Thermometer error correction

Barometer Correction—After applying the corrections for thermometer error, correct each reading for deviation of the barometric pressure from normal by adding algebraically the correction calculated as follows ... 

Constant errors arise when you use equipment that is incorrectly calibrated or that functions incorrectly the same way each time. The constant error is not a result of your care or lack of it, but rather a faulty instrument. For example, a thermometer might be calibrated to read 1 C too low at all temperatures. Unless you calibrate the thermometer and correct for this miscalibration, even temperatures measured with great care will be 1°C too low. Constant errors also arise when you use procedures that create bias in your measurements. For example, if the volume of a liquid is measured in one container and the mass of the liquid is measured in another container after a transfer, the mass, and thus the density calculated from it, will usually be too small because not all of the liquid will be transferred when it is poured from one container to another. Such a constant error can be eliminated by changing the procedure to eliminate the transfer so that volume and mass can be measured in the same container. 

When temperatures above 150°C are measured, thermometer errors can become significant. For an accurate melting point with a high-melting solid, you may wish to apply a stem correction to the thermometer as described in Technique 13, Section 13.4. An even better solution is to calibrate the thermometer as described in Section 9.9. 

The comparatively inexpensive long-scale thermometer, widely used by students, is usually calibrated for complete immersion of the mercury column in the vapour or liquid. As generally employed for boiling point or melting point determinations, the entire column is neither surrounded by the vapour nor completely immersed in the liquid. The part of the mercury column exposed to the cooler air of the laboratory is obviously not expanded as much as the bulk of the mercury and hence the reading will be lower than the true temperature. The error thus introduced is not appreciable up to about 100°, but it may amount to 3-5° at 200° and 6-10° at 250°. The error due to the column of mercury exposed above the heating bath can be corrected by adding a stem correction, calculated by the formula ... 

Tn addition to the error due to the exposed stem, ordinary chemical thermometers of low cost are subject to errors due to irregularities in the bore and sometimes the scale graduations may not be very accurate. " It is therefore essential to check the thermometer at several temperatures against the melting points of pure solids or the boiling points of pure liquids as described below. The application of an exposed stem correction will of course be unnecessary if the thermometer is calibrated in this way. A calibration curve may then be drawn upon graph " paper from the data thus obtained. Temperatures at intervals of about 20° are marked as abscissae and the corrections to be added or subtracted as ordinates the points thus obtained are then connected by a smooth curve. The thermometer correction at any temperature may be read directly from the curve. 

Most of the instruments, commonly used in an analytical laboratory, such as UV-Spectrophoto-meter, IR-Spectrophotometer, single—pan electric balance, pH-meter, turbidimeter and nephelometer, polarimeter, refractometer and the like must be calibrated duly, before use so as to eliminate any possible errors. In the same manner all apparatus, namely pipettes, burettes, volumetric flasks, thermometers, weights etc., must be calibrated duly, and the necessary corrections incorporated to the original measurements. 

The ordinary mercury-in-glass thermometer as covered by ISO 177014 is in such common use that it is rather badly taken for granted. In practice, much of the variability associated with testing at a set temperature can be traced to the misuse of thermometers. They should be calibrated frequently, carefully inspected for separation of the mercury, and immersed to the correct depth. The worst errors are usually found with low temperature thermometers and, hence, particular care should be taken when conditioning or testing at sub-zero temperatures. Precision thermometers are covered by ISO 653 - 65615 18 and there is a British standard for laboratory thermometers19. 

Temperature variations in the instrument are a source of error and electrical power dissipation is limited to avoid the effects of self-heating. This is achieved by means of the four lead system shown in Fig 6.25ft. This minimises any effects of variations in temperature on the resistance RCL of the connections between the RTD and the bridge and is used normally with digital thermometers and data acquisition systems where the sensor non-linearity is corrected within the computer software. 

For most purposes a partial-immersion thermometer need not be stem corrected because of a few degrees variation in room temperature or a few degrees error in the immersion level. On the other hand, it is usually worthwhile to apply stem corrections to readings of a total-immersion thermometer when it is used in partial immersion, particularly when reading temperatmes well removed from room temperatures. 

The most important method of achieving temperature control is to use a sensitive thermometer that generates an electrical signal, such as a resistance thermometer, thermistor, or thermocouple. Comparison of this signal with a reference signal that establishes the set point provides an error signal to a feedback circuit that controls the power to the heater. Thns at r temperature deviation from the desired value detected by the sensor will be corrected antomatically. 

In making a determination it is necessary to consider a number of sources of error, and to correct for them. Thus we must make corrections for the change in temperature of the walls of the calorimeter vessel, of the thermometer itself, of the stirrer with which the hquid is agitated to secure efficient mixing, etc. 

For most purposes, it is convenient to use partial-immersion thermometers, because the corrections required at any temperature will usually be smaller. However, regardless of which type is used, it should be calibrated against some thermometers for which the calibrations are known. The common practice of calibrating thermometers by using the melting points of known substances will often lead to considerable errors if the capillary method is used. 

Thermometer correction. The temperature which is read on the thermometric scale must be corrected because there are several errors in such determinations. One source of error arises from the construction and calibration of the thermometer. The bore of the capillary may not have the same diameter throughout further, the scale graduation and the calibration of low-priced thermometers are not very accurate. A second source of error is the method used in the common melting point apparatus. The common thermometer has been calibrated while totally immersed in a bath. In the melting-point apparatus described, only a part of the stem is immersed. The column of mercury above the oil bath has a lower temperature than that at which the thermometer was calibrated. Therefore either a thermometer calibrated by partial immersion should be used or a correction must be made for the unequal heating of the mercury in the stem of the thermometer. Although thermometers calibrated by partial immersion are available, the latter practice is the more common. 

In order to make corrections for errors due to construction, it is necessary to calibrate the thermometer. This is best accomplished the placing of the thermometer side by side with another thermometer already calibrated by the Bureau of Standards and then taking three to four readings over the entire scale. Another method of calibration is to compare the readings of the thermometer with the freezing or boiling points of reference compounds. Several compounds which are used for the calibration of thermometers are listed in the Appendix. 

Note that you cannot check the accuracy of the pipette unless you know that the balance and the thermometer are accurate too. However, you can estimate the precision with which you can measure 25 cm3 with the pipette. If the balance is correctly calibrated and set-up properly, the random error due to weighing will be very small compared with that due to pipetting. 

The accuracy attainable with a liquid-in-glass thermometer is limited by the characteristics of the thermometer itself. Instability of the thermometric liquid, nonuniformity of capillary bore, and inaccuracies in scale graduation are the important factors. Uncertainties in corrections for the emergent stem may greatly limit the accuracy of partial-immersion thermometers. Generally, partial-immersion thermometers are assigned an uncertainty of 0.3°C in their calibration, whereas total immersion thermometers may have an uncertainty as small as 0.03°C. Observer errors add to the uncertainty but with care these can usually be made relatively small. 

Your thermometer may be a major source of error in melting point measurement. Occasionally thermometers for routine laboratory use may not be accurate and may read up to 10°C high or low. To avoid this problem you should always calibrate your thermometer to determine any error and be able to correct for it. 

Correct the recorded temperatures by the amount of the error of your thermometer and obtain the average of these four values. (4) What is this temperature called ... 

Record the correct values of the melting point of tin and that of the eutectic, and those you obtained. Add to your observed value 6° in the case of the melting point of tin and 3° in the case of the eutectic. These values are the so-called stem corrections, that is, the amounts that must be added to the observed temperatures to correct for the error due to the fact that the mercury in the stem of the thermometer was not heated to the temperature of the mercury in the bulb. 

It should be noted that all the thermometers listed above except the ground-joint type (Fig. 363) are calibrated for total immersion in other words, they read correctly only when the whole of the mercury thread is at the measured temperature. Short thermometer wells can thus give rise to appreciable errors. According to Piatti and Marti [20] the error is small only if the mercury bulb is just immersed in the oil in the well. In practice, total immersion thermometers should be mounted so as to be completely immersed in the vapour or liquid, although the ease of reading is thereby adver.sely affected. Otherwise a correction for emergent stem has to be applied (r/. 

---