Comparative Tests of Electric Voltage and Resistance with the Use of Virtual Measurement Systems

The article discusses issues related to virtual measurement systems and presents the results of comparative measurements of resistance and electric voltage measured with various measuring devices, including virtual NI Elvis systems. The accuracy and measurement errors were compared, as well as the advantages and disadvantages of the devices and methods used for the measurements. Moreover, a computer simulation of the measurements in the Multisim software was performed, and its results were compared with the actual indications of the meters.


INTRODUCTION
The constant development of technology in various industries (medical, electronic, chemical, etc.) [ 1,2]. forces the complete replacement of "old" technologies with new solutions, often by improving known technical solutions. Improving the applied solutions instead of replacing them completely allows for a signifi cant increase in effi ciency and quality improvement while keeping the costs of introduced changes low [3]. Classic measurements in electrical metrology are performed with the use of multimeters and analog or digital devices (in newer solutions). There is a trend in the world to switch to digital measurement methods [4]. For over a dozen years dedicated measurement cards and PCs have been used to perform metrological measurements [ 5,6], Many activities and production processes are automated to eliminate monotonous human activities or to save time and money [ 7,3].
By automating a given process (or activity), the user gains more control over it, speeds up the measurements, allows for saving current data to a fi le and for further processing and analysis of the results. Thanks to the appropriate interpretation of variables -by a dedicated algorithm included in the program, the user can be informed (e.g. by means of visual controls) about the failure states of the device and any errors that may occur in it. The purpose of this study was to establish the permissible ranges, accuracy and errors of measurements, and to present the advantages and disadvantages of measuring devices and methods used in modern electronics.
reading the obtained results directly from the display screen on the measuring instrument [8][9][10].

Indirect method
Indirect measurement is the direct measurement of quantities other than the one you want to obtain, but are obtained as a result of calculations or transformations. This method is often referred to as the technical method. For example, for measuring resistance, two instruments should be used: an ammeter and a voltmeter. In an ideal voltmeter, the resistance value should be infi nitely large, and in an ideal ammeter, it should equal 0. Due to the fact that measuring devices have a fi nite internal resistance that falsifi es the measurement result, the ammeter can be connected in two ways, either behind or in front of the voltmeter, the so-called measurement with the method of correctly measured current and correctly measured voltage [8,9,11] .

An example of an indirect method
A conventional system for measuring the internal resistance of a voltage source (indirect method) is based on the use of two multimeters measuring the current and voltage and a manual mechanical switch (Fig. 1). The test consists in connecting the actual system, measuring electrical quantities (measuring the voltage with the switch open, and then measuring the current intensity with the closed switch) and calculating the value of the internal resistance of the voltage source.
The presented measurement method is timeconsuming with a large number of measurements for diff erent voltage sources. The same measurement can be performed using the presented automated measuring stand. Automation of the stand was achieved by connecting the NI ELVIS II* module to a PC into an appropriately modifi ed conventional measurement system (Fig. 2).
In the described measurement system, the hardware part of measurement data acquisition is performed on the National Instruments NI ELVIS II* laboratory module [ 12,13].
NI ELVIS II. It is characterized by a compact housing integrated with 12 of the most commonly used laboratory instruments; including oscilloscope, digital multimeter, function generator, regulated DC power supply. The whole is connected to a PC via a USB cable. On the module one can effi ciently connect and run various prototype electronic circuits, and the presence of a contact plate eliminates the need to solder discrete and integrated elements [3]. In place of "removed multimeters" (for the purposes of connection to the ELVIS set) specially prepared adapters/signal connectors have been added. These are appropriately marked small plastic enclosures with banana plugs installed in them. The modules have been marked accordingly, as: "V" (Volt voltage measurement), "A" (A current measurement), BAT (battery connection) [14].
The role of the mechanical switch was taken over by the relay electronic circuit built on the board included in the NI ELVIS II* set. The actual test stand is presented in Figure 3.
The measurement control program was prepared in the LabVIEW environment of the National Instruments company (G language ) [15,16] (Fig. 4). The prepared program measures the U1, U2 voltage and the current intensity I. After this operation, the obtained results are immediately displayed on virtual meters: voltmeters and a single ammeter. After taking the measurements, the program automatically calculates the resistance value of the voltage source and the load resistance. The program displays the results of the calculations with the value of the voltages U1, U2 Diagram of an automated system for measuring the internal resistance of a voltage source [3] and the current I in appropriately labeled visual indicators (Fig. 5). Additionally, the program interface is equipped with the "Waveform Chart" windows displaying the U1 and U2 voltage values during the course the measurement (in the form of [3]. By replacing the traditional measurement with a virtual measuring instrument, it is possible to quickly compare and demonstrate changes in measurement results for various measured elements of the same type (voltage source resistance for several batteries or resistance for several values, etc.) [3].

COMPARATIVE STUDIES OF RESISTANCE
The results of resistance measurements measured with selected measuring devices were compared, i.e. digital multimeters, analog meters, NI Elvis system, milliohmmeter, whereas simulations of resistance measurements were carried out   [3] using the Multisim application. Analogous tests were performed for the measurements of the electric voltage source. Measurement errors were calculated according to the formulas for each device and measurement method. The absolute measurement error was calculated for each of the measuring devices using the following formula (1): Then, using the formula (2), the relative measurement error was calculated: The absolute errors of the meters were calculated using the formulas (3)(4). For a digital meter and a milliohmmeter acc. to formula (3): For analog meters acc. to formula (4): For NI Elvis acc. to formula (5): ∆X = ppm of reading + ppm of range (5) where: ppm stands for 'parts per milion'.
The research began with the direct method of measuring the resistance. Resistance standards were measured with the values: 0.01 Ω, 0.1 Ω, 1 Ω, 1,000 Ω, 10,000 Ω. Table 1 shows the obtained relative errors in the measurements of the resistance using the above-mentioned instruments.
Analyzing data from Figure 6, it can be noticed that in the direct method the largest measurement error, up to 900%, occurred at very low resistances of 0.01 Ω and 0.1 Ω for analog meters. However, with higher values, above 100 Ω, the error was practically imperceptible. Computer simulation in the Multisim software, based on a mathematical algorithm, is a virtual measurement pattern without errors. Hence, for each of the tested resistances, an error equal to 0 was obtained. An extremely accurate device turned out to be a milliohmmeter based on the Kelvin bridge. For the measured resistances, the relative error did not exceed 0.1%. The NI Elvis circuit appeared to be more precise than the analog meters. The measurement with the NI Elvis system required writing a program in the LabVIEW application (Fig. 7).
Automatic resistance measurement only required connecting a resistor and clicking the "run" button in the application; the fi nished results were obtained in a spreadsheet. The measurement itself took place in a very short time compared to other instruments. The choice of the used ohmmeter has a great infl uence on the fi nal result of the resistance measurement. Figure 8 shows the instrument errors depending on the measuring range.
Subsequently, an indirect method was used to measure the resistance. The measuring system was prepared for measurement with the method of correctly measured current and voltage. Measurements were carried out with the use of the following devices: digital multimeters, analog meters, NI Elvis system, and simulations of measurements were performed in the Multisim software. Figure 9 shows the relative errors of measurements for individual meters.
The greatest measurement errors occur at low resistances. This is due to the inaccuracy of the meters used and the resistance of the wires and terminals, which have a signifi cant eff ect at low resistance values. A distinctive error in both methods occurred using NI Elvis system. The reason for this is a single measuring range of the ammeter. At low current fl ows (while measuring the resistance of 10,000 Ω), the NI Elvis system is seriously fl awed. In addition, the system was unable to measure two values simultaneously:   7. Scheme of the program for measuring resistance using the direct method in LabVIEW voltage and current. This necessitated the use of two platforms and increased the time taken to perform the measurements as the automatic measurement program could not be used. Nevertheless, an additional function of the NI Elvis system to measure the NI Elvis Two-Wire Current-Voltage Analyzer current-voltage characteristic was tested. In the case of resistance, where this dependence is proportional, the function did not work well because the results were very different from the actual value.

COMPARATIVE TESTS OF ELECTRIC VOLTAGE
A laboratory power supply, model APS3005S, was used to test the DC voltage.  From the obtained results presented in Table 2, it can be seen that in the case of the NI Multisim simulation program, the obtained results of the direct current value are equal to the value of the generated voltage. NI Multisim is only a measurement simulation, an equivalent of a virtual reference method based on a mathematical algorithm based on physical formulas.
For very low voltages of 1.5 V, 2.5 V DC, the digital multimeter shows greater measurement accuracy, so its measurement error is smallest, not exceeding 5%. For higher DC voltage values, the DMM achieves measurement results comparable to the results obtained with the NI Elvis device. Then, comparative voltage measurements were made with selected measuring tools. Having considered the errors of the measuring devices, comparative results were obtained for each of the devices (Fig. 10).
Analog meters used for direct current measurements, are marked by a small measurement error in comparison with digital multimeters, and even with virtual devices. The RIGOL model DG1022 digital generator was used to carry out the testing of alternating current, which has the function of generating alternating voltage in the  (Table 3) was performed for an alternating voltage signal of 1.5, 2.5 and 5 V.
For analog meters, the measurement error increases with the voltage value, reaching the level of 6% for the voltage of 5 V. Measurement with a digital multimeter is more accurate for low voltage values, and in the case of voltage increase above 5 V, the measurement results are marked  by an increasing measurement error. The NI Elvis platform and the LabVIEW environments for low sinusoidal voltage are characterized by the best measurement accuracy, with the relative measurement error of up to 0.3%. The authors have conducted an extensive series of analogical tests of alternating current for signals with a triangular and rectangular shape. For these signals also, the values obtained from the compared measuring devices were compared. Interestingly, the NI Elvis layout appeared to be the most accurate. A detailed summary analysis of the practical selection of a measuring device for alternating current is displayed in Figure 11.
Measurements with the NI Elvis platform for alternating current in the form of sinusoidal, square and triangular signals were characterized by the lowest relative errors, none exceeding 0.6%. In turn, for a digital multimeter, it can be seen (Fig. 11) that for a sinusoidal signal, the relative errors reached a maximum of 1.4%. The evidently least precise measurement results originate from analog meters.

CONCLUSIONS
The advantage of virtual measuring systems is not only the operation speed and efficiency, but also the versatility, e.g. one stand equipped with the NI ELVIS II* module can be used in several different configurations and replace several different classic measuring devices. The obtained results of the comparative electric voltage tests indicate that, in the case of measuring the value of the DC electric voltage, the NI Elvis measuring device produces a large measurement error. Compared to analog meters of 2 and 0.5 classes and a digital multimeter, significantly better results were achieved by the NI Elvis platform when measuring AC voltage for three tested signals: sinusoidal, triangular and rectangular. The advantage of the NI Elvis platform is the ability to create measurement programs using the LabVIEW environment. As a result, it is a possibility of saving and post-hoc processing of the measurement results obtained.