This technical article will discuss the Electromagnetic Compatibility (EMC) IEC61000-3-3:2013 testing standard, including relevant extracts from IEC61000-4-15(instrumentation oriented standard) and explanations relating to analysis of voltage fluctuations and flicker on public low voltage supply systems. We will also describe how N4L ensure full compliance to the standard, including our Harmonics and Flicker Analyzer (PPA5511/PPA5531), AC Power Source and Impedance Network (Zref)
The formal definition of flicker is the “impression of unsteadiness of visual sensation induced by a light stimulus whose luminance or spectral distribution fluctuates with time” (IEC 60050-161,1990)
Flicker can be considered as a symptom, resulting from the modulation of a load and its effect on its own terminal voltage. Whilst connected to a voltage source with finite impedance, any load modulation will cause voltage fluctuations on the supply line.
IEC61000-3-3:2013 is primarily concerned with the limitation of such fluctuations and subsequent flicker upon public low-voltage systems. IEC61000-3-3:2013 specifies the limits of voltage changes produced by electrical and electronic equipment when tested under these specified conditions.
IEC61000-3-2:2013 is applicable to electrical and electronic equipment with an input current rating equal to or less than 16A per phase. Should a device fail the limits specified in IEC61000-3-3:2013 tested with the specified source impedance specified in section 6.4 (Zref), it may be retested to show conformity with IEC61000-3-11 (discussed in a separate article). Part 3-11 is applicable to all equipment with rated input ≤75A per phase and is subject to conditional connections.
Zref is the internationally agreed reference source impedance for low-voltage supply networks, as illustrated in the diagram below.
G – Voltage Source
EUT – Equipment Under Test
M – Measuring Equipment
S – Supply Source including reference impedance and voltage generator output impedance
RA = 0.24Ω jXA = 0.15Ω @ 50Hz
RN = 0.16Ω jXN = 0.10Ω @ 50Hz
Note: The reference impedance within the source (S) block includes both the output impedance of the AC Source (G) and the reference impedance network (RA, jXA, RN, jXN)
|Figure 1. (IEC61000-3-3:2013,2013)|
In real world compliant flicker testing, Zref is a physical impedance network consisting of a resistive element and an inductive element that is placed in between the AC Power Source and the equipment under test. N4L can provide the entire test system including Programmable AC Power source, Impedance Network and Flicker (+Harmonics +Power) Analyzer. We will discuss the different elements of the test system below;
1. Programmable AC Power Source – Section 6.3 of IEC61000-3-3:2013 describes the requirements for the AC Power Source for compliant Flicker Testing, this can be summarized in the table below.
|IEC61000-3-3:2013 Test Supply Voltage (AC Power Source) Requirements|
|Requirement||N4L N4Axx Programmable AC Power Source Performance|
|Open Circuit Voltage shall be the rated voltage of the equipment||0-300Vrms Single Phase, 520Vrms Three Phase|
|Test Voltage Maintained within ±2% of the nominal value||±0.1%|
|Frequency Stability ±0.25Hz||±0.01Hz|
|THD < 3%||Better than 0.3%|
2. Zref – Reference Source Impedance – Section 6.4 of IEC61000-3-3:2013 describes the requirements for Zref
IEC61000-3-3:2013 refers to IEC/TR 60725 which specifies the in-phase and quadrature component of the reference impedance Zref. It is important to note that Zref includes the impedance of the reference network and the output impedance of the AC source, the output impedance of the N4A series AC Power Sources is very low and can be considered negligible for this application.
3. Flicker Analyzer – Section 4.2.2 of IEC61000-3-3:2013 refers the reader to IEC61000-4-15:2010 which lays out the specification and signal processing requirements for compliant flicker measurements. All N4L PPA5511 and N4L PPA5531 Harmonics and Flicker Analyzers offer full compliance to IEC61000-4-15. Furthermore, N4L are able to offer IEC61000 Harmonics and Flicker calibration to ISO17025 from our in house UKAS Laboratory.
Commonly known as “d” values, voltage change characteristics constitute three separate parameters – d(t), dc, dmax and Tmax.
Pst, or “short term flicker severity” is the calculated short term flicker as per IEC61000-4-15, unless otherwise specified the Pst evaluation period is 10 minutes.
Long term flicker sensitivity (perceptibility long term) is calculated as follows;
Where Pst, i (i = 1,2,3, …) are consecutive readings of short term flicker severity Pst.
Should the device under test fail the limits specified in IEC61000-3-3 when tested with Zref, it is permitted for a manufacturer to specify that the equipment must be connected to a supply interface point which exhibits an impedance lower than that of Zref (the terminal/socket at which the device will be connected to the mains supply). This does make sense as basic electronic theory teaches us that voltage drop across any impedance (whether it is real or complex) is proportional to that impedance. If the source impedance at the connection point of the supply is lower (for the same load modulation) you will experience lower voltage fluctuations at the connection point, thus lower flicker readings.
Assessment of relative Voltage change:
The basis of flicker measurement centers around the voltage change characteristics when measured at the terminals of the equipment under test. N4L use the “Flickermeter” method for assessment of Pst. Other methods exist and are specified in section 4.2.1 of IEC61000-3-3, these methods are however accompanied by caveats such as that the nominal voltage (U(t)) must be known to use the simulation method.
It can be considered as the difference between the line and neutral Uhp values, remembering that Uhp is the half period RMS value between consecutive zero crossings.
Uhp(t1) is the half period Vrms at time = t1
Uhp(t2) is the half period Vrms at time = t2
ΔUhp(t) = Uhp(t1) – Uhp(t2)
This voltage change at the equipment under test’s terminals (or the difference between Uhp(t1) and Uhp(t2)), is a result of a change in voltage drop across the complex reference impedance (Zref) caused by a chang in the complex fundamental current IEC61000-3-3 describes the complex current as;
ΔIp = Active current change
ΔIq = Reactive current change
If the equipment under test exhibits an inductive load to the generating equipment, for single phase and symmetrical three phase equipment the voltage change can be approximated to;
ΔUhp = | ΔIpR + ΔIqX |
Where ΔIp and ΔIq are the in phase and quadrature parts of the current change respectively. R and X are the in-phase and quadrature elements of the complex impedance.
Relative Voltage Change
Relative voltage change is given by the following formula;
d = ΔUhp/Un
As we have already stated, dmax is the maximum absolute voltage change, therefore the dmax,i evaluation ends as soon as a new steady state voltage condition is established. Alternatively the evaluation of dmax,i will end when the measurement period finishes.
Positive dmax,i is recorded when the maximum voltage change is a result of a reduction in voltage since the previous dend,i
Negative dmax,i is recorded when the maximum voltage change is a result of an increase in voltage dend,i
The Pst = 1 curve is a well known concept, it is however rarely understood. The graph below (taken from IEC61000-3-3:2013) illustrates the overall concept, we must remember that if the performance of the equipment under tests lies under the Pst = 1 curve then the disturbance to the supply network is deemed to be acceptable in the short term (we will distinguish between short term and long term later in this article).
|Figure 2. (IEC 61000-3-3:2013,2013)|
If we analyze various working points on the Pst = 1 curve it is possible to determine the relative voltage change (d value) and the number of voltage changes per minute at that perturbation level.
For a 230V system a relative voltage change of 1% resulting in a Pst of 1 would require 25 voltage changes per minute. Figure 3 below is annotated at the 1% voltage change point and we can follow the graph across to the intersecting value on the y-axis at approximately 25 voltage changes per minute.
|Figure 3. (IEC 61000-3-3:2013,2013)|
We can state that for a voltage change of 1% at a frequency of 25 changes per minute, we would achieve a Pst equal to 1.
If we cross refer to table D1, found in Annex D of IEC61000-3-3 we discover that the findings above correlate to the test protocol table.
|Table 2. (IEC 61000-3-3:2013,2013)|
Long term flicker Plt shall be calculated with a value of N = 12 (see formula below)
The standard states that it is only necessary to assess Plt for devices that are to be operated for longer than 30 minutes at a time.
The limits specified in IEC61000-3-3:2013 are applicable to voltage fluctuations and flicker measured at the supply terminals of the device under test. This is an important point, as any current carrying component with measurable impedance, there will be a voltage drop across it. A current carrying cable connecting the AC Source to the equipment under test is no different and if the Flickermeter is placed at the output terminals it is important to remember that the overall impedance of the test circuit will include the interconnecting cabling. N4L have characterized this additional impedance and supply known value interconnecting cables with our IEC61000 test systems and compensated Zref networks accordingly.
|Voltage Fluctuation and Flicker Limits|
|Pst||Less than or equal to 1.0|
|Plt||Less than or equal to 0.65|
|Tmax||Accumulated time of d(t) with a deviation exceeding 3.3% during a single voltage change at the equipment under test terminals must not exceed 500ms|
|dc||The maximum relative steady state voltage change must not exceed 3.3%|
Maximum relative voltage change (between two half periods) shall not exceed:
4% without additional conditions
6% for equipment that is;
– switched manually
– switched automatically more than twice a day and is also fitted with a delayed restart not less than a few tens of seconds. Alternatively a manual restart after a power supply interruption.
7% for equipment that is;
– attended whilst in use
– switched on automatically or intended to be switched on automatically no more than twice per day. Must also be fitted with a delayed restart of not less than a few tens of seconds (or manual restart) after a power supply interruption.
Pst and Plt limits need not be applied to voltage changes caused by manual switching.
Nor do the limits need to be applied to voltage changes associated with emergency switching or emergency conditions, this is sensible as such occurrences will be very rare.
For voltage changes caused by manual switching, equipment is deemed to comply without further testing providing the maximum input current (including inrush current) evaluated over each 10ms window between zero crossings does not exceed 20A. Further to this, the supply current after inrush must be within 1.5A. If the maximum relative voltage change d(max) caused by manual switching is determined via measurement methods (such as the N4L PPA5511 and PPA5531) shall be measured as per the “Inrush Test Caused by Manual Switching” section below. The test procedure described is detailed in Annex B of IEC61000-3-3:2013.
The accuracy requirements are detailed in section 6.2 of IEC61000-3-3 and are tabulated below for your reference;
|Flickermeter Accuracy Requirements|
|Parameter||Limit||N4L PPA55x1 Performance|
|Current||±(1% + 10mA)||0.01% Reading + 0.038% Range|
|Total Measurement Uncertainty, including tolerance of Impedance Network and Impedance of AC Power Source||±8%||<1%|
The reference impedance is a conventional impedance and consists of a resistance and inductance in each phase. The values specified in IEC61000-3-3, an example of our reference impedance is shown below;
|Reference Impedance 50Hz (60Hz)|
The observation period is indicated in section 6.5 of IEC61000-3-3, the recommended period Tp for measurement of flicker is shown in the table below;
The standard recommends that the observation period should cover the part of the operation cycle that includes the most “unfavorable sequence of voltage changes”.
Assessing Pst requires the cycle of operation to be repeated continuously unless otherwise specified in Annex A of IEC61000-3-3. The minimum time it takes to restart the equipment, for equipment that automatically stops shall be included in the observation period if the cycle of operation is shorter than the observation period.
The assessment of Plt does not require the cycle of operation to be repeated unless otherwise stated in Annex A, when the cycle of operation is less than 2hrs and the equipment is not used continuously.
Cycle of operation: 35 minutes
Four consecutive Pst values are measured during an initial period of 40 minutes.
Remaining eight Pst values in the 2Hr test period are deemed to be zero.
The standard provides guidance on general test conditions for devices not mentioned in Annex A, advising that the equipment should be set up to produce the most unfavorable conditions with respect to voltage fluctuations at the input terminals of the equipment when connected to the power source and reference network.
The equipment should only use settings and programmable modes that are mentioned by the manufacturer in the instruction manual or that are otherwise likely to be used.
The underlying message of the standard is one that the equipment must be working in a similar manner to that which it would be in normal day to day use when in the field. The standard concentrates on simulation of the equipment when connected to a real power grid, therefore it makes sense to ensure that the equipment being tested is being used in a realistic manner.
Guidance for Motors
When a motor is first started, the back emf created by a rotating rotor across the static motor windings is zero, therefore when the rotor mechanical frequency is zero the supply current will be maximum. It is sensible to conclude that a locked rotor test (in which the rotors mechanical frequency is zero) could be used to determine the maximum inrush current. A secondary effect of maximum inrush current is a large dmax measurement, in fact the maximum dmax will occur during a locked rotor test and this approach is acceptable for dmax assessment.
Separately Controlled Circuits
For equipment consisting of separately controlled circuits, each system should be considered as an entirely separate item of equipment. This is providing that each sub circuit is not designed to be switched at exactly the same moment.
If several sub circuits are designed to be switched simultaneously then each group of circuits should be considered a separate item of equipment.
A system contains 14 sub circuits
Circuits 1,3,5 and 7 all switch simultaneously (System A)
Circuits 2, 4, 6 and 8 ~ 14 all switch simultaneously (System B)
Therefore, two “Items of equipment” are separately tested for the 14 sub circuits.
If a control system regulates only part of a load, the fluctuations in the supply voltage produced by each variable part should be separately considered.
Annex A provides mores specific product type related information regarding test conditions, a brief overview of the recommended test conditions are tabulated below;
|IEC61000-3-3 Annex A|
|Product Type||Type Test Conditions|
No requirement to test Plt for domestic cookers
Pst performed at steady state, unless specified
Tested using standard saucepans with water quantity specified in table A.1 of IEC61000-3-3:2013
Boiling – Perform 5 tests at boiling point and take the mean to calculate final result
Frying – Fill pan to 1.5 times quantity in table A.1, set temperature to 180 deg C, confirm temperature with thermocouple
Power Settings – Test all discrete power stages up to a maximum of 10. If non discrete stages are available separate range into 10 discrete steps.
Baking Ovens –
Test empty with door closed
Place thermocouple in center of oven
Set temperature to 220 deg C (conventional ovens), 200 deg C (hot air ovens)
Test with doors closed, unless otherwise stated
If settings are available – set to lowest, middle and highest settings and record worst result.
Baking Oven/Grill Combinations –
Test with door closed
Place thermocouple in the center of the oven
Set temperature to 250 deg C (or closest available temperature to this value)
Microwave Ovens –
Use a 1000g glass bowl of water as a load
Test at low, medium and third power stage (less than or equal to 90% maximum power)
Worst result recorded
Use lamp rated at power rating of equipment
Plt and Pst evaluation only required for equipment likely to produce flicker, such as disco equipment
No limits apply to individual lamps, such as fluorescent tubes
Incandescent Lamps (≤1000W), Discharge Lamps (≤600W), LED Lamps (≤200W) are deemed to pass dmax without the requirement for testing.
Luminaries with higher ratings that do not comply to the limits specified in IEC61000-3-3 shall be subject to conditional connection in accordance with IEC61000-3-11
Complete laundry program, normal wash cycle, 60deg C cotton without pre wash program if available
Rated load of double hemmed press washed cotton cloths (70x70cm, 140g/m2 ~ 175g/m2)
Water temperature 65degC (heater element), 15degC (other)
Heating elements not controlled by programmer – preheat to 65degC
Heating elements with no programmer – preheat to 90degC or lower if steady state is reached
Ignore simultaneous switching of motor and heater for assessment of dc, dmax and Tmax
Evaluate Pst and Plt
Fill with textiles, 50% dry load double hemmed press washed cotton cloths (70x70cm, 140g/m2 ~ 175g/m2)
Soak with 25degC water, adding 60% of original mass
Where possible, perform test at maximum and minimum heat settings
Evaluate Pst and Plt
Operate continuously with the door closed
Adjust thermostat to the mid value
Empty the cabinet, no heat required
Do not evaluate Pst and Plt
|Copying Machines, Laser Printers and Similar|
Test for Pst at the maximum rate of copying
Copy medium is white blank paper, 80g/m2 if not otherwise stated
Evaluate Plt in standby mode
|Vacuum Cleaners||Do not evaluate Pst and Plt|
|Food Mixers||Do not evaluate Pst and Plt|
Do not evaluate Plt
For portable tools without heating elements fitted do not evaluate Pst
Switch on for 10 minutes
For hand held dryers Plt shall not be evaluated
Evaluate Plt by switching on for 10 minutes
If power ranges are available, test complete range up to a maximum of 20 stages (for discrete steps)
|Television, Audio, Computer, DVD and similar equipment||Refer to Annex A, if no special test conditions are appropriate, only compliance to dmax is required.|
The table is not exhaustive and reference to Annex A in IEC61000-3-3 is recommended, further test conditions recommendations are made for other equipment such as water heaters, audio-frequency amplifiers, air conditioners, heat pumps, commercial refrigeration equipment and arc welding equipment.
Annex B describes test conditions for measuring dmax, also known as voltage changes at the terminals of the equipment under test caused by manual switching.
When an item of equipment is manually switched on or off, current inrush to the EUT causes voltage changes over the supply network. A statistical method is used in order to achieve a repeatable measurement of dmax.
The procedure includes 24 separate measurements, described below;
The highest and lowest recorded dmax values are deleted and the arithmetical average of the remaining 22 values is taken.
Annex C describes determination of steady state voltage and voltage change characteristics.
This section of the standard is a modified version extracted from IEC61000-4-15:2010, we should remember that the two standards would not exist without the other. IEC61000-4-15 sets out requirements of the measurement system whereas IEC61000-3-3 describes the pass/fail limits and practical implementation of a flicker test.
There are two distinct conditions that should be considered, one is where the voltage half period RMS remains in a steady state, the other is a period of “change”. This change can be a rise or fall of the half period RMS magnitude.
In order for us to understand the terminology and descriptions within annex C, it is important for us to define the parameters related to steady state and changing voltages.
Half period RMS Voltage (Uhp) –
This is the half period RMS (root mean square) value of the test waveform, taken from 0~180º and 180~360º, you will notice that the “half wave” is windowed with respect to zero crossing.
|Half Period RMS as specified in IEC61000-3-3 (Uhp)|
Figure 5 illustrates the data points used (the PPA5511/PPA5531 sample rate is much faster, at 2Ms/s) to calculate the half period Voltage RMS between 0~180º, this process is then repeated for the second half of the cycle (180~360º). It is clear that Uhp becomes a function of time which leads us on to our second definition.
Half period RMS Voltage characteristics Uhp(t)
The behavior of successive half period Vrms values is a function of time, therefore Uhp is used.
Relative Half period RMS Voltage dhp
Expressed as a ratio in comparison to to nominal voltage, this compares the most recently measured Uhp value to the nominal voltage (Un) as follows.
dhp(t) = Uhp(t) / Un
Steady State Voltage Change dc,i
Value of the difference between two consecutive steady state voltage changes, normally expressed as a percentage of the nominal voltage Un.
dc,i = [dend,i-1 – dstart,i]/Un
dc Measurement Example:
|Steady State Voltage Change Example|
Figure 6 illustrates the behavior of the half period RMS over time when a “steady state change” occurs as per the IEC61000-3-3 standard.
dc,i = [dend,i-1 – dstart,i]/Un
dc,i = [230-227.7]/230 = 0.01 = 1%
This is an example of a steady state voltage calculation, you will notice that the next steady state occurs after 1 second at the 227.7Vrms half period value.
Maximum Voltage Change dmax,i
This is the maximum recorded voltage difference between a steady state condition and a following half period Vrms ( dhp(t)) during the test period.
dmax,i = max [dend,i-1 – dhp(t)]
Polarity is indicated, the direction of polarity is dependent upon the direction of the voltage change in comparison to the dend,i-1 value.
The dmax value is a result of a reduction in voltage then the value of dmax is positive.
The dmax value is a result of an increase in voltage then the value of dmax is negative.
Maximum steady state Voltage change during observation period dc
dc is the highest absolute value of all dc,i values recorded during a test.
dc = max i [|dc,i|]
Maximum absolute Voltage change during test period dmax
Highest absolute value of all dmax,i values during the test period.
dmax = max i [|dmax,i|]
Voltage deviation d(t)
d(t) is the deviation of the current half period Vrms(dhp(t)) to the previous dend,i-1 inside a voltage change characteristic. Polarity is optional for this parameter and a voltage drop is considered positive.
d(t) = dend,i-1 – dhp(t)
Steady State Voltage Change and Voltage Change Characteristics
A steady state condition is defined as when the half period r.m.s. voltage Uhp remains within a tolerance band of ±0.2% for a minimum of 100 half cycles of the fundamental frequency (50Hz), this clearly equates to a steady state for 2 seconds.
At the beginning of a flicker test, the initial reference voltage is defined to be the average r.m.s. voltage acquired during the preceding second to the commencement of the test period. This is the starting reference value for dc and dhp(t) calculations and well as for the calculation of dmax and d(t) measurements.
If a steady state condition cannot be established during the test, dc shall be reported as zero.
As the test progresses, an average Uhp,avg is calculated from the last 100 (50Hz) Uhp values. The Uhp,avg value is used to determine whether or not the steady state condition continues, it is also used as the reference for dc, dmax and Tmax in the event voltage changes occur.
What this essentially means is that a 100 sample average, Uhp,avg of the half period r.m.s. voltage is continually taken, this average value is then compared to the last half period r.m.s. voltage Uhp. If the last half period r.m.s. voltage is within 0.2% of the Uhp,avg value then the steady state condition continues.
A new steady state condition dc,i is determined after a voltage change has occurred by calculating an initial value – dstart,i = dhp(t=tstart) is used. A tolerance value of 0.2% is set and a steady state condition is detected if 100 half period r.m.s. values are within the 0.2% band.
The reason Uhp,avg is used is to prevent slowly changing line voltages triggering a dc or dmax indication.
The last value of Uhp that lies within the 0.2% tolerance is denoted as dend,i. The new half period r.m.s. value following this final value within the 0.2% limit is denoted as dhp. This value is used as the starting value for determining the value of the next steady state condition, known as dc,i+1.
Once a new Uhp,avg is present (after 100 half cycles), a new steady state condition is measurement is taken. If any value falls outside the tolerance band for the 100 half periods, then this new value is used as the starting point for the steady state condition, until a period of 100 consecutive half cycles fall within the tolerance band of 0.2%.
N4L’s flicker meter verification tests are performed as per “Determination of steady state voltage and voltage change characteristics, as defined in IEC61000-4-15:2010”, specifically referring to Annex C.
Both reference voltage change characteristic patterns are generated which are described in table C.1 and C.2 of IEC61000-3-3:2013.
N4L have developed an AC Power source capable of generating the required Vrms change characteristics to simulate the waveforms described in C.1 and C.2, other waveforms can be generated as per customer requirements with the generator of the N4A06 6KVA AC Power source.
|Table C.1 – Test Specification for dc – dmax – td(t)|
The waveform is generated using “SEQUENCE” mode in the N4A06 AC Power Source, utilizing highly accurate timing algorithms.
The waveform is then measured with the PPA5511 flicker meter, with “normal” d(t) measurement settings.
|PPA5511 Verification : td(t) = 500ms|
The waveform in table C.2 is then generated and verified, a td(t) value of 600ms is expected.Note that td(t) is displayed as Tmax on the PPA5511.
|Table C.2 – Test Specification for dc – dmax – td(t)|
|PPA5511 Verification : td(t) = 600ms|
It should be noted that in figures 8 and 10 the values for d(t), dc and Tmax are all well within the accuracy limits of the IEC61000 standards.
Flicker analysis is a complex subject, procedures must be in place in order to perform thorough verification of the analyzer. Newtons4th are an accredited ISO17025 Calibration Laboratory (UKAS) and our scope includes testing of Harmonics and Flicker to IEC61000-3-3 (IEC61000-4-7), ISO17025 calibration is available directly from Newtons4th.