**Abstract**

In this paper, the stability of the new internal multi-range thermal voltage converter which designed and implemented at NIS, Egypt is studied. It is constructed from multi-range resistors connected in series with a thermoelement which are mounted in the same box. This new thermal voltage converter is used for the ac voltage measurements in the range from 10 V to 750 V. The stability study is applied on the 100 V range as an example. The studied parameters are the drift of the e.m.f with time, sensitivity, and ac-dc transfer difference. The stability of the new TVC is compared with the stability of the traditional TVC. This study proves that the new thermal converter is stable and under control. The electrical model for the new thermal voltage converter is also predicted in this paper by using “LT-Spice” program. The simulated ac-dc difference is calculated from the electrical simulation and compared with the practical ac-dc difference to evaluate the efficiency of the predicted simulation. The evaluation proves that this model is reliable to determine the ac-dc difference in the frequency ranges above 100 kHz.

Keywords: AC-DC Transfer Standard, Multi-Range Resistors, Thermal Converter Characteristics, Electrical Simulation.

**1. Introduction **

AC voltages at frequencies up to 100 MHz is most accurately measured by comparison with dc voltage using ac-dc transfer standard which is also called thermal converter (TC) that responds nearly equal to ac and dc voltages [1]. TCs are widely used in most the national metrology institutes as a primary standards for driving alternating quantities; voltage, current from known dc quantities [2]. The main component of the TC is the thermoelement (TE). The TCs are used for ac voltages up to 1 kV which sufficient for industry by using range resistors that connected in series by a coaxial connector with the TC to provide the thermal voltage converter (TVC).

At NIS, Egypt, a new internal multi-range TVC had been previously designed and implemented at NIS, Egypt to cover the ac voltage ranges from 5 V to 750 V with range resistor values varied from 2 kW to 150 kW [3]. The new TVC is constructed from multi-range resistors connected in series with a single-junction TE. This construction is mounted in the same box. Each range resistor covers a certain ac voltage range. The required one is selected by using a 6-position selector switch. Setting the multi-range resistors combined with the TE in the same box has many benefits. From these benefits is removing the contact resistance between the range resistor connector box and the thermoelement connector box when they mounted separately. Using one TE to cover a wide range of ac voltages reduces the cost. It also helps in protecting the connector from wearing-off during the connection and separation of the two boxes [3].

Better performance for this new multi-range TVC requires putting into account its characteristics such as the dc response time, the drift of the electromotive force (e.m.f) at constant input voltage with time, and ac-dc transfer difference stability. Furthermore for the accurate measurement of ac voltages, it is essential to simulate the ac-dc transfer standard [4] to obtain its ac-dc differences especially at high frequencies. So, this parameters and the electrical modeling of the new TVC is evaluated and studied in this paper.

**2. New TVC Stability**

Accurate measurements by using the new TVC require taking into consideration its specific characteristics such as its response in time, and the drift of the e.m.f at the different ranges [5]. Thus, the stability of the new TVC is automatically studied for the 100 V range as an example by using suitable LabVIEW programs which have been prepared to measure these parameters [6]. The parameters which their stability are studied, are the drift of the e.m.f with time; stability time, dc response time, sensitivity, and ac-dc transfer difference. In this study the new TVC is directly connected to a calibration system Fluke model 5720A, and the TVC output is connected to a reference multimeter (DMM) Fluke model 8508A to measure its e.m.f. All the used instruments are under computer controlled.

**2.1 Stability Time and DC Response Time of the New TVC**

The stability of the new TVC is automatically detected during 2 minutes by using a LabVIEW program. This test determines its dc response time and stability time to put it into consideration during the measurement process of ac voltage to get the most accurate results. Its stability is compared with the stability of the traditional TVC. Fig. 1 shows the stability time of the new TVC and the traditional TVC at 100 V range.

It is shown from the figure that the stability time required for the new TVC at range 100 V to reach to its stability value; 6.866606 mV; is 60 sec. So for accurate measurements, this time should be waited before recording the output e.m.f from the DMM. It`s also found that the new TVC is more stable than the traditional TVC which requires about 115 sec. to be stabled. Fig. 2 shows the dc response time of the new TVC and the traditional TVC at 100 V range.

It is found that the time required for the new TVC at range 100 V to reach to 63% of their stability value is around 2 sec and the traditional TVC requires 4.5 sec.

**2.2 Sensitivity Stability of the New TVC**

The sensitivity of the TVC; factor n is considered a measure of the power dissipated in the heater and varies approximately as the square of the heater current. However, the device deviate significantly from a square low response as the heater current approaches the rated value. The response appears to be nearly square law (n = 2) at very low currents, but n is usually 1.6 to 1.9 at rated heater current [7]. This sensitivity is measured once as a receiving test, and its corresponding value is put into the software program when ac-dc measurements are done by using this TVC. The sensitivity is obtained from [8]:

Where DV is a small change in the heater voltage, DE is the corresponding change in the output e.m.f. and E is the output e.m.f. at the nominal test voltage V.

This test is also performed automatically by using a specific LabVIEW program and repeated several times. Fig. 3 shows the sensitivity stability of the new TVC at 100 V range. It is shown that all the results are under control and stability.

**2.3 Transfer Difference Stability**

The stability means the ability of the standard to keep its performance and characteristics constant with time. For the ac-dc transfer standard, there is a small change in its output value over time and over changes in other variables such as temperature with the input left constant. So, the measurement of the ac-dc difference is automatically repeated seven times. Fig. 4 shows the ac-dc transfer difference stability of the new TVC for 100 V range at frequency 10 kHz. It is shown that all the results are under control and stability which reflect the precision of the new TVC.

**3. Electrical Model of the new TVC**

The most difficult task in forming; designing and implementing a new TVC as an ac-dc transfer standard is determining its high frequency difference [2]. The frequency characteristic of the new TVC is estimated by making an electrical model to simulate its performance at different high frequencies. There are many components affect on the frequency characteristic of the new TVC such as parasitic components in the internal circuit. These parasitic components like lead inductance, stray capacitance, skin effect, and dielectric loss between the high and the low parts of the heater [9]. There are also some components should put into consideration during modeling the circuit such as parasitic capacitances, and heater and internal range resistors resistances and inductances. These components are evaluated by an electrical model and measured by using LCR meter Agilent model E4980A in the frequency range of 1 kHz to 100 kHz.

The evaluation of the predicated model is done by comparing the practical measured results with the simulated results for the practically calibrated ranges to validate it [10]. The agreement between the measured and simulated results confirms the excellent reliability of the predicted model. By using this simulation, the ac-dc difference for the high frequency ranges; 1 MHz and above can be predicated. The used simulation program for this study is “LT-Spice”. The ac-dc differences of the new TVC are measured practically to determine the simulation efficiency at the different frequencies. Fig. 5 shows the electrical model of the new TVC.

The simulated ac-dc differences are calculated by simulating and evaluating the power produced due to flowing ac current in the heater and the power produced due to flowing dc current. Table (1) shows the practical and the simulated ac-dc differences of the 100 V new TVC at the different frequencies by switching on S3 and keeping the other switches off.

Table (1) Practical and simulated ac-dc differences of the 100 V

Nominal Frequency |
Practical ac-dc difference, ppm |
Simulated ac-dc difference, ppm |
Simulation Efficiency, % |

1 kHz |
2.1 |
1.83 |
88.8 |

10 kHz |
-11.9 |
-8.63 |
72.5 |

100 kHz |
13.0 |
-8.89 |
68.4 |

It is shown that the simulation efficiency is high especially at the frequency range 1 kHz which confirms the reliability of the predicted model. This difference is slightly increased by increasing the frequency. This difference because there are some errors become significant in the practical operation such as Tompson and Peltier effects, skin effect, dielectric loss and capacitance and inductance errors. It proves that this model is reliable to determine the ac-dc difference in the frequency ranges above 100 kHz.

**4. Conclusion**

Better performance for the new multi-range TVC requires putting into account its characteristics such as the dc response time, the drift of the e.m.f at constant input voltage with time. Furthermore for the accurate measurement of ac voltages, it is essential to simulate the ac-dc transfer standard to obtain its ac-dc differences especially at high frequencies. So, the stability of the new TVC for some parameters is automatically studied and evaluated for the 100 V range as an example by using suitable LabVIEW programs. It is shown that the stability time required for the new TVC to reach to its stability value is less than the traditional TVC. In addition, the time required for the new TVC to reach to 63% of their stability value is around 2 sec and for the traditional TVC is around 4.5 sec. It proves that decreasing the distributed capacitances and inductances, and eliminating the contact resistance improves the stability of the new TVC than the traditional TVC. The sensitivity and the transfer difference stability show that the new TVC is under control and stability which reflect its precision. The evaluation of the predicated model shows that the simulation efficiency is high especially at the frequency range 1 kHz. This difference is slightly increased by increasing the frequency due to some errors which become significant and appear in the practical operation. It proves that this model is reliable to determine the ac-dc difference in the frequency ranges above 100 kHz.

**References**

[1] Yasutaka Amagai, and Hiroyuki Fujiki, "Improved Electrothermal Simulation for Low-Frequency Characterization of a Single-Junction Thermal Converter", IEEE Transaction on instrumentation and measurement", DOI: 10.1109/TIM.2014.2313952, Published online 2014.

[2] Ilya Budovsky, and Barry D. Inglis, "High-Frequency AC-DC Differences of NML Single-Junction Thermal Voltage Converters", IEEE Transaction on instrumentation and measurement", Vol. 50, No. 1, Feb., 2001.

[3] Rasha S.M. Ali, "New internal multi-range resistors for ac voltage calibration by using TVC", Measurement Science and Technology Journal, Vol. 26, No. 10, Oct., 2015.

[4] Y.Amagai, H.Fujiki, "Numerical and Experimental Investigations of Low Frequency Properties of Single-Junction Thermal Converters", Conference on Precision Electromagnetic Measurements (CPEM), pp. 440-441, 2012.

[5] U. Pogliano, and B. Trinchera, "RMS Voltage Measuring System for Precise Evaluation of Electric Quantities", Metrology and Measurement Systems Journal, Vol. XIV, No. 4, pp. 555-562, 2007.

[6] Rasha S.M. Ali ,“Automatic Evaluation of the Thin-Film Multijunction TVC performance in the Highly Accurate AC Voltage Measurements”, Journal of Control & Instrumentation, Vol. 4, Issue 1, PP. 1-10, 2013.

[7] Earl S. Williams “The Practical Uses of AC-DC Transfer Instruments”, NBS TECHNICAL NOTE 1166, October 1982.

[8] B. Pal, S. Ahmad, and A. K. Govil, "Automation and Evaluation of Two Different Techniques to Calibrate Precision Calibrators for Low Frequency Voltage using Thermal Devices", MAPAN-Journal of Metrology Society of India, DOI 10.1007/s12647-012-0038-5, Published online 11 January 2013.

[9] "Technical Reference for ET2001 Thermal Voltage Converters", Nano-Electronics Research Institute/AIST-Japan, Version 3.01, April 2010.

[10] Rasha Sayed A. Mohammed, “Design and Implementation of Precise Thermal Converters for the Highly Accurate AC Voltage Applications”, ph. D. Thesis, Faculty of Engineering, Ain Shams University, 2009