211.020 General. Direct inquiries to, and mark shipments Electrochemistry Section, MET 211.021 Standard cells. A247, National Bureau of Standard 301-921-2721. (a) Unsaturated standard cells will be accepted for calibrati Bureau only from public utilities and others having operations such a nature as to require calibrations by the Bureau. by the (b) Normally, unsaturated cells require about two weeks for t complete calibration. The cells are kept in a thermally insulated cabin and, after the values have become reasonably constant, readings of the emfs are taken daily for a period of ten days. If the emf continues fluctuate, or is unusually low, or if the cell shows other abnorm indications, the nature of the failure is stated. Unsaturated cells a not likely to be injured by normal transportation (mail or express) if th are packed carefully. Because of the possible hazard from freezin shipment during very cold weather should be avoided. (c) Saturated cells for test in thermostatically controlled o baths should be transported by messenger because the cells should never inverted or tipped from an upright position by more than 45° in a direction. Saturated cells of the unshippable type housed thermoregulated air baths also should be transported by messenger and und power if possible. Saturated cells of the shippable type housed thermoregulated air baths should be packed carefully and shipped und power if possible. If the cells cannot be shipped under power, shipme during very cold weather should be avoided because of the possible haza from freezing. To ensure the highest accuracy in reported values, the em of saturated cells are measured while they are kept at least 6 to 8 wee in an NBS temperature-controlled oil bath or in their own air bath. thermoregulat 211.021a 211.021b Cadmium standard cell (unsaturated type) determination electromotive force with uncertainty of 0.005 percent. fir °C, Cadmium standard cell (saturated type), measurement of the 211.021c Each additional saturated cell of a group. 211.021d 211.021e Cadmium standard cell (saturated type), measurement of the fir cell of a group at any temperature between 20 and 35 °C, exce 28 °C, in a thermostatically controlled oil bath. Each additional cell of a group (at any temperature between and 35 °C, except 28 °C, in a thermostatically controlled o bath). .021z Special measurements not covered by this schedule. Electrical Instruments 211.030 General. Direct inquiries to, and mark shipments for the attention of: ctrical Instruments Section, MET - B162, National Bureau of Standards, hington, DC 20234. Telephone: 301-921-2727. 211.031 Standard resistors for current measurements (shunts). (a) Calibration: Normally the Bureau calibrates only resistors 0.04 percent accuracy or better. Test results for all high quality oilled resistors whose power rating does not exceed 50 watts normally will reported with an uncertainty of 50 ppm. In general, well designed airled resistors whose power rating does not exceed 25 watts will be -orted to the same uncertainty. A11 other air-cooled resistors for rent measurements will be reported with an uncertainty of 0.01 percent. (b) Design: A standard resistor for current measurements is a r-terminal resistor, for which the resistance is defined as the ratio of open-circuit potential difference between the potential terminals to current through the current terminals. The resistance value will be inite and reproducible only if the current flow pattern at the potential minals is completely reproduced. This flow pattern should be fixed by istor design to be independent of the way in which current is introduced the current terminals and of the location of leads on the potential minals. In some instances where this has not been done the type ation of connections to the current terminals can be specified quately to fix the flow pattern at the potential terminals. (c) High-current resistors: and (1) Resistors for high currents (above about 1,000 A) require siderable power SO that their temperature rise between low and rated rent, and the resulting change in resistance, will depend not only on ir design, including means provided for dissipating heat, but also on connecting bus bars and their junctions to the resistor. Bus bars of erous cross section may carry away a significant part of the heat erated in the resistor; inadequate bus bars may actually contribute to heating of the resistor. In addition, contact resistance at the points connection to the bus bars, unless carefully minimized, may contribute reciably to the heating. (Contact resistance resistance of bolted connections ends on area of contact, cleanness of surfaces, and pressure.) istance determinations made in the laboratory at rated current, refore, may be of little value because the working temperature ditions cannot be duplicated. The best experimental procedure to use in h cases is to place the standard in a temperature-controlled enclosure I measure its resistance with a comparatively low test current when it is ted uniformly to temperature approximating that at which it will operate service. From data at two or more elevated temperatures, combined with at the operating temperature can be read, provided this temperature determined by the user with the resistor under the actual operati conditions. (2) Changes in resistance may also result from strains in t resistance element produced by mechanical forces incidental to clamping t resistor connections, as well as from inherent internal expansi constraints on resistor parts, or forces from the magnetic field field produc by the current. (d) Test schedule: When first submitted for test, resisto should be tested with about 20 percent of rated current and with full rat current; normally when resubmitted for test, determinations need need be ma with only 20 percent of rated current; once stability is proved, t resistor need not be recalibrated at intervals of less than two years. 211.031a-1 Initial determination of resistance of a single-range resist 211.031a-2 211.031b 211.031c 211.031d 211.031z voltage Same as a-1, except current rating above 300 A but not exceed 1,000 A. Determination of resistance on another range of a multiran resistor, at 30 percent rated current or less (current rati not to exceed 300 A). Test according to item 211.031a or b having been made, for additional determination at another test current (not exceed 1,000 A). Twenty determinations of resistances corresponding to 9 pl Determinations of resistance at temperatures above ro 211.032 Fixed d-c voltage dividers (volt boxes). range of t A resistive voltage divider is used to extend the measured by a potentiometer. Its ratio for any range is obtain by dividing the voltage across its input terminals by the open-circu voltage across the section to be connected to the potentiometer. (a) Calibration: (1) Normally, NBS calibrates only dividers for which t maker's stated ratio accuracy is 400 ppm or better. Dividers normally a calibrated by direct comparison with the NBS ratio standard. Values ratio normally are reported with an uncertainty of 50 ppm. However, t values of ratio of some special types having combined humidit temperature, internal heating, and high voltage effects less than 10 described (2) Silsbee-type dividers of design similar to that in NBS RP1419 are calibrated by the Dunfee method (NBS Paper 67C1-114). Values of ratio normally are reported with an uncertainty of 10 ppm. (b) Humidity and ambient temperature effects: (1) The insulating structure of a divider is equivalent to a network of high resistances in parallel with one or more of its wire-wound precision resistance elements. Thus, changes in insulation resistance as a result of variations in surface or volume moisture may affect the ratios. Such ratio changes normally are less on low than on high ratios. This effect can be reduced or eliminated by constructions which provide built-in guard electrodes, maintained at appropriate potentials. Another effect of humidity is produce changes in the values of the precision wire-wound resistors. The magnitude of this effect varies with coil construction with wire size and coating. Because humidity effects may reach equilibrium only after days (or even weeks), it is recommended that laboratory humidity be held continuously at or below 50 percent. and (2) Changes in ambient temperature should have little effect on ratio if all the resistance elements have small and equal temperature coefficients. (c) Internal heating and high voltage effects: (1) Internal heating, which includes both the self-heating of the high-resistance coils and the transfer of heat to other coils (proximity heating), may significantly change the ratio of a divider. (2) The magnitude of this change and the time required for a state of "ratio equilibrium" to be established varies with divider construction and with operating and ambient conditions. For example, many dividers constructed of manganin resistors exhibit heating curves (ratio versus time) that go through a maximum within 10 minutes after the divider is energized then decrease to an equilibrium value. The time required to reach equilibrium usually is two hours or less. For the highest ranges of a divider having a relatively large operating current the difference between the maximum value and the equilibrium value can be as large as 200 ppm. The Silsbee-type and special dividers that are constructed of resistance elements having small and equal temperature coefficients and designed to have low power dissipation per element usually have ratio changes less than 10 ppm due to internal heating. (3) Corona and other high voltage effects negligible for voltages up to 1500 volts. (d) Test schedule: usually are (1) Dividers are calibrated on each range at rated voltage. However, the first calibration test should also include a complete test at reduced voltage (20 percent rated voltage or less) to provide significant information on heating effects. The general procedure at NBS for calibrating a divider by the direct comparison method is to measure the ratio at rated voltage on the highest range as a function of time and then proceed to the lower ranges in descending order. Approximately five to fifteen minutes are required for ratio equilibrium to be established on these lower ranges. Dunfee method at reduced voltage to minimize any self-heating and proximit heating effects. A comparison test against the NBS ratio standard is als performed to determine the ratio change that may arise from interna heating when the divider under test is energized at rated voltage on the highest voltage range. (3) Once stability has been established, a divider should no require recalibration at intervals less than 2 years. 211.032b 211.032c 211.032d 211.032e Determination of ratio and internal heating effects of a on divide one range at rated voltage, not to exceed 1,500 V, and at ratio not to exceed 10,000/1. Determination of ratio of a divider at rated voltage on a additional range. Determination of ratio of a divider at a reduced voltage on range tested in 211.032a or b. Determination of each ratio of a multirange Silsbee-type divider Determination of the internal heating effect of a Silsbee-typ divider on one range at rated voltage, not to exceed 1,500 V. 211.032z Measurements on dividers not covered by this schedule. 211.033 RMS ac-dc transfer standards (voltage and current con verters and comparators, 20 Hz to 50 kHz, 1 mA to 20 A, and 0.5 to 600 V). (a) Ordinarily only ac-dc transfer standards and thermal con verters of 0.05 percent rated accuracy or better are accepted for test which consists of ac-dc difference determinations by the procedures 0 items 211.033a to c. See 211.038 for tests of peak ac-dc comparators. (b) Ac-dc difference tests consist of determinations of the dif ferences between the quantities (current, voltage, or power) required t give the same response (output) of the transfer standard on alternatin current and on reversed direct current (the average of the two direction: of direct current). The alternating quantity, Qa required for a give response of the instrument or converter it then 3' = Qd (1 s) where Qd i a : the average quantity required for this response on reversed direct current as determined by d-c standards, and S is the small fractional ac-d difference. therma (c) The ac-dc differences of commercial audio-frequency transfer standards usually are negligible ( < 0.005%) at intermediat frequencies. They become significant at the higher higher audio frequencie (often increasing approximately as the square of the frequency) because of the residual reactances of the circuit elements. However, generally al but the high-current and high-voltage ranges have frequency influences les than 0.01 percent up to at least 20 kHz. kHz (the (d) For these reasons, tests are recommended on each range at 2 upper limit for best accuracy at present), or at the upper rate |