Investigations of Electrical and Optical Parameters of Some LED Luminaires—A Study Case

: In the paper, the electrical and optical parameters of wirelessly controlled LED lamps of the Hue type are analysed. Features of this class of wirelessly controlled LED lamps given by the producer are described and a measuring setup for the electrical and optical parameters of the considered LED lamps is proposed. The results of measurements of the two investigated lamps with variable hue, based on RGB chips or emitted white light of different CCTs (correlated colour temperatures), are shown and discussed. Attention was given to the parameters characterising the inﬂuence of the tested LED lamp on such electric power quality indicators as PF (power factor) and THD (total harmonic distortion). Therefore, the inﬂuence of voltage supply and the selected value of power density of the emitted light on the THD and PF is analysed and discussed.


Introduction
Both white LED luminaires (with constant or variable correlated colour temperature (CCT)) and colour LEDs (with constant or variable hue) are used in lighting techniques [1][2][3]. LED lamps consist of four main components [4,5]: LED modules, LED drivers, optical lens and the lamp case with a heat-sink, which makes it possible to effectively remove the heat generated in this lamp.
The important aim of lighting systems is to obtain a desirable colour of emitted light and luminous flux. Classical LED lamps do not show such functionality and the colour or CCT of emitted light is fixed in the stage of production of these lamps. On the other hand, it is known from the literature [1,2,[5][6][7][8][9][10][11][12] that the luminous flux emitted by power LEDs can be easily controlled by changing the value of the LEDs' current, and the colour of the emitted light can be regulated using RGB diodes and by adjusting the value of the current flowing through each diode.
The papers [8][9][10][11][12][13][14][15][16] describe the control methods and application systems for RGB LEDs. The influence of selected RGB LED control methods on the dynamic parameters of lighting systems is also presented. On the other hand, there is a lack of measurement results describing the influence of the parameters of the signal-controlling RGB LEDs on the electrical and optical parameters of selected LED lamps. Thus, these papers present a measurement setup to obtain the RGB LED characteristics.
In the papers [2][3][4][5][6]17,18], some results of experimental studies on the influence of the operating temperature on the operating parameters of solid-state light sources and LED lamps are presented. We can notice a significant impact of the operating temperature of LED lamps on the value of the emitted luminous flux and on the shift in the optical spectrum generated by the mentioned semiconductor light sources. In the paper [19], a closed-loop feedback electronic circuit is proposed for the real-time white point adjustment lamp. The results of the THD measurements are obtained using a simplified method. This parameter describes the influence of harmonic frequencies on the shape of the current received from the electro-energy network. This simplification causes big discrepancies between the measured results of this parameter with a single-phase power analyser and with an oscilloscope. This paper, which is an extended version of the paper [47], presents a comparison of the exploitation parameters of two selected wirelessly controlled LED lamps. One of them emits light with three basic colours (blue, green and red), whereas the other-white light-with three arbitrarily selected values of CCT (warm, cold and neutral). The greatest emphasis is placed on the comparison of the obtained parameters characterising the quality of the electrical energy consumed by the tested LED lamps. The waveform and the spectrum of the current feeding the tested LED lamps are also analysed.
Section 2 contains a description of the wireless control of LED lamps of the Hue type. Section 3 presents the measuring setup used to characterise properties of the tested lamps. In Section 4, the results of measurements of selected characteristics of the examined lamps are provided and discussed.

Tested Lamps
Wirelessly controlled colour and dynamic white LED lamps of the Hue type which emit optical radiation in three different colours: Green, red and blue, or three different CCT values, were selected for experimental investigations. The colour and CCT value of the emitted light were selected by means of the application Philips Hue 3.0.1 dedicated to mobile devices with Windows, iOS and Android operating systems. This application makes it possible to configure up to 50 LED lamps of the Hue type and choose a hue from the palette of sixteen million colours. It also enables the regulation of intensity of the light within the range from 1 to 100% with a step of 1%. The examined lamp of the type White and Colour Ambiance model 9CK is characterised with the maximum value of the emitted luminous flux equal to 806 lm. The electrical power consumed from the electro-energy network is equal to 9.5 W [37].
The colour of the emitted optical radiation in the case of the LED lamp White and Color Ambience is obtained by means of RGB LEDs. These LEDs are controlled with three separate PWM (pulse-width modulation) signals, and the colour of illuminance is regulated by changing the value of the suitable level of the duty cycle for each PWM signal. This is one of the methods to obtain a suitable colour of the emitted light and it is widely described in the literature [4,7,12,18,34]. Communication between the selected LED lamp and the mobile device is based on the ZigBee protocol described in the standard IEEE 802.15.4 [35,37].
An indispensable component of the system of the remote control is a communication bridge-the Hue Control Bridge. It is used for remote communication with the end devices of the Hue system (with LED lamps) [37]. The Hue communication bridge operates in the waveband from 2.4 to 2.484 GHz in the ISM range. The bridge uses this frequency range to communicate with the examined LED lamp and with the mobile device. The Wi-Fi router and the Hue communication bridge are connected with the LAN (local area network) by the RJ-45 cable. These devices are supplied power from the electro-energy network from a feeder producing 5V DC voltage.

Measurement Setup
In order to measure characteristics of the examined LED lamps of the Hue type, a measuring setup was designed and constructed. This setup makes it possible to measure optical and electrical parameters of the examined LED lamps. In order to measure optical parameters of the selected LED lamps, the light-tight measuring-track containing a PCV pipe with a length of 50 cm was built. Figure 1 presents the block scheme of the measurement setup.  This setup contains an autotransformer ATS-REG1.2 [48], two multimeters Unit 804 [49] realising the function of an ammeter and a voltmeter, the current probe Tektronix TCPA300 [50], a single-phase power analyser Tektronix PA1000 [51] and an oscilloscope Gw Instek GDS 2104a [52]. The power supply voltage of the examined LED lamp is regulated with the use of the autotransformer connected to the electro-energy network. The voltage is regulated within the range from 100 to 230 V, making it possible to obtain different values of illuminance and power density of the emitted light for the examined LED lamps of the Hue type.
The illuminance obtained on a surface of the probe at a distance l = 50 cm from the investigated lamp is registered by an illuminance meter L200 [53], whereas the power density of emitted light in the axis of the tested lamps at a distance l from these lamps is registered by a radiometer HD2302 [54]. The measuring probe of these devices was placed at the end of the optical path. In order to measure spectral characteristics of the examined LED lamp, an Ocean Optic USB 650 spectrometer [55] was used. The measurements of optical parameters were software-controlled from the PC.
In order to examine the quality of the electrical energy received from the electro-energy network, an oscilloscope and a current probe were used. These instruments enabled measurements of the spectrum of the current received from the electro-energy network by fast Fourier transformation (FFT) processing. Additionally, the spectrum of this current was also measured by a Tektronix PA1000 single-phase power analyser [15].
The view of the used measurement setup is presented in Figure 2. Tektronix power analysers have highly extensive input circuits, which allow more accurate FFT processing than a good-class oscilloscope. Therefore, the input current waveform of the tested LED lamps can be represented more accurately and with a higher number of harmonics. In the case of the Tektronix PA1000 power analyser, it is possible to register up to 70 harmonics of current, while in the case of spectrum measurements with a digital oscilloscope using the fast Fourier transform (FFT) function, the number of these harmonics does not exceed 20. This is due to the use of less complex input filters. This setup contains an autotransformer ATS-REG1.2 [48], two multimeters Unit 804 [49] realising the function of an ammeter and a voltmeter, the current probe Tektronix TCPA300 [50], a single-phase power analyser Tektronix PA1000 [51] and an oscilloscope Gw Instek GDS 2104a [52]. The power supply voltage of the examined LED lamp is regulated with the use of the autotransformer connected to the electro-energy network. The voltage is regulated within the range from 100 to 230 V, making it possible to obtain different values of illuminance and power density of the emitted light for the examined LED lamps of the Hue type.
The illuminance obtained on a surface of the probe at a distance l = 50 cm from the investigated lamp is registered by an illuminance meter L200 [53], whereas the power density of emitted light in the axis of the tested lamps at a distance l from these lamps is registered by a radiometer HD2302 [54]. The measuring probe of these devices was placed at the end of the optical path. In order to measure spectral characteristics of the examined LED lamp, an Ocean Optic USB 650 spectrometer [55] was used. The measurements of optical parameters were software-controlled from the PC.
In order to examine the quality of the electrical energy received from the electroenergy network, an oscilloscope and a current probe were used. These instruments enabled measurements of the spectrum of the current received from the electro-energy network by fast Fourier transformation (FFT) processing. Additionally, the spectrum of this current was also measured by a Tektronix PA1000 single-phase power analyser [15].
The view of the used measurement setup is presented in Figure 2. Tektronix power analysers have highly extensive input circuits, which allow more accurate FFT processing than a good-class oscilloscope. Therefore, the input current waveform of the tested LED lamps can be represented more accurately and with a higher number of harmonics. In the case of the Tektronix PA1000 power analyser, it is possible to register up to 70 harmonics of current, while in the case of spectrum measurements with a digital oscilloscope using the fast Fourier transform (FFT) function, the number of these harmonics does not exceed 20. This is due to the use of less complex input filters. The total deformation of the waveform of the supplying current can be characterised by means of the coefficient of total harmonic distortion THD. The value of this coefficient is described by the formula where I k denotes the k-th harmonic of the current and n is the number of used harmonics. where Ik denotes the k-th harmonic of the current and n is the number of used ha The unfavourable influence of the device current received from the electr network on this network is typically characterised by the power factor PF meas power analyser. The power analyser used in the setup measures the value of t factor based on the definitional example of the form [46]     The unfavourable influence of the device current received from the electro-energy network on this network is typically characterised by the power factor PF measured by a power analyser. The power analyser used in the setup measures the value of the power factor based on the definitional example of the form [46] where T denotes the period of waveforms of current i in (t) and voltage v in (t), and V in,RMS and I in,RMS are root mean square (RMS) values of the supplying voltage and current, respectively. The calculations were performed using the following formula [15]: where the THD coefficient is estimated by Formula (1) and based on the data measured and collected with the oscilloscope.

Results
In order to test properties of the selected LED lamps, a set of parameter measurements was performed with different RMS values of sinusoidal feeding voltage at a frequency equal to 50 Hz. The measurement setup described in Section 3 was used to measure optical power density, illuminance and RMS values of the input current at the steady state. Additionally, waveforms of illuminance and spectral characteristics of the examined LED lamp were measured. Measurements of the total harmonic distortion THD and power factor PF coefficients were performed for different RMS values of feeding voltage. All the parameters which were measured at different values of feeding voltage were also measured for various values of dimming factor d, characterising the emitted light, in the range from 0 to 100% that was controlled using a special application designed for the Hue system. Figure 3 presents the measured dependences of the supplying current as a function of parameter d and for two different voltage supply values equal to 120 (dashed lines) and 230 V (solid lines), respectively. The measurements were performed for both LED lamps.
supply. The measurements for a colour LED lamp ( Figure 3a) indicate t consumption of current corresponds to red-coloured light, and the least to light. This is associated with the spectral sensitivity of the human eye [56] ing Figure 3b, it should be noted that the highest power consumption corre tral white light, and the lowest-to cold white light. The observed changes ing current correspond to properties of the LED drivers used in the test input current of these drivers depends on the current consumed by the fe order to obtain the selected value of luminous flux, the red LED needs current than the blue or the green LED. For the lamp emitting white lig CCTs, changes in the current Iin,RMS are much smaller than for the lamp em light.  As expected, an increase in the fixed value of parameter d caused an increase in the RMS value of the supplying current, but this increase was nonlinear. In the range of the change in dimming factor d < 30%, the value of the current I in,RMS was nearly constant. In this range, the input current was consumed mostly by the LED driver, whereas only a small part of this current was consumed by the LEDs. The current decreased with voltage supply. The measurements for a colour LED lamp ( Figure 3a) indicate that the highest consumption of current corresponds to red-coloured light, and the least to green-coloured light. This is associated with the spectral sensitivity of the human eye [56]. When analysing Figure 3b, it should be noted that the highest power consumption corresponds to neutral white light, and the lowest-to cold white light. The observed changes in the supplying Energies 2021, 14, 1612 7 of 18 current correspond to properties of the LED drivers used in the tested lamps. The input current of these drivers depends on the current consumed by the feeding LEDs. In order to obtain the selected value of luminous flux, the red LED needs a much higher current than the blue or the green LED. For the lamp emitting white light of different CCTs, changes in the current I in,RMS are much smaller than for the lamp emitting coloured light. Figure 4 shows the dependence of the supplying current I in,RMS as a function of the input voltage V in,RMS in the range from 100 to 230 V for two LED lamps emitting the coloured (a) and white (b) light. The measurements were carried out for a constant value of dimming factor d = 100%.
As can be seen, the value of Iin,RMS current decreases with Vin,RMS voltage ship can be observed for both the tested LED lamps, regardless of the co emitted by the LED lamp. The changes in Iin,RMS current are much bigge emitting the white light. They are equal to 20%, whereas for the colour lam exceed 10%. The obtained relationship Iin,RMS(Vin,RMS) is typical for switch-mo plies. Such power supplies are used as LED drivers in the tested LED lam differences in the values of Iin,RMS obtained at emitted light of different colou CCT result from differences in the forward voltages of the used RGB LEDs The following figures show the dependence of the optical parameter two selected Hue-type LED lamps on the value of parameter d ( Figure 5) an length of the emitted light ( Figure 6). As can be seen, the value of I in,RMS current decreases with V in,RMS voltage. This relationship can be observed for both the tested LED lamps, regardless of the colour and CCT emitted by the LED lamp. The changes in I in,RMS current are much bigger for the lamp emitting the white light. They are equal to 20%, whereas for the colour lamp, they do not exceed 10%. The obtained relationship I in,RMS (V in,RMS ) is typical for switch-mode power supplies. Such power supplies are used as LED drivers in the tested LED lamps. Observed differences in the values of I in,RMS obtained at emitted light of different colours or different CCT result from differences in the forward voltages of the used RGB LEDs [57].
The following figures show the dependence of the optical parameters recorded for two selected Hue-type LED lamps on the value of parameter d ( Figure 5) and on the wavelength of the emitted light ( Figure 6).  The following figures show the dependence of the optical parameters reco two selected Hue-type LED lamps on the value of parameter d ( Figure 5) and on th length of the emitted light ( Figure 6).  In Figure 5, the measured dependences of irradiance at a distance of 50 cm from the LED luminaires on the fixed value of parameter d for both LED lamps: Coloured (a) and dynamic white (b), are shown.
The considered dependence is a monotonically increasing function. It is worth noticing that at the maximum value of parameter d, irradiance is largest for the dynamic white LED lamp, and smallest for the colour LED lamp. It can be noticed that for the colour LED lamp, the values of irradiance corresponding to these colours of the emitted light differ between each other three times. It is easy to observe that the obtained dependence Φ E (d) is nonlinear. It can be approximated by a square function. Such a dependence makes it possible to precisely regulate the emitted luminous flux in the range of its small values.
The measurements of parameters characterising the illuminance of the examined LED lamps were performed, too. These measurements show constancy of this parameter, which testifies to the efficient removal of the heat generated in the examined LED lamps [5,11] or suitable compensation of the changes in internal temperature of light sources by a change in the value of the current flowing through these sources.
In Figure 6, normalised spectral characteristics of the examined colour and dynamic white LED lamps operating respectively in the mode of emission of the red, blue and green light and the cold, warm and neutral white light are presented.   In Figure 5, the measured dependences of irradiance at a distance of 50 c LED luminaires on the fixed value of parameter d for both LED lamps: Colou dynamic white (b), are shown.
The considered dependence is a monotonically increasing function. It is w ing that at the maximum value of parameter d, irradiance is largest for the dyn LED lamp, and smallest for the colour LED lamp. It can be noticed that for the lamp, the values of irradiance corresponding to these colours of the emitted between each other three times. It is easy to observe that the obtained depen is nonlinear. It can be approximated by a square function. Such a dependen It can be noticed that when the lamp operates in the mode of emission of red light, its spectral characteristic is narrowest. In turn, in the mode of emission of blue light, two maxima are visible on the spectral characteristics corresponding to blue and green colours. By analysing the spectrum of white LED lamp radiation, it can be noticed that all the spectra have two local maxima. Warm light is closer to red, neutral light is better balanced and cold light has more blue component. Figures 7-10 show the dependences characterising the shape of the supplying current for two selected LED lamps emitting radiation in the whole visible light range. These dependences were registered with the fixed values of the V in,RMS voltage. Figure 7 illustrates the waveforms of the supplying current of the colour LED lamp (a) and the dynamic white LED lamp (b) for the value of feeding voltage V in,RMS = 230 V and parameter d = 100%. Figure 8 shows the relative values of amplitudes of selected harmonics normalised to the amplitude of the harmonic of the fundamental frequency for the colour LED lamp (a) and the dynamic white LED lamp (b).
In Figure 7, visible deviations in the measured current from the desirable sinusoidal waveform in both the examined lamps are observed. It is worth paying attention to large current impulses occurring in the regions of extrema of the waveforms of the network voltage. The shape of these waveforms shows the capacitive character of the input impedance of the feeder contracted in both the examined lamps and the use of a full wave rectifier in them. The impulses of the supplying current even reach 150 mA. waveform in both the examined lamps are observed. It is worth paying att current impulses occurring in the regions of extrema of the waveforms o voltage. The shape of these waveforms shows the capacitive character of th ance of the feeder contracted in both the examined lamps and the use of a fier in them. The impulses of the supplying current even reach 150 mA.  In Figure 8, the spectra of harmonics of the supplying current of only odd harmonics for both the investigated lamps are visible. The dominant role is played by the first harmonic, but in the measured spectra, harmonics to the number 19 are visible. Higher harmonics defined by the norm [55] are below the resolution level of the used instruments. It should be noticed that the relations between the values of each harmonic in the spectrum of supplying current is different for each colour of light emitted by the tested lamp or for different CCT values characterising white light. Figure 9 illustrates the waveforms of the supplying current of the green light LED lamp (a) and neutral white LED lamp (b) for the fixed value of parameter d = 100%. Figure  10 shows the relative values of particular harmonics normalised to the amplitude of the fundamental harmonics for the green LED lamp (a) and neutral white LED lamp (b). By analysing Figure 9, it can be observed that the LED lamp emitting green light is characterised by a more distorted supplying current waveform. The level of distortions increases with the decrease in the RMS voltage V in,RMS , which manifests itself in higher distortions of the instantaneous value of the current waveform supplying the LED lamp. It should also be noted that for the LED lamp emitting the green light, the peak-to-peak value of the supplying current does not change with the change in the RMS value of the line voltage and amounts to approximately 250 mA. Analysing the shape of the supplying current of both the values of power supplies of selected LED lamps, the capacitive nature of the impedance of solid-state light sources can be noticed.
for both the investigated lamps are visible. The dominant role is played by the first har monic, but in the measured spectra, harmonics to the number 19 are visible. Higher har monics defined by the norm [55] are below the resolution level of the used instruments. should be noticed that the relations between the values of each harmonic in the spectrum of supplying current is different for each colour of light emitted by the tested lamp or fo different CCT values characterising white light.  for both the investigated lamps are visible. The dominant role is played by the first har monic, but in the measured spectra, harmonics to the number 19 are visible. Higher har monics defined by the norm [55] are below the resolution level of the used instruments. I should be noticed that the relations between the values of each harmonic in the spectrum of supplying current is different for each colour of light emitted by the tested lamp or fo different CCT values characterising white light.         The analysis of the results presented in Figure 10 shows that the LED lamp emitting the green light is characterised by higher values of the normalised harmonics compared to the white LED lamp, which also confirms the greater current waveform distortion of the green LED lamp.
Using the values of each harmonic of the supplying current measured by means of the current probe and the oscilloscope, the THD values are calculated for the supplying current of the colour and dynamic white lamps operating in the modes of emission of the whole spectrum of visible light. The values of the THD calculated using Formula (1) for different values of feeding voltage are marked in Figure 11  It should also be noted that for the LED lamp emitting the green light, the peak-tovalue of the supplying current does not change with the change in the RMS value o line voltage and amounts to approximately 250 mA. Analysing the shape of the supp current of both the values of power supplies of selected LED lamps, the capacitive n of the impedance of solid-state light sources can be noticed.
The analysis of the results presented in Figure 10 shows that the LED lamp em the green light is characterised by higher values of the normalised harmonics comp to the white LED lamp, which also confirms the greater current waveform distorti the green LED lamp.
Using the values of each harmonic of the supplying current measured by mea the current probe and the oscilloscope, the THD values are calculated for the supp current of the colour and dynamic white lamps operating in the modes of emission o whole spectrum of visible light. The values of the THD calculated using Formula (1 different values of feeding voltage are marked in Figure 11 by means of dashed lines results of measurements of the THD value of the supplying current of the examined performed with the use of the Power Analyzer PA1000 are also shown in the figure. T results are marked with solid lines. These analyses were performed in a wide ran changes in the RMS voltage Vin,RMS. As can be observed in Figure 11a, the THD values of the supplying current stro As can be observed in Figure 11a, the THD values of the supplying current strongly depend on the colour of the emitted light and on feeding voltage. Analysing the results obtained by means of the power analyser, it is visible that in the considered range of changes in the feeding voltage, the maximum of the dependence of THD(V in ) occurs. For each of the considered colours of the emitted light, the lowest THD value at the nominal value of feeding voltage equal to 230 V is obtained. At this value of feeding voltage, the greatest THD value of current of the power supply is obtained at the emission of red light and reaches 65%. The lowest THD value equal to 45% is obtained at the emission of green light. In the case of the dependence of THD(V in ) calculated with the use of Formula (1) and the measurements performed by means of the oscilloscope, it is clearly visible that these values are considerably different from the values obtained by means of the power analyser. The differences in the THD values obtained by means of both methods even reach 15%.
As can be observed in Figure 11b, the THD values strongly depend on the CCT of the emitted light and feeding voltage. It can also be noticed that the dependences of the THD coefficient as a function of the feeding voltage for warm and neutral white light are monotonically increasing functions. With the nominal value of the voltage of the tested LED lamps equal to 230 V, it can be observed that the highest THD value is obtained for neutral white light and is 59%, while the lowest THD value is obtained for warm white light and is equal to only 46%. In the case of the values calculated using Formula (1) and the measurements made with the PA1000 power analyser, it can be seen that these values differ significantly from each other. The observed differences between the THD values obtained with the use of various measuring devices result from the construction of these devices and can even reach 5%.
These differences between results of THD obtained from measurements and calculations can be a result of the finite resolution of instruments used to measure the values of each harmonic. As it was mentioned above, this resolution does not make it possible to measure high-order harmonics, which can influence the value of THD. Figure 12 shows the dependence of the power factor PF on the RMS voltage of feeding voltage V inRMS . The measurements were carried out for two LED lamps emitting light in the spectrum of visible light. These analyses were performed in a wide range of changes in the RMS value of feeding voltage from 100 to 230 V. The values of the power factor PF were measured using a single-phase power analyser PA1000.
It is visible that the value of the power factor is a decreasing function of feeding voltage. At the nominal value of feeding voltage, the value of the power factor is contained in the range from 0.35 to 0.4 for the colour lamp. This is a very small value (lower than 0.5 declared by the producer [37]), which shows that the considered lamp has a very unfavourable influence on the power plant. For the lamp emitting white light, the value of PF is in the range from 0.55 to 0.7 (lower than 0.8 declared by the producer [37]). The obtained results of measurements show that no power factor correction circuit was applied in the feeder of the examined lamp. depend on the colour of the emitted light and on feeding voltage. Analysing the results obtained by means of the power analyser, it is visible that in the considered range of changes in the feeding voltage, the maximum of the dependence of THD(Vin) occurs. For each of the considered colours of the emitted light, the lowest THD value at the nominal value of feeding voltage equal to 230 V is obtained. At this value of feeding voltage, the greatest THD value of current of the power supply is obtained at the emission of red light and reaches 65%. The lowest THD value equal to 45% is obtained at the emission of green light. In the case of the dependence of THD(Vin) calculated with the use of Formula (1) and the measurements performed by means of the oscilloscope, it is clearly visible that these values are considerably different from the values obtained by means of the power analyser. The differences in the THD values obtained by means of both methods even reach 15%.
As can be observed in Figure 11b, the THD values strongly depend on the CCT of the emitted light and feeding voltage. It can also be noticed that the dependences of the THD coefficient as a function of the feeding voltage for warm and neutral white light are monotonically increasing functions. With the nominal value of the voltage of the tested LED lamps equal to 230 V, it can be observed that the highest THD value is obtained for neutral white light and is 59%, while the lowest THD value is obtained for warm white light and is equal to only 46%. In the case of the values calculated using Formula (1) and the measurements made with the PA1000 power analyser, it can be seen that these values differ significantly from each other. The observed differences between the THD values obtained with the use of various measuring devices result from the construction of these devices and can even reach 5%.
These differences between results of THD obtained from measurements and calculations can be a result of the finite resolution of instruments used to measure the values of each harmonic. As it was mentioned above, this resolution does not make it possible to measure high-order harmonics, which can influence the value of THD. Figure 12 shows the dependence of the power factor PF on the RMS voltage of feeding voltage VinRMS. The measurements were carried out for two LED lamps emitting light in the spectrum of visible light. These analyses were performed in a wide range of changes in the RMS value of feeding voltage from 100 to 230 V. The values of the power factor PF were measured using a single-phase power analyser PA1000. It is visible that the value of the power factor is a decreasing function of feeding voltage. At the nominal value of feeding voltage, the value of the power factor is contained in the range from 0.35 to 0.4 for the colour lamp. This is a very small value (lower than 0.5 declared by the producer [37]), which shows that the considered lamp has a very unfavourable influence on the power plant. For the lamp emitting white light, the value of PF is in the range from 0.55 to 0.7 (lower than 0.8 declared by the producer [37]). The obtained results of measurements show that no power factor correction circuit was applied in the feeder of the examined lamp.
PF values were also calculated using Equation (3) and the calculated values of THD shown in Figure 11. The obtained values of calculations are much higher than the measured values of this parameter. For example, at VinRMS = 230 V, values of the PF are in the range from 0.85 to 0.95, and they are higher than the value declared by the producer in [37]. According to information given in [15], Equation (3) can be used only in the case when the value of THD is nearly zero.
Comparing the obtained results of the measurements of the power factor to the values of this parameter obtained for other light sources, and among others in the paper [25], it can be stated that the use of the examined lamps, to a considerably higher degree, results in the worsening of the quality of electrical energy in the power plant than classical bulbs and popular LED lamps. The obtained values of PF for the examined lamp are nearing the values corresponding to a fluorescent lamp.

Conclusions
In the paper, the results of measurements and calculations illustrating electrical and optical properties of LED lamps of the Philips Hue-type emitting light of selected colours or white light of different CCTs were presented. A manner of regulating useful parameters of the considered lamps was presented, and the characteristics illustrating the relation between the values of the selected optical parameters and the measured values of the emitted luminous flux and the spectral characteristics of the examined lamps were shown.
Attention was given to properties of the LED drivers feeding lighting components of the examined lamps. The influence of voltage supply and the colour or CCT of the emitted light and dimming factor d of the lamp on the RMS value of the supplying current, its waveform, the coefficient of the total harmonic distortion and the power factor was discussed.
The measurements were performed with the use of the authors' measuring setup. PF values were also calculated using Equation (3) and the calculated values of THD shown in Figure 11. The obtained values of calculations are much higher than the measured values of this parameter. For example, at V inRMS = 230 V, values of the PF are in the range from 0.85 to 0.95, and they are higher than the value declared by the producer in [37]. According to information given in [15], Equation (3) can be used only in the case when the value of THD is nearly zero.
Comparing the obtained results of the measurements of the power factor to the values of this parameter obtained for other light sources, and among others in the paper [25], it can be stated that the use of the examined lamps, to a considerably higher degree, results in the worsening of the quality of electrical energy in the power plant than classical bulbs and popular LED lamps. The obtained values of PF for the examined lamp are nearing the values corresponding to a fluorescent lamp.

Conclusions
In the paper, the results of measurements and calculations illustrating electrical and optical properties of LED lamps of the Philips Hue-type emitting light of selected colours or white light of different CCTs were presented. A manner of regulating useful parameters of the considered lamps was presented, and the characteristics illustrating the relation between the values of the selected optical parameters and the measured values of the emitted luminous flux and the spectral characteristics of the examined lamps were shown.
Attention was given to properties of the LED drivers feeding lighting components of the examined lamps. The influence of voltage supply and the colour or CCT of the emitted light and dimming factor d of the lamp on the RMS value of the supplying current, its waveform, the coefficient of the total harmonic distortion and the power factor was discussed.
The measurements were performed with the use of the authors' measuring setup. The measurements of the RMS value of the supplying current show that it is an increasing function of parameter d and a decreasing function of voltage supply. The waveforms of the supplying current registered during the experimental part of the research confirm strong deformation, and they can constitute a proof that in the LED driver, a block of the power factor correction was not applied. The performed measurements and calculations of the THD and PF also showed that the examined lamps unfavourably influence the quality of energy in the power plant. One should pay attention to the fact that the values of the THD and PF obtained by means of the power analyser accept far less favourable values than in the case of using the oscilloscopic measurements and the simplified analytic formulas given in the literature. The observed differences result probably from the fact that the waveband, wherein the harmonics values of the supplying current are measured, is considerably wider in the case of the power analyser. When the oscilloscope is used, the fast Fourier transformation taking into account only about 20 harmonics frequencies of the network are identified, but the norm [55] needs 25 harmonics. Therefore, the disturbances of frequency higher than 1 kHz are not taken into account while determining the THD value. By contrast, the power analyser uses 70 harmonics. The obtained results show that it is worth taking into account a change in the law regulations and increase in the number of harmonics of the supplying current for LED lamps connected to the power plant.
The findings presented in this paper can be useful for designers of lighting systems with LED lamps. One should also take into account the influence of the properties of electronic devices on the power plant to make allowance for harmonics of higher order which correspond to the current impulses connected with the use of switch-mode power supplies in the LED driver contained in LED lamps. The objective of further investigations would be a proposal of systemic solutions making it possible to decrease the THD and increase the PF value characterising the waveforms of the supplying current of the examined LED lamps.

Conflicts of Interest:
The author declares no conflict of interest.

Abbreviations
T j junction temperature T a ambient temperature P opt optical power E e surface power density of the emitted light Z the (t) electrical transient thermal impedance v D forward voltage i D diode current