^{1}

^{1}

^{*}

^{1}

A set of seven single junction thermal converter Micropotentiometers (μPots) has been constructed at the National Institute for Standards (NIS), Egypt. This set has been built to cover the low ac voltage ranges from 2 mV to 200 mV at frequencies from 40 Hz up to 20 kHz. The construction of the μPots set has been presented and an adopted calibration method has been performed as well. This method has been performed by means of a step-down procedure using a Digital Multi-Meter (DMM). The scaling procedures have been carried out in sequential steps starting from the calibration of the 200 mV-μpot by using DMM that is accurately calibrated at its 200 mV ac voltage range down to 2 mV-μPot. Furthermore, a new automatic calibration system has been established to achieve the scaling procedures. This system has been specially designed using Laboratory Virtual Instrument Engineering Workbench (LabVIEW) software to overcome the deficiencies of manual methods. The automatic calibration has been investigated of all mPots at different frequencies. The ac-dc differences for the μPots and their uncertainty evaluation from 2 mV to 200 mV at different frequencies from 40 Hz to 20 kHz have been determined.

The electrical SI units are defined and realized as dc electrical quantities. ac electrical quantities can’t be directly determined in terms of SI units, so ac quantities have been determined in terms of the dc quantities. In order to determine ac quantities, it is necessary to transfer ac to dc [

The constructed set consists of seven SJTC µPots having output ac voltage with nominal values of 2, 5, 10, 20, 50, 100 and 200 mV. Each µPot consists of an ultra high frequency (UHF) single junction thermalelement in series with a low inductance radial output resistor. In the SJTC energy dissipated by an ac current flowing through a heater resistor, raising its temperature above the ambient is compared to the energy dissipated by a dc current flowing through the same heater. The increase in the temperature of the heater at ac and dc, proportional to the dissipated energy, is measured using a thermocouple. A relative difference between the response of the converter to ac and dc inputs, called ac-dc transfer difference, is determined from these two measurements. The SJTC which is used in each of our adopted μPots construction has been tested at NIS. It shows innovative design and produces extremely low transfer error [

The second main component in constructing each μPot is the radial resistor. It consists of ten equal resistors ensured an optimum frequency response, as it is a special design annular shape film resistor. These ten resistors are connected in parallel and placed between the center pin and the outer flange of an N-type female connector. It is securely soldered into the output N-type coaxial connector and screwed into the μPot case [

Step-down scaling of the ac voltages from 200 mV to 2 mV is shown in

used from full to half of their nominal voltages with almost the same ac-dc differences [

The scaling procedures have been automatically performed through implementation of an automated calibration system. This automated system accelerates the measurement procedures, eliminates the operator’s errors, and significantly improves the measurement process. Moreover, it allows statistical proceeding of the results in rather short time. General Purpose Interface Bus (GPIB) card and cables are used to communicate and control the automated system hardware with its software. The new software LabVIEW program is designed to perform the following: control the calibrator output, perform measurement procedures with the specified times for the warming-up and the steady-state of the TE, collect measurements data indicated by DMM, and generate data sheets at the end for saving the measurement results. The scaling procedures from 200 mV to 2 mV are carried out in several sequential steps.

The first step of the scaling procedures is the calibration of Fluke 8508 A DMM at a 200 mV ac voltage range using the 300 mV thin-film MJTC. This automated calibration system consists of a highly accurate programmable calibrator (Fluke 5720 A) to precisely source both alternating and direct voltages, highly sensitive, 8.5 digits, DMM (Fluke 8508 A), 8.5 digits, DMM (HP 3458 A) to measure the μPot output emfs and 300 mV thin film MJTC.

The multifunction calibrator, which is the alternating and the direct voltages source is connected using a teeconnector to both the thin film MJTC and the Fluke 8508 A DMM which is recently calibrated at its dc voltage mode.

Alternating voltage in addition to direct voltage in positive and negative polarities is sequentially applied in the sequence (dc^{+}, dc^{−}, ac) from the calibrator. Enough time is allowed during the automated program adjustment to warm up the TE. Also, it is important to wait a frequently predicted time before recording the DMM output electro motive forces (emfs) indicated by HP 3458 A DMM until the MJTC reached its steady-state. This sequence is repeated for ten times. The corresponding MJTC output emfs readings have been automatically recorded from the 3458 A DMM in specially prepared

excel sheet. Furthermore, the 8508 A DMM readings are recorded in the same excel sheet. The difference between the MJTC responses due to the applied dc average voltage and the ac voltage (δ_{diff}) is calculated using the equation:

WhereE_{ac}: output emf for the AC input voltage.

E_{dc}: mean emfs values for forward (dc^{+}) and reverse (dc^{−}) voltages.

n: response characteristic varies between 1.6 and 2, depending on the amplitude of the input signal [

The accurate applied value at 200 mV ac voltage, V_{ac}, is calculated from:

Whereδ_{MJTC}: ac-dc transfer difference of the MJTC.

δ_{diff}: difference between the MJTC responses due to the dc average voltage and the ac voltage.

V_{dc}_{+}, V_{dc}_{−}: the accurate applied dc voltages 8508 A DMM readings and the accurate applied value obtained from equation 2 is then used to obtain the 8508 A DMM correction at its 200 mV ac voltage in order to get its accurate calibrated value.

The second step is the calibration of the 200 mV-μPot using the Fluke 8508 A DMM which is calibrated from the previous step in its 200 mV ac voltage range. This automated calibration system consists of 5720 A calibrator, 8508 A DMM, 3458 A DMM to measure the μPot output emfs, 200 mV-μPot and 900 W resistor.

In order to perform this step, the calibrator is connected to the 200 mV-μPot through a 900 Ω series resistor to pass 5 mA to the μPot. 3458 A DMM is connected to measure the μPot output emf readings.

The 200 mV-μPot ac-dc difference (δ) is determined from the following equation:

Where,

: calibrated 200 mV ac voltage of 8508 A DMM after putting its correction obtained from the previous step.

V_{dc}_{+}, V_{dc}: calibrated 200 mV dc voltages of 8508 A DMM.

δ_{diff}: difference between the 200 mV-μPot responses.

In the third step, the calibrated 200 mV-μPot is used to calibrate the Fluke 8508 A DMM at its ac 100 mV range. The automated calibration system consists of 5720 A calibrator, 8508 A DMM, 3458 A DMM, 200 mV-μPot and 900 W resistor. As shown in

The 8508 A DMM correction at its 100 mV ac voltage is then obtained from the 8508 A DMM recorded readings and from:

WhereV_{ac}: the accurate calibrated ac voltage applied on the 8508 A DMM.

δµ_{Pot}: ac-dc transfer difference of the 200 mV-μPot.

The calibrated 100 mV range DMM is used for the calibration of 100 mV-μPot. Then, the calibrated 100 mV- μPot is used to calibrate the DMM at 50 mV range and so on till reaching to the calibration of the 2 mV-μPot by using the automated calibration systems described in the second and third steps.

The values of the n-factor test of our µPot set are within (1.76 to 1.91). These values of n-factor are very acceptable [

Generally, the considered expanded uncertainty is a round approximation of the calculated uncertainty. For example, the considered expanded uncertainty for the 100 mV-μPot at 40 Hz becomes 30 ppm. _{exp}) of all the calibrated μPots from 200 mV to 2 mV at the specified frequencies 40 Hz, 1 kHz, 10 kHz and 20 kHz.

It is found that the ac-dc differences of the 200 mV to 2 mV µPots are from 3.1 ppm to 150 ppm at the different frequencies from 40 Hz up to 20 kHz. µPots can be used to generate ac voltage signals in millivolt ranges with ac-dc differences ranging from 20 ppm to 1000 ppm [

The established single junction thermal converter µPots set provides a reference standard ac voltage source in the millivolt region. It offers stable output ac voltage ranges from 2 mV up to 200 mV at frequencies range from 40 Hz to 20 kHz. To achieve low ac voltage traceability at NIS, an automated DMM method has been performed for the scaling of ac voltage from 200 mV to 2 mV by means of a step-down procedure. The LabVIEW program which is constructed to automatically calibrate the µPots significantly improves the measurement process as well as it allows statistical proceeding of the results in rather short time. The ac-dc differences for the µPots at different frequencies from 40 Hz to 20 kHz are determined. It is clearly shown that the ac-dc differences of the µPots are within reasonable and acceptable ranges. The expanded uncertainty of the calibrated µPots is also evaluated with 95 % confidence level at these frequencies.

The authors thank Dr. Umberto Pogliano for his guidance and advices.