of said thermocouple, means supporting said structure in inwardly spaced symmetrical relation to said inner housing, said heat-conductive structure having projections of limited cross-sectional area engaging said sub-housing for support therefrom to limit the heat-conduction path there-between, sub-housing and supporting structure carried by said main housing and having extensions of limited cross-sectional area disposed in heatconductive relation with said sub-housing, the cross-sectional areas of said heat-conductive paths, the mass of said heat-conductive structure and the thickness of said sub-housing being selected for [the same] change in temperature [as the hot junctions and of the cold junc tions] of the cold junctions in the same manner as change in temperature of the hot junctions as respectively affected by changes in ambient temperature.

26. A temperature-responsive device comprising at least one thermocouple, means supporting said thermocouple with its hot junction disposed for application of heat thereto, a thermal shunt electrically insulated from said hot junction but thermally intimately connected thereto and forming at least one heat-conduction a main housing enclosing said Lil 26 4 path from said hot junction to the cold junction of said thermocouple, a sub-housing forming an enclosure for said at least one thermocouple and said thermal shunt, said sub-housing being formed of a material having high thermal conductivity, heat-conductive structure supporting the cold junctions of said thermocouple, means supporting said structure in inwardly spaced symmetrical relation to said inner housing comprising extensions of relatively small cross-sectional area extending [radially] axially of said [inner housing] sub-housing, said heat-conductive structure having projections of limited cross-sectional area engaging said subhousing for support therefrom to limit the heatconduction path therebetween, a main housing enclosing said sub-housing and supporting structure carried by said main housing and having extensions of limited cross-sectional area disposed in heat-conductive relation with said sub-housing, the cross-sectional areas of said heat-conductive paths, the mass of said heat-conductive structure and the thickness of said sub-housing being selected for [the same] change in temperature [as the hot junctions and of the cold junctions] of the cold junctions in the same manner a change in temperature of the hot junctions as respectively afiected by changes in ambient temperature.

27. A temperature-responsive device comprising at least one thermocouple having a hot junction and a cold junction, means supporting said thermocouple with its hot junction disposed for application of radiant heat thereto and for controlling temperature of the hot junction and of the cold junction so that they change in like manner with transient changes in ambient temperature comprising a sub-housing of good heatconducting material surrounding said thermocouple, and heat-conductive structure in large part spaced from said sub-housing for supporting the hot junction in radiant energy-receiving relation to an opening through said sub-housing, said heat-conductive structure having extensions forming heat-conducting paths between it and said sub-housing, the dimensions and masses of said heat-conducting structure, said housing and of said extensions controlling the temperature changes of said cold junction in response to change in ambient temperature in like manner with change in temperature of said hot junction with change in ambient temperature.

28. A temperature-responsive device comprising a group of thermocouples, a heat-conductive structure, having an opening at the center, supporting the hot junctions within the opening and supporting the cold junctions from said heatconductive structure in good thermal relation therewith, a sub-housing of good heat-conducting material enclosing said thermocouples and said structure, projections of limited area and of good heat conduction interconnecting said structure and said housing, one wall of said sub-housing having an opening for admission of energy to said hot junctions through a path avoiding said heat-conductive structure, the cross-sectional areas of said extensions being selected with reference to the mass of said heat-conductive structure and with respect to the mass of said [inner housing] sub-housing for regulation of the temperature change of said cold junction in manner corresponding with temperature change of said hot junctions due solely to change in ambient temperature.

29. The combination set forth in claim 28 in which said wall comprises a diaphragm coated with magnesium oxide to render more visible an image of the area from which radiation is directed to said thermocouples.

. 30. In a radiant energy sensing device having a detector element to be sighted upon a source of radiant energy for measuring the intensity thereof, a diaphragm with an opening for passage of radiant energy to a receiving area of said detector element, an optical means, including a Window of heat-resistant material which passes radiant energy in the visible and in the infra red portion of the spectrum, for projecting an image of said source toward said opening and upon said diaphragm, an image-receiving area of said diaphragm having a coating of magnesium oxide to {enhance the Visibility]! make visible said image [range] of said source for accurate sighting of said device.

31. A heat-sensitive device comprising heatreceiving areas spaced one from the other, thermocouples connected in electrical opposition with the hot junctions thereof at one area and the cold junctions at the other area, and solid heatconductive means interconnecting said target areas for interchange of heat, said heat-conductive means having a size and a temperature coefficient for controlling the interchange of heat between said areas to maintain unchanged the output from said thermocouples upon change only of ambient temperature.

32. An ambient temperature independent thermopile comprising means to support a plurality of electrically conductive elements substantially parallel to each other, a plurality of other electrically conductive elements connecting said first-mentioned elements to form a group of differentially connected thermocouples with two distinct groups of junctions, substantially symmetrical target areas associated with each group of junctions and auxiliary solid heat-conducting means connected between said areas and of size and temperature coefficient to control the interchange of heat between said areas thermally to maintain unchanged. the output from said group of thermocouples upon change only of ambient temperature.

33. A heat-sensitive device comprising a heatsensitive element having an electrical characteristic varied in accordance with change in the temperature of the element, the change in the electrical characteristic for equal changes of temperature when the heat-sensitive element is ata higher temperature being materially greater than when at a lower temperature, a support adjacent said element, solid heat-conduction paths between said element and said adjacent support including conductors forming electrical connections thereto, said solid paths having a small total cross-sectional area. to minimize loss of heat through them, heat also being lost through heat-loss paths other than through said solid paths as through a radiation path, said heat-conduction paths having temperature efilcients of difiering sign and having dimensions such that when heat developed by radiation directed upon said element is dissipated through said-heat-loss paths at' the same rate as received approximate equality is established'be'-' tween a first ratio of the sum of the respective heat-loss factors of all heat-loss paths to the sum of the respective products of each of said heat-loss factors and the respective tempera-'.

ture coefficients of thermal conductivity of each of sai-d paths and a second ratio of d temper-a 28 tare-voltage constant A to a temperature-volt age constant C, said constants A and 0 having values which satisfy the equation AV: (A+CT) AT where AV=the voltage change produced by change in said characteristics for a small change in tem perature, AT, and

T=the ambient temperature above a reference level.

tlon path, the ooejllcienl of thermal conductivity of at least one of said paths being different in sign from other of said paths, and the dimensions establishing the'heai loss through each of the paths establishing approximate equality between u first patio of the sum of the respective heat-loss factors of all heat-loss paths to the sum of the respective products of each of said heat-loss factors and the respective temperature coefilcients of thermal conductivity of each of said and a second ratio of a temperaturevol'lage constant A to a temperature-voltage constant C, said constants A and C having values satisfy the equation AV:(A+CT) AT where AV the voltage change for a small change in temperature, AT, and Tzthc ambient temperature above a reference level.

35. A heat-sensitive device for measuring apparalus, said device comprising a heat-receiving region varying in temperature in response to variations in ambient temperature and with heat applied thereto, electrical means including an element at said heat-receiving region having an electrical characteristic which varies with temperature to produce a potential difiereuce AV, a plurality of heat-loss paths by means of which heat is lost from said heat-receiving region, said loss paths including at least a radiation path, and solid conduction paths made small to minimize loss of heat for maximum rise in temperature of said element with application of heat thereto, at least one of said heat-loss paths having a loss jactor and a coefficient of thermal conduct-lolly which in relation to the loss factors and temperature coefficients of thermal conductivity of the remaining heat-loss paths produce approximate equality between a first ratio of the sum' of the respective heat-loss factors of all heat loss paths to the sum of the respective products ofeach of said heat-loss factors and the respective temperature coefficients of thermal conductivity AV: A-]CT) AT where 29 AVzthe characteristic change for a small change UNITED STATES PATENTS in temperature AT, and Number Name Date Tz the ambient temperature above a reference 1533740 Kemath Apr 14 1925 level- 553 789 Mo ller se u pt. 15, 1925 WILLIAM FASTTE" 5 2,186,948 Adler Jan. 16, 1940 2,357,193 Harrison Aug. 29, 1944 REFERENCES CITED OTHER REFERENCES The following references are of record in the Temperature, Amer. Inst. of Physics, 1941,

file of this patent or the original patent: 10 pages 12154217.

Certificate of Correction Reissue No. 23,615 January 13, 1953 \VILLIAM G. FASTIE It is hereby certified that error appears in the printed specification of the above numbered patent requiring correction as follows:

Column 14, iines 24, 25, and 26, for

i read g column 19, line 48, for hunt read shunt. column 23, line 40, for of heat read of the heat and that the said Letters Patent should be read as corrected above, so that the same may conform to the record of the case in the Patent Ofiice.

Signed and sealed this 7th day of J uly, A. D. 1953.

THOMAS F. MURPHY,

Assistant Commissioner of Patents.

Images