0% found this document useful (0 votes)
56 views6 pages

Switching Power Supplies:: Analysis of Waveform Distortion and Absorbed Powers

Download as pdf or txt
Download as pdf or txt
Download as pdf or txt
You are on page 1/ 6

Switching Power Supplies: Analysis of Waveform Distortion and Absorbed Powers

Roberto Langella and Alfredo Testa


Dipartimento di Ingegneria dellInformazione Second University of Naples Aversa (CE), Italy Email: roberto.langella@ieee.org, alfredo.testa@ieee.org

Abstract Analyses to have a quantitative idea of the real behavior of Switching Power Supply (SPS) units from the point of view of waveform distortion and power absorption are developed. Results of several experimental tests performed by means of a test system built up at Laboratories of the Second University of Naples (Italy) are firstly reported. Then, suitable models are proposed to easily predict the behavior of SPS sets supplied by the same busbar when their number varies and these models are validated by means of accurate time domain simulations. Experimental and numerical results are discussed. Keywords- Power Quality, Harmonic Pollution, Revenue Meters.

distribution networks supplying single-phase residential customers due to the large diffusion of compact fluorescent lamps and, particularly in Italy and in other similar climate countries, of air conditioning apparatus (and other domestic apparatus) based on inverter technology. So in the future, the percentage of power absorbed by SPS on the total power for each customer will overcame by far the percentage of linear loads. On the other hand, the commercial revenue meters of active and reactive energy, generally, use algorithms and techniques developed for the sinusoidal regimen and that, extended to the non-sinusoidal periodic regimen, introduce de facto metrics that sometimes correspond to the definitions of the powers introduced in the relevant literature and standards and sometimes not [1-4]. The use of these metrics could have consequences on the distribution efficiency and costs attributing to the utilities the costs of incremental power losses due to the harmonic circulation. Unavoidably, this will reflect in increased energy price also for those customers that do not significantly contribute to the harmonic pollution [5-8]. In this paper, to have a quantitative idea of the real behavior of switching power supply units from the point of view of waveform distortion and power absorption, the results of several experimental tests performed by means of a test system built up at Laboratories of the Second University of Naples (Italy) are firstly reported. Then, suitable models are proposed to easily predict the behavior of SPS sets when the number of sets varies and these models are validated by means of accurate time domain simulations. In what follows, after some recalls on basic definitions, the experiments performed are described and the models proposed are reported together with some significant experimental and numerical results.

I.

INTRODUCTION

The majority of modern electronic equipments use switching power supplies (SPSs). These differ from older units in that the traditional step-down transformer and rectifier is replaced by direct controlled rectification of the supply to charge a reservoir capacitor from which the direct current for the load is derived by a method appropriate to the output voltage and current required. The advantage is that the size, cost and weight is significantly reduced and the power unit can be made in almost any required form. The disadvantage is that the power supply unit draws pulses of current which contain large amounts of third and higher harmonics and significant high frequency components. A simple filter is fitted at the supply input to bypass the high frequency components from line and neutral to ground but it has no effect on the harmonic currents that flow back to the supply. This means that these units interact with the supplying network exchanging harmonic powers and not only absorbing power at fundamental power frequency. The diffusion of the aforementioned kind of load is expected to be more and more high in next years in LV

II.

SOME BASIC DEFINITIONS OF POWERS UNDER NON-SINUSOIDAL CONDITIONS

For steady-state conditions, under the hypothesis that DC and interharmonic components are not present, a nonsinusoidal instantaneous voltage or current has two distinctive components: the power system frequency components v1(t) and i1(t), and the remaining terms vH(t) and iH(t) that contain all the other harmonics:
v(t ) = v1 (t ) + v H (t ) = 2V1 sin (t 1 ) +

h 1

2Vh sin (ht h ) , (1)

N = 1 Q v(k + ) i (k ) , (7) d N k =1 with =N/4 which corresponds the numerical implementation of: 1 +T1 T (8) Q = v t + 1 4 i (t )dt , T1 which is a definition valid only in sinusoidal regimen. It is possible to demonstrate by means of simple mathematical manipulations [7] that (8) can be seen as: =Q + (9) Q

i (t ) = i1 (t ) + iH (t ) = 2 I1 sin (t 1 ) +

h 1

2 I h sin (ht h ) .

(2)

A. Active and Reactive Power and Power Factor In single-phase case, denoting the fundamental power period by T1, the fundamental power frequency by f1 = 1 / T1 , the total active power can be defined as:

that is to say the summation of Q1 with an error term, d, which is function of the odd harmonic active powers and of the even harmonic reactive powers. As for the power factor, it is: P P d = . (10) P = Fd 2 2 2 2 +Q P + (Q1 + d ) P d d
Finally, it is important to observe that, with the instrumentation considered, only the active power measurement corresponds with the definitions introduced in literature while, the reactive power measurement and, as a consequence, the power factor introduce incorrect metrics de facto at least meaningless. III. LABORATORY EXPERIMENS

P=

1 T1

+ T1

Vh I h cos h v(t ) i(t )dt = h =1

(3)

where is the instant in which the calculation of P starts and being H=N/2 the maximum harmonic order according to the sampling theorem and h=h-h the phase angle difference between the phasors Vh e Ih. On the other hand, the Budeanus definition of reactive power is widely adopted, QB:

T h =1 The IEEE Standard 1459 [4] states that the power factor definition: P (5) PF 1 = 1 S1 plays the same significant role that the fundamental power factor has in non-sinusoidal single-phase systems giving reliable information about the lines utilization. QB =

+T

i1 (t ) v1 (t )dt dt + QBH = Vh I h sin h (4)

To have a quantitative idea of the real behavior of an SPS unit, several experimental tests were performed by means of a test system built up at Laboratories of the Second University of Naples. A. The switching power supply under study The switching power supply unit considered for experimental activities is the AT unit of 150 W of rated Power. The ac supply voltage is rectified and filtered to obtain an unregulated DC voltage. Then, a DC/ DC conversion operated by means of a high frequency solid state switch, a high frequency transformer and secondary rectifiers and filters allow to obtain the requested different DC voltages. A feedback control of the switch Base/gate drive circuit assures exact values of the output voltages of +5, +12, -5, -12 V. To the purposes of the experimental activities performed, the supply unit has been shared into three parts as shown in Fig. 1, obtaining six measurement sections. Load resistances R5V and R12V were added to emulate the presence of a real PC load in different working conditions, that is to say to control by varying their values the output ( and input) powers. Fig. 2 scheme defines also the symbols utilized for the quantities in the result discussion. B. The Test System A simplified block diagram of the test system is reported in Fig. 2. It is based on: arbitrary waveform power generator (Pacific Power source 3120AMX), power analyzer (Norma D6000), PXI system.

B. Metrics de facto adopted by commercial digital meters For the digital meters fs is the sampling frequency of the instrument and N=T1/fs the number of points in a period of the fundamental frequency f1=1/T1. Here, it is assumed, moreover, the absence of causes of uncertainty, that is to say ideal transducers, perfect synchronization between fs, N and T. What remains is the uncertainty related only to metric adopted. the reading For the measure of the active power, let us call P
d

of the digital meter:


N = 1 v( k ) i (k ) , (6) P d N k =1 It is evident that the results of (6) coincide with those of (3), so digital meters are not able to separate the system power frequency contribute from harmonic contributes. As for the reactive power measurement, let us consider as Q
d

the power reading of the digital meter:

IA

IB

Measurement Section Cb

ICb

Measurement Section Db I

Db

VCb VA VB VCa
Measurement Section A Measurement Section B Meas. Section Ca

VDb ICa VDa


Measurement Section Da

R12V IDa R5V

Figure 1. SPS Measurement sections.

The main characteristics of the 3-phase arbitrary waveform power generator produced by Pacific Power (model 3120AMX) are: i) Maximum Power: 12 kVA; ii) Frequency Range: 20 Hz to 50 kHz; iii) Line Regulation: 0.027 mV; iv) Load Regulation: 0.00135 mV; v) THD: 0.1 %; vi) Voltage Ripple and Noise: -70 dB. The main characteristics of the power analyzer produced by LEM (model Norma D6000) are: i) overall accuracy of 0.05% for current and voltage measurements and <0.1% for power measurements; ii) frequency range: from DC to 1 MHz. All the instrumentations are controlled by a PXI system produced by National Instruments (model PXI-1020) equipped with a controller NI PXI 8176 with a 1.2 GHz processor, and 128 MB ram, data acquisition board NI PXI-4472, arbitrary waveform generation board NI PXI-5411. The software implementing the test procedures and coordinating the devices has been developed in LabView environment and everything has been made automatic, thanks to the remote control of all the instruments.
TEST SYSTEM Single phase voltage supply

The results obtained in terms of voltage, current, active power and reactive power in case of 36 W of power supplied to final loads (R5V and R12V) are reported for the most significant harmonic orders in Tabs. I- II; for each quantity both absolute value and relative value as percentage of the corresponding fundamental power frequency value are reported. The metrological characterization of the Test System warranties for the results accuracy.
2.5 Real Supply Ideal Supply

a)

100%
2

1.5 Voltage [%] 1 0.5 0

50

150

250

350 450 Frequency [Hz]

550

650

750

100 Real Supply Ideal Supply 90

b)

Ctrl lines

80

70

60 Current [%]

50

PXI
IEEE 488

40

30

Figure 2. The test system.

20

10

C. Experimental results The results obtained with reference to two different supply conditions are reported in Fig. 3. Fig.3a shows the spectrum of supplying voltages: white bars stay for the minimum THD supply conditions that can be obtained by means of the available Power Source, that means ideal laboratory conditions, and black bars for a real condition corresponding to measured distortion in the LV network supplying the laboratory. Fig. 3b shows the harmonic current spectrum absorbed by a SPS in conditions that correspond to supply voltage of Fig. 3a. By comparing the ideal spectrum with the real one, the effects of voltage distortion are evident and correspond to a remarkable reduction of the harmonic current absorbed when the supply voltage is distorted.

50

150

250

350 450 Frequency [Hz]

550

650

750

Figure 3. Harmonic spectrum during laboratory tests on a SPS unit, in two different conditions: white bars stay for ideal supply and black bars for real supply conditions: a) applied voltages, b) absorbed currents.

Tab. I refer to the ideal supply conditions, while Tab. II to distorted supply conditions. The sign of harmonic active powers of section A are always negative evidencing that SPS acts as power generator at harmonic frequencies. By comparing among them the results in the same section in the different supply conditions, it results evident the variations of a noticeable amount regard exclusively the measuring section A quantities, that is to say the absorbed quantities. In particular, the injected harmonic powers absolute values are sensibly greater in case of distorted supply.

TABLE I. Meas. Section A A A A B B Ca Cb Da Db Freq [Hz] 50 150 250 350 0 100 0 0 0 0 V [V] 220.33 0.21 0.10 0.12 303.37 0.07 155.16 155.18 4.91 12.14 % 100 0.0 0.0 0.0 100 0.0 100 100 100 100

IDEAL SUPPLY CONDITIONS I % [A] 0.31 100 0.29 95.5 0.27 88.9 0.24 79.0 0.20 100 0.29 145 0.16 100 0.17 100 4.86 100 1.01 100 P [W] 66.2 -0.1 0.0 0.0 60.6 0.0 25.2 25.9 23.9 12.3 % 100 0.0 0.0 0.0 100 0.0 100 100 100 100 Q [VAr] -13.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 % 100 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Supply conditions Ideal Conditions Distorted Conditions

TABLE III. [%] 91.7 92.5 BC [%] 84.3 84.8

SPS EFFICIENCIES CD [%] 70.8 69.9 [%] 54.8 54.8 ' [%] 91.5 90.6 ' [%] 54.7 53.7

IV.

MODELING

TABLE II. Meas. Section A A A A B B Ca Cb Da Db Freq [Hz] 50 150 250 350 0 100 0 0 0 0 V [V] 225.94 3.56 3.98 3.90 301.88 0.05 154.32 154.40 4.9 12.14

DISTORTED SUPPLY CONDITIONS % 100 1.6 1.8 1.7 100 0.0 100 100 100 100 I % [A] 0.30 100 0.25 82.3 0.16 54.2 0.07 24.2 0.20 100 0.27 135 0.16 100 0.17 100 4.8 100 0.99 100 P [W] 66.7 -0.5 -0.7 -0.2 60.4 135.4 25.1 26.1 23.7 12.1 % 100 0.7 0.9 0.4 100 0.0 100 100 100 100 Q [VAr] -11.1 -0.7 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 % 100 6.6 0.0 1.3 0.0 0.0 0.0 0.0 0.0 0.0

In general, it is necessary to take into account the interactions of SPSs with the supply network and the other loads as illustrated in Fig. 4 with reference to a simple situation. The harmonic currents injected by a SPS, Ihin, depend on the combined effects of the background harmonic voltage, Eh, and of the voltage drop caused by the entire amount of harmonic current, including the other load contributions, Ihn, absorbed on the same feeder in the system equivalent harmonic impedance, Zh.

Vn E
h

SPS1 SPS2

Zh

Ihn
PCC

Ihin

SPSn-1 SPSn

Figure 4. Equivalent circuit for modelling a power supply of a SPS taking into account of pre-existing harmonic voltages.

D. Conversion Efficiency Starting from the results shown in Tabs. I - II, the SPS efficiencies have been evaluated considering the cascade of elements and measuring sections in Fig. 1. In what follows denotes the total efficiency from the input to the output while ij the efficiency from section i to section j. It results: P + PCb P AB = B 100, BC = Ca 100, PA PB (11) P + PDb P + PDb CD = Da 100, = Da 5 100, PCa + PCb PA
where reference is made to the total powers defined according to (3). In case of input power at section A, the difference between the input total power and the input power at the fundamental power frequency is remarkable, and it is interesting to define also the efficiencies evaluated with reference to powers absorbed at fundamental frequency, 'AB and ': P +P P (12) ' AB = B 100, ' = Da 1 Db 100. 1 PA PA

It is useful to refer to a simple condition in which i) the only nonlinear loads present in a LV distribution feeder are SPSs, of the same power and identical characteristics, balanced over the three phases, and ii) the pre-existing voltage distortion Eh is nill. In particular, the second hypothesis in reality is near to be verified for third harmonic voltages. The single-phase network equivalent impedance as seen by the PCC is assumed, for simplicity sake, as:

Zh=Rh+jXh,
h h

(13)

where R and X take into account, in a single phase equivalent, the effects of the frequency value and of sequence of harmonic currents at each harmonic order, in the hypotheses of symmetry assumed. In particular, for triplen harmonic the presence on the neutral conductor of the three identical currents originating from the three different phases and of the conductor reduced section are considered. The following hypothesis is assumed for the currents absorbed by each load when the number of connected loads varies from 1 to a reference value N, per phase:
h h I in = I11 n h I11 I1hN , N

(14)

The results obtained are reported in Tab. III. It is interesting to observe as the efficiency estimated with reference to the fundamental frequency powers, ', are sensibly lower than the corresponding ones obtained for total powers, ; this is particularly evident in distorted conditions. In fact they take into account also the harmonic powers injected into the supply system and dissipated in it.

being Ih11 the current absorbed by a single load when it is the only load connected and Ih1N the current absorbed by a single load when the total number of connected loads is N per phase. An example of spectra of these two currents are reported in Fig. 3b (white bars for Ih11 and black bars for Ih1N). The total current flowing in the single-phase equivalent line impedance is:
h h In = n I in .

(15)

V Voltage n [V] [V] PCC

In these conditions, the voltage drop on the line impedance produced by n loads per phase is:
h Vnh = Z h nI in ,

20

a)

10

(16)
h n

150

250

350

450

550

IN Current [A]

with a contribution by a single load equal to:

4 2 0

b)

V = Z I =V /n .
h in h h in

(17)

Ihn [A]

h I1N I [A] [A] in Current

The contribution of a single load to the total harmonic active power dissipated on the line resistance is:
h h = R n I in Pin

150

250

350

450

550

0.4 0.2 0

c)

[ ]

h Ih Ih = R n I11 n 11 1N = N
2

150

250

350 Frequency [Hz]

450

550

= Rn I

[ ]

h 2 11

(N n)k N + n N n n h 2 + (1 = R I11 n NkN NkN N

[ ]

Figure 5. PCC voltage, Vn, total current through Zh, In, and current absorbed by the i-th load, Iin, for harmonic orders from 3 to 11 and for n = 1,,10.
0

a)

P npower [W] [W] Active

8) being kN=Ih11/Ih1N and the total Joule losses are: (19) Equations (13)-(19) allow to estimate the behavior of n SPSs per phase with approximations deriving from the assumption of the relation (14) which assumes a linear proportion between the variations of the current injected by each SPS and the supply voltage variation for each harmonic order. More in general, (14)-(19) could be written substituting Ih11 with Ih1m, being m a number of loads less than N. In the reality, the behavior of the SPS input rectifier is as far from that of the proposed model as the harmonic order increases. Furthermore, in real situations the non linear loads are not all equal and their supplying nodes are not coincident and this determines cancellation effects which can be taken into account by the so called diversity factors [9], not considered here for the sake of simplicity. V. NUMERICAL EXPERIMENTS
h , Pnh = n Pin

-10 -20 -30 -40 -50 150 250 350 450 550

0 Active in power [W] -1 -2 -3 -4 -5 150 250 350 Frequency [Hz] 450 550

b)

Figure 6. total active power absorbed by the loads, Pn, and the active power absorbed by the i-th load, Pin, for harmonic orders from 3 to 11 and for n = 1,,10.

V3 [V]

To have an idea of the approximations introduced by (14) numerical simulation were performed by means of the tool Power Systems Blockset (PSB) under Matlab with reference to an SPS whose current spectrum in sinusoidal supply conditions is the same of Fig. 3b (white bars) and to a supply busbar characterized by an impedance similar to that of the line supplying the laboratory in the experiment in Sec. III. The number of SPS per phase considered has been varied from 1 to 10 with unitary increments. Fig. 5 illustrates the results of the simulation in terms of PCC voltage, Vhn, total current through Zh, Ihn, and current absorbed by the i-th load, Ihin, for harmonic orders from 3 to 11 and for n = 1,, 10. Fig. 6 reports the total active power absorbed by the loads, Phn, and the active power absorbed by the i-th load, Phin, for harmonic orders from 3 to 11 and for n = 1,, 10. Figs. 5 and 6 allow to appreciate the different behaviors at the different harmonic orders. The cumulating effects related to an increasing number of SPS are more evident at lower harmonic orders, mainly for the 3rd harmonic.

Figs. 7 and 8 allow a comparison among the Model results and the PSB simulation results for 3-rd (a), 5-th (b) and 7-th (c) harmonics in terms of PCC voltages, Vhn and Vhin , and of PCC active powers, Phn and Phin , respectively, versus the number of SPSs per phase. The inaccuracy effects introduced by approximation (14) in the model results are greater for n = 5, 6, that is far from the extreme values (1 and 10) where they are zero, increase with the harmonic order and affect powers more sensibly than voltages.
3 2 1 0 1 2 3 4 5 6 7

Ph [W]

a)
Simulations Analitic

V 3n V 3in
8 9 10

Harmonic V5 voltage [V] [V]

V 5n V 5in

b)

10

1.5 1 0.5 0 1 2 3 4 5

V7 [V]

V 7n V 7in
6 7 8 9 10

c)

Figure 7. PCC voltages, Vn and Vin , versus the number of SPS per phase for 3rd (a), 5th (b) and 7 th (c) harmonics: __ PSB simulations, Model results.

60 P3 [W] 40 20 0 1 2 3 4 5

a)
Simulations Analitic

P3n P3in
6 7 8 9 10 supply efficiency Supply efficiency 0.95

Active Power P5 [W] [W]

b)
2

0.9

P5n
1 2 3 4 5 6

P5in
7 8 9 10

0.85

0 1

P7 [W]

0.5

P 7n P7in

c)

0.8

S of (20) S of (21)
1 2 3 4 5 6 7 8 9 10

0.75 1 2 3 4 5 6 7 8 9 10

number of Number ofloads SPS

Figure 8. PCC active powers, Pn and Pin , versus the number of SPS per phase for 3rd (a), 5th (b) and 7th (c) harmonics: __ PSB simulations, Model results.

Figure 10. Efficiency values of the supply network versus the number of SPS, obtained from simulation:_____ defininition (20) and ------- definition (21).

Fig. 9 reports the efficiency values of the SPS rectifier section versus the number of SPS as defined by formulas (11) and (12), respectively and obtained from the results of the simulations performed. It is evident the difference between and with the higher number of loads. Fig. 10 reports the efficiency values for the supply network versus the number of SPS obtained as for Fig. 9 and defined by the formulas:

VI.

CONCLUSIONS

S =

Pnh
1 PG

(20)

Analyses to have a quantitative idea of the real behavior of switching power supply units from the point of view of the power absorption are developed. Results of several experimental tests performed by means of a test system built up at Laboratories of the Second University of Naples (Italy) are firstly reported. Then, suitable models are proposed to easily predict the behavior of SPS sets when the number of sets varies and these models are validated by means of accurate time domain simulations. Experimental and numerical results are discussed. ACKNOWLEDGMENT The authors thank eng. Claudio Fulgeri for its valuable support during experimental activity. REFERENCES
[1] IEEE Working Group on Nonsinusoidal Situations, Practical Definitions for Powers in Systems with Nonsinusoidal Waveforms and Unbalanced Loads, IEEE Transactions on Power Delivery, Vol. 11, No. 1, Jan. 1996, pp.79101. Emanuel, A. E., Apparent Power Definitions for Three-Phase Systems, IEEE Transactions on Power Delivery, Vol. 14, No. 3, July 1999, pp. 76772. A. Ferrero, Definitions of electrical quantities commonly used in nonsinusoidal conditions, ETEP, Vol. 8, 1998 pp.235-40. IEEE Standard 1459-2000, Definitions for the measurement of electric power quantities under sinusoidal, nonsinusoidal, balanced or unbalanced conditions, IEEE PES, PSIM Committee, June 2000. L. S. Czarnecki, Comments on Active Power Flow and Energy Accounts in Electrical Systems with nonsinusoidal Waveforms and Asymmetry, IEEE Transaction on Power Delivery. Vol. 11 Issue: 3, Jul. 1996. R. Arseneau , B. Hughes, Selecting revenue meters for harmonic producing loads, IEEE International Conference on Harmonics and Quality of Power, Paper hpq 054, Lake Placid NY, Sept. 12-15, 2004. R. Langella, A. Testa A Circular Convolution based Algorithm for Electric Energy Measurement Under Non-Sinusoidal Conditions, IMTC 2007, Warsaw, Poland, May 1-3, 2007. M. Ferrucci, R. Langella, A. Testa, Sulla Definizione delle Potenze e sul Conteggio dellEnergia Elettrica in Regime Periodico Deformato, proc. of 101st Italian National Conference AEIT, Capri, September 2006 (in Italian). A. Mansoor, W.M. Grady, A.H. Chowdhury, M Samotyj, An Investigation of Harmonics Attenuation and Diversity Among Distributed Single-Phase Power Electronic Loads, IEEE Trans. on Power Delivery, volume 10, number 1, pp. 467-473, January 1995.

P1 . 'S = n 1 PG

(21)

being P1G the active power before the impedance Zh. The differences among the results based on the definitions (11) and (20), that consider the total power (3) to charge the customers, and those of definitions (12) and (21), respectively, that assume the power at fundamental frequency to charge the customers, are evident. The first choice introduces a discount for distorting customers and penalizes the network, that means the utility, and this can reflect on the prices of electrical energy for all customers.
1

[2]

[3] [4]

0.95 load efficiency SPS Rectifier efficiency

[5]

0.9

[6]
0.85

[7]
0.8

of (11) of (12)
1 2 3 4 5 6 7 8 9 10

[8]

0.75

number of loads Number of SPS

[9] Figure 9. Efficiency values of the SPS rectifier section versus the number of SPS, obtained from simulation:___ defininition (11) and ---- definition (12).

You might also like