Isotech 962 User manual

MODEL 962 HANDBOOK
ISSUE 02 02/05
1
MODEL 962
HOTSPOT
The company is always willing to give technical advice and assistance where appropriate. Equally,
because of the programme of continual development and improvement we reserve the right to
amend or alter characteristics and design without prior notice. This publication is for information
only.
Isotech North America
158 Brentwood Drive, Unit 4
Colchester, VT 05446
Phone: [phone]
Fax: [phone]
www.isotechna.com
[email protected]

MODEL 962 HANDBOOK
ISSUE 02 02/05
2
CONTENTS
Page No:
1. Introduction 4 - 5
2. A New Temperature Platinum Resistance Thermometer
for 1990 6 - 10
3. Returning thermometer 962 to Isotech 11
4. Performance Report 12 - 15
DIAGRAM:
Packing Instruction for Returning thermometer 16

MODEL 962 HANDBOOK
ISSUE 02 02/05
3
GUARANTEE
This instrument has been manufactured to exacting standards and is guaranteed for twelve months
against electrical breakdown or mechanical failure caused through defective material or
workmanship, provided the failure is not the result of misuses.
In the event of failure covered by this guaranteed, the instrument must be returned, carriage paid, to
the supplier for examination, and will be replaced or repaired at our option.
FRAGILE CERAMIC AND/OR GLASS PARTS ARE NOT COVERED BY THIS GUARANTEE
INTERFERENCE WITH, OR FAILURE TO PROPERLY MAINTAIN THIS INSTRUMENT MAY
INVALIDATE THIS GUARANTEE
RECOMMENDATION
The life of your ISOTECH Instrument will be prolonged if regular maintenance and cleaning to
remove general dust and debris is carried out.
We recommend this instrument to be re-calibrated annually.
Serial No:........................
Date:.............................
[email protected]
www.isotechna.com
Fax: [phone]
Phone: [phone]
Colchester, VT 05446
158 Brentwood Drive, Unit 4
Isotech North America

MODEL 962 HANDBOOK
ISSUE 02 02/05
4
INTRODUCTION
You or your company has just paid a considerable sum of money to purchase this 962.
Although the materials to make the 962 are very costly, it has also taken up to 1000 hours to prepare
the 962 for your use.
It is not a product that can be assembled, tested and sold, it is produced, calibrated, aged and
recalibrated, until its characteristics are stable enough to meet the exacting needs of ITS-90.
The weeks of work that go into its production make each 962 more than another product. Each 962
already has a character and a history before it leaves us. Please look after it. Regard yourself as its
custodian rather than its owner.
1. Always keep the 962 in its case when not in use.
2. When in use, support the handle.
3. Cool and store the 962 in the same place as you normally use it.
4. Each time before you use the 962 clean off all traces of grease by using a suitable alcohol and
dry thoroughly.
5. Even go so far as using gloves to handle the 962, and keep the gloves clean.
6. Quartz is glass. It is a supercooled liquid. At 800C and above, your 962 will bend and bow if
you do not support it along its complete length. At high temperatures I recommend that the
962 be housed inside a close fitting recrystallised Alumina closed ended tube, which has been
pre-fired to 1000C.
7. Quartz is transparent in two senses of the word. At temperatures above 700C metallic
vapours can pass through the quartz and attack the pure platinum sensing element. Isotech
have developed a product which can be attached to the 962 to prevent this happening. A unit
is provided free with each 962.
8. Remember 962's cannot be repaired!
9. Do not drop or knock.
10. Do not put lateral pressure on the stem, it is rigid and will break
11. The handle and top 40mm of the stem should not be heated over 50°C. The cable will
withstand 200°C.
To keep the original characteristics:
1. Do not use the standard in conditions where there is vibration or mechanical shock.
2. Avoid thermal shock - allow the standard to warm-up and cool-down slowly.
3. Do not exceed the temperature limits.
4. If used above 450°C, always cool slowly to 450°C, and maintain at 450°C for approximately
½ hour, and then withdraw to ambient.
5. Always cool a hot SPRT in a thermometer rack where the SPRT can cool in an upright
position, never cool in a horizontal position.

MODEL 962 HANDBOOK
ISSUE 02 02/05
5
The cable is connected as follows:

MODEL 962 HANDBOOK
ISSUE 02 02/05
6
A NEW HIGH TEMPERATURE PLATINUM
RESISTANCE THERMOMETERS FOR 1990
John P. Tavener
Isothermal Technology Limited
Pine Grove, Southport
Merseyside PR9 9AG England
It is not often that one can pin down to the hour of the day when a product is born, however, I was
sitting at the back of the ITS-90 workshop session at N.C.S.L. in 1987. Concern was expressed that
there were no western suppliers of the new S.P.R.T. required for 1990 to work to the Silver Point.
During the discussion I conceived the notion that I should produce such a product.
Knowing that the problems to be surmounted were larger than my experience could solve, technically
and financially, I sought expert advice upon my return to England.
My first call was to N.P.L. where the staff were most helpful and provided me with much background
material - in particular the articles written by John Evans proved most useful and informative.
My second call was to a manufacturer of quartz parts where the design which was emerging was
altered to suit practical manufacturing techniques, again the staff proved most helpful and cooperative.
My last call was to the supplier of platinum who was charged with the task of producing the highest
alpha wire.
Defining the Problem
In defining the assembly of the High Temperature Standard Platinum Resistance Thermometer, the
history of the device was considered.
Resistance Value
Thermometers have been produced with Ro of between 0.25and 2.5. Clearly the lower the
resistance the smaller the effects of parallel resistance and because thicker wire can be used, the
more likelihood of a mechanically stable thermometer so I initially chose 0.25. Higher resistances
could be introduced at a later date assuming the design proved successful.
Diameter
The thermometer has to be calibrated inside Freeze and Triple Point Cells which smallest sizes
suggested a diameter of less than 8mm - 7.5mm was chosen.
Length
Again, the main consideration for the length is to accommodate existing Freeze Point Apparatus. 600
to 700mm seems typical. 650mm was chosen.

MODEL 962 HANDBOOK
ISSUE 02 02/05
7
SHEATH MATERIAL
The only material ever considered was quartz and I did not feel qualified to challenge this assumption.
INTERNAL CONSTRUCTION
The Sensing Element
Many constructions have been used. The main options seemed to be crucifix or spade. Since the
spade option has half the points of contact as that of the crucifix, it was chosen. Laser cutting of the
quartz ensured a smooth, polished surface for the coil.
The Leads
Again, two options presented themselves.
Little bits of quartz covering the platinum leads fed, at 50mm intervals through 4 holed quartz discs or
single lengths of quartz capillary tubing running from the sensing winding to the handle.
This decision is one of the key points in the construction, since all failures in experimental
thermometers have occurred with leads open or short circuiting.
I will reserve my decision, conclusion and reasoning for my lecture.
The Handle
In the handle, the platinum wire interfaces with the cable which connects the thermometer to the
reading instrument.
It is also the area where the assembly is vacuumed. It needs to be small so as not to add a large
weight to a long, thin fragile stem. This is the second key area, and the one in which I take most
pleasure in the solutions.
Atmospherics
It is necessary to ensure that the total assembly and all constituents start clean and remain clean and
that over the useful temperature range they do not produce any substances or have properties which
might affect the thermometer's performance. Such as parallel resistances. Or absorb substances
from outside the assembly, which might damage the performance of the sensor. This can be done by
copious applications of acids, alkalis, rinsing, oxidising and many other precautionary techniques,
these have been described adequately elsewhere.

MODEL 962 HANDBOOK
ISSUE 02 02/05
8
MECHANICAL CONSIDERATIONS
Temperature Coefficients
Quartz has the unique property that it does not expand (or contract), in consequence it does not crack
- like other glass or ceramics. It is pure and will not contaminate its contents, unlike metals such as
nickel or steels. It can be used with good properties at temperatures around 1,000C.
Platinum, the other constituent in the product has a temperature expansion of 0.9% per 1,000C
expanding up to 5mm. Since the quartz does not expand, the platinum is forced to move with respect
to the quartz. Provided the mechanism is understood and the magnitude defined, the shape, size and
constraints of the assembly can be designed to accommodate these movements without work
hardening stress or strain.
In the lecture I will show the exact solution.
The Unexpected
During his researches, John Evans discovered contaminants which were eventually traced to sources
outside the thermometer. These, he concluded were caused by vapours passing through the quartz
sheath at high temperatures.
This problem has a number of solutions, the easiest is being tested and will be patented by the time
this lecture is presented.
The Guard Ring
This, almost mythical aspect of high temperature thermometers has its proponents and opponents.
The most cogent voice I have consulted suggests the dubious advantage of this approach, because of
the temperature gradient along the sheath.
My own solution has been a commercial one of offering either a guarded or non-guarded version.
Does It Work
The products at this level of Metrology are only 40% design, 60% consists of evaluation, history and
confidence. Even before I began this project I established our National Physical Laboratory's
willingness to evaluate such a product.
Firstly however, I needed to establish some initial figures, so as not to waste N.P.L.'s valuable time.
By bending rules at all the suppliers we produced the first two units in 4 weeks. Whilst these were
being evaluated, a further 4 were built to a slightly modified design based on improvements from the
first two.
One unit from the first two and one from the second batch were sent to N.P.L. for evaluation, after we
had aged and tested the thermometers for up to 1,000 hours.

MODEL 962 HANDBOOK
ISSUE 02 02/05
9
Initial impressions from N.P.L. suggest that after cycling to Tin, Zinc, Aluminium and Silver Points.
1.Ro changed by less than 2mK
2.Alpha by less than mK
3.R, Ag by less than 2mK
It is anticipated that full details will be available before N.C.S.L.
AC/DC
Arguments range as to the best means of measurement of such a supreme device. Although it is
beyond the scope of a sensor manufacturer to discuss this aspect, since I am a practical person I will
allow reference to Berry who suggested that since polarisation effects have time constants of 3
minutes or more, only very slow A.C. is suitable. I hope to be able to offer comparative results by the
time of lecture.
The Final Solution
The final solution is nothing new. Just a new combination of good ideas.
Its derivation is the total history of efforts provided by all the worthy researchers into high temperature
sensors, Evans, Chattle ....... added to it has been the expertise of one of Europe’s foremost quartz
manufacturers and the world's top producer of pure platinum.
The final alpha values, as measured at N.P.L. confirmed my own findings that the alpha values were
very high; typically 0.003928. The importance of this figure is not the high value, but that the high
value shows that no contamination has been introduced during manufacture, or in use subsequent to
the manufacture.

MODEL 962 HANDBOOK
ISSUE 02 02/05
10
Limits of Expectation
ITS-90 committee does not expect that the standard P.R.T. will last as long, or with such stability as
that standard adopted in IPTS 1968.
The main aspect of the new thermometer is that it should be an improvement on the type R or S
thermocouple.
In conclusion, with such a short, if respectable history of the Isotech 962 thermometers, I submit that
there is now (at least) a western source of thermometers for ITS-1990.

MODEL 962 HANDBOOK
ISSUE 02 02/05
11
RETURNING YOUR THERMOMETER TO ISOTECH
REPACKING INSTRUCTIONS:
Place a small foam spacer under the end of the Quartz sheath and one under the handle.
Then place two larger pieces over the end of the sheath.
PREPARING THE BOXES
Place the two foam blocks over the closed case and lower into inner box. Pad out ends with
polystyrene chips to stop the unit from moving - seal the box, lower it into the outer box and fill with
chips, then lift the inner box slightly so the polystyrene chips completely surround it. Place the
protective sheath if supplied on top of the chips and seal the box.
Always remember to label the box thoroughly with “fragile” and “this way up” labels and arrange
adequate insurance cover.
Your unit should now be ready to send safely.

MODEL 962 HANDBOOK
ISSUE 02 02/05
12
PERFORMANCE REPORT
STANDARD THERMOMETER
Various producers of standard platinum resistance thermometers quote the performance of their
product in various ways:
Tinsley say Reproducibility ±0.001°C
of their 5187SA Accuracy ±0.01°C over 0 to 100°C
25.5thermometer
Y.S.I. DO NOT SAY ANYTHING
Chino quote various
drift rates from 0.001K/year for the 25.5up to 630 to 0.01°C/100
hours when a thermometer is used above 850°C. Their
reproducibility is 0.001K.
CLAL drift 10 cycles - 500 hours at 660°C = 0.01
The following definitions from the International Vocabulary of Metrological terms show that
"Accuracy", "Reproducibility" and "Drift" do not define the performance of an SPRT.
The real criteria is how accurately can an SPRT measure temperature and with what uncertainty.
Using a "Measurements International" bridge it is possible to resolve the signal from an S.P.R.T. to
±0.00005K.
In a fixed point cell it is possible to realise the ITS-90 values from about .15mK for the Triple Point
of Water and Melting Point of Gallium to between 2 & 4mK at the Silver Point.
With a fixed resistor and the best calibration 0.05 ppm uncertainty is possible.
In addition is the reproducibility of the SPRT and various other small uncertainties; from all this the
overall uncertainty can be estimated.
Uncertainties also vary depending on the temperature range of calibration
Isotech's 909 and 962 thermometers are sufficiently stable to conform to uncertainties translated
overleaf Models 670 and 96178 can be calibrated as tabulated overleaf or at extra cost to the
uncertainties of our latest schedule.
The annual drift, although of interest is of little importance to the calibration process. Certainly 1mK
per year drift is an impossibility if the thermometer is used regularly from 90K to 661°C and Chino
have refused to replace a thermometer with higher drift rates than their literature states.

MODEL 962 HANDBOOK
ISSUE 02 02/05
13

MODEL 962 HANDBOOK
ISSUE 02 02/05
14
The enclosed uncertainties apply on condition that the measurements of resistance at the triple
point of water are reproducible within the uncertainty given at that temperature. Uncertainties for
thermometers which do not meet this requirement will be increased by a factor of 3 or 10, as
appropriate. For the time being, metal-sheathed thermometers will be calibrated by comparison
with standard thermometers at the requested points. Calibrations at the triple point of argon are
obtained from comparisons with standard thermometers in a bath of liquid nitrogen.
Calibration below 0°C is optional. Ranges of calibration using combinations of fixed points other
than those given above, may be accommodated on request.
3.05 Accuracy of Measurement
The closeness of the agreement between the result of a measurement and the (conventional) true
value of the measurand.
3.07 Reproducibility of Measurements
The closeness of the agreement between the results of measurements of the same measurand,
where the individual measurements are carried out changing conditions such as:
-method of measurement
-observer
-measuring instrument
-location
-conditions of use
-time
Notes
1. A valid statement of reproducibility requires specification of the conditions changed.
2. Reproducibility may be expressed quantitatively in terms of the dispersion of the results.
3.09 Uncertainty of measurement
An estimate characterising the range of values within which the true value of a measurand lies.
Note Uncertainty of measurement comprises, in general, many components. Some of these
components may be estimated on the basis of the statistical distribution of the results of series
of measurements and can be characterised by experimental standard deviations. Estimates of
other components can only be based on experience or other information.

MODEL 962 HANDBOOK
ISSUE 02 02/05
15
5.16 Stability
The ability of a measuring instrument to maintain constant its metrological characteristics.
5.18 Drift
The slow variation with time of metrological characteristic of a measuring instrument.

MODEL 962 HANDBOOK
ISSUE 02 02/05
16
RE- PACKING INSTRUCTIONS FOR THE RETURN OF THERMOMETER 909/962