How an EMI Filter Affects Your Product's Performance

INTRODUCTION

This article covers how the insertion of an EMI filter affects product performance. In particular, this article addresses other attributes to examine in EMI filter design above and beyond attenuation. Examples include temperature effects, overload, distortion, harmonics, voltage drop, and safety.

RATED VOLTAGE

The rated voltage of the filtered device is a continuous volt- age that establishes the voltage the EMI filter capacitors must withstand without breakdown or malfunction. Proper capacitor derating guidelines should be followed to ensure the filter is reliable when operated continuously at its rated voltage.

RATED CURRENT

Maximum current drawn and the maximum ambient temperature of the filtered device establish the rated current for the filter. In particular, the performance and magnetic properties of chokes are highly dependent on the core material. Impedance diminishes as current and temperature increase, and the choke saturates.

RATED FREQUENCY

The behavior of the filter capacitors help influence the rated operating frequency of EMI filters.

IN-RUSH CURRENT

In-rush current occurs when the filtered device is first turned on. At initial turn-on, peak current can be much larger than normal steady state current. Depending on the magnitude of this peak and its duration, filter components may overheat, breakdown or malfunction. Off-the-shelf filters are tested per MIL-F-15733 at 140% rated current, at rated frequency, for 15 minutes to verify that they will survive current overload conditions.

VOLTAGE DROP

Both AC and DC voltage drop tests should be conducted to ensure that the voltage dropped across the impedances of the filter does not negatively impact the voltage seen at the load.

VOLTAGE OVERSHOOTS AND SPIKES

EMI filters contain inductance in the chokes. Fast switching of dV/dt circuits results in very high amplitude overshoots and spikes that cause stress to the insulation of the windings. This phenomenon could greatly impact the reliability of the filter.

Figure 1 – Generic power-line filter

HARMONIC DISTORTION AND INSTABILITY

The issue of harmonic distortion and instability pertains to filters incorporated within switching power supplies. The negative input impedance of a switching power supply can oscillate in conjunction with the output impedance of the fil- ter. The problem is avoided by utilization of the Middlebrook Criteria, which dictates that the output impedance of the filter must be less than the reflected load impedance of the switch- ing power supply.

Oscillation can occur if two or more filters follow each other in series. If the filters are high Q (> 2) they can detune each other. This phenomenon shifts the cut-off frequency into the bandpass region which reduces the input voltage to a point where the filtered device may no longer function. Total ca- pacitance of the cascaded could also result in higher line and harmonic currents which adds heat to the filter.

VOLTAGE REGULATION, CLAMPING OR SMOOTHING

EMI filters should not be relied on to provide voltage regula- tion, clamping or smoothing. Excessive distortion of the pow- er line frequency is minimized by keeping the value of induc- tive reactance small. Capacitors used to suppress differential mode (DM) noise should have reactance that is no less than 100 times the filtered device’s input impedance.

SAFETY REQUIREMENTS

Safety requirements applicable to the filtered device must be taken into consideration when designing an EMI filter. The most prominent requirements are leakage current, component ratings, dielectric withstand, insulation resistance and temperature rise. Complete filters purchased as off-the-shelf items should already have mandatory safety agency approvals, however product safety testing on the end product is still required.

Leakage Current: The allowable earth leakage current limit of the filtered device (entire device) is driven by the applicable safety standard, such as IEC 61010-1 for measuring equipment or IEC 60601-1 for medical devices. The highest limit allowed is 3.5mA and the lowest is 0.1mA for medical devices if contact with a patient is involved. These limits dictate the values of the line-to-earth connected common-mode capacitors (Y-caps). Leakage current is calculated as:

Leakage Current = V * 2�F * C * 1.2 μA

Where:

• V is maximum rated mains supply voltage;
• F is 50 or 60Hz;
• C is in µF;
• 2 represents a 20% capacitor tolerance

If common-mode emissions are a problem and leakage current requirements prevent adding more Y-cap capacitance, about the only other solution is the addition of more common-mode (CM) choke inductance. This can be done either by increasing the value of the CM-choke or by adding a second CM-choke to the EMI filter design.

Capacitor Ratings: Safety standards such as IEC 60384-14 describe the rating requirements for X-caps (line-to-neutral, differential-mode) and Y-caps. Both types carry mains volt- ages continuously and must be specially rated to do so. A fire hazard is created with the failure of any X-cap, whereas the failure of a Y-cap results in a fire hazard and potential shock hazard.

There are different classes of X- and Y-caps depending on application for X-caps and the type of insulation connected across for Y-caps.

X1-caps are high pulse rated and are specified to withstand by 50µs (500W source impedance) impulses between 2.5kV and 4kV. X2-caps are for general purpose uses and therefore a little less robust than the X1-types. X2-caps are pulse rated to only 2.5kV. X3-caps are also of the general use type and pulse rated to a maximum of 1.2kV. 

Classification of Class X Capacitors

Subclass	Peak Impulse Voltage in Service	IEC 60664-1 Installation Category	Application	Peak Impulse Voltage Up applied before voltage test
X1	> 2.5 kV ≤ 4.0 kV	III	High pulse application	When CR ≤ 1.0 µF UP = 4 kV When CR > 1.0 µF UP =  kV
X2	≤ 2.5 kV	II	General purpose	When CR ≤ 1.0 µF UP = 2.5 kV When CR > 1.0 µF UP =  kV
X3	≤ 1.2 kV	-	General purpose	None
NOTE: The factor used for the reduction of UP for capacitance values above 1.0 µF maintains ½ x CRUP constant for these capacitance 2 values; CR is in F.

Classification of Class Y Capacitors

Subclass	Type of Insulation Bridged	Range of Rated Voltages	Peak Impulse Voltage before Endurance Test
Y1	Double insulation or reinforced insulation	≤ 500 V	8.0 kV
Y2	Basic insulation of supplementary insulation	≥ 150 V ≤ 300 V	5.0 kV
Y3	Basic insulation of supplementary insulation	≥ 150 V ≤ 250 V	None
Y4	Basic insulation of supplementary insulation	≤ 150 V	2.5 kV
NOTE 1: For definitions of basic, supplementary, double and reinforced insulation, see IEC 61140. NOTE 2: Y2 capacitors may be substituted for Y1 capacitors of the same or higher UR.

Y1-caps are for used where double or reinforced insulation is required. These capacitors have rated voltages ≤ 500V, with pulse ratings of 8kV.

Y2-, Y3-, and Y4-caps have rated voltages of < 150V (Y4-caps) to ≥ 150V and ≤ 250V (Y2 and Y3-caps). These capacitors are used where only basic or supplementary insulation is re- quired. Y2-caps have a 5kV pulse rating, whereas Y4-caps are rated 2.5kV. Y3-caps have no pulse rating.

Both X- and Y-caps are endurance tested for 1000 hours. X-caps at tested at 1.25 times the rated voltage with a 1 kV AC dielectric withstand tested performed for 0.1 seconds each hour. Y-caps tested at 1.7 times rated voltage with a 1 kV AC dielectric withstand test performed for 0.1 seconds each hour.

Endurance for Class X Capacitors: The capacitors shall be submitted to an endurance test of 1,000 h at upper category temperature at a voltage of 1,25 UR except that once every hour the voltage shall be increased to 1,000 Vrms for 0.1 s. Each of these voltages shall be applied to each capacitor individually through a resistor of 47 Ω ± 5 %.

Endurance for Class Y Capacitors: The capacitors shall be submitted to an endurance test of 1,000 h at upper category temperature at a voltage of 1.7 UR, except that once every hour the voltage shall be increased to 1,000 Vrms for 0.1 s. Each of these voltages shall be applied to each capacitor in- dividually through a resistor of 47 Ω ± 5 %.

A common mistake in the application of X- or Y-caps in EMI filter design is not selecting the proper class for the applica- tion, like using an X2 when an X1 is required, or using a Y2- cap to bridge double or reinforced insulation where a Y1-cap is required.

Temperature Rise: Temperature rise of an EMI filter falls under the category of both safety and reliability. During safety agency approval testing, both the filter and end product are tested at maximum worst case loading conditions as expected in service.

Applied voltage is set at rated voltage plus 10%. For example, a filter specified with a rated input voltage of 240 VAC is tested at 264 VAC. The end product is configured to produce its maximum load. It’s good practice to apply 20% margin on top of maximum load to ensure design robustness.

Thermocouples are placed on components and their temperature is determined once the temperature has stabilized. If any measured temperature exceeds component specifications, they should be replaced with those that have the proper rating. Running a filter too hot will reduce its expected life.

Dielectric Withstand: There are two different types of dielectric withstand tests (also known as hi-pot, short for high potential). Both tests are performed on the end-product with the filter installed as intended for normal use.

The first test is performed as a type test on prototypes during product development and safety agency evaluation at various stages (pre and post) of tests. In some cases, this test is performed after the end product has been exposed to 48 hours (21 days) of humidity preconditioning at 93% (± 3) relative humidity at 40°C (± 2°C). A high potential is applied to the line and neutral inputs tied together on one side and chassis ground on the other. A table is given below for the various voltages.

Voltage Proof (Filter Connected to Mains)

Type of Insulation	Range of Rated Voltages Line/Line (Test A) or Line/Ground (Test B or C)	Test A	Test B or C
			V a.c.	V d.c.
Basic	< 150 V	4.3 U 1 d.c. R	900	1260
Basic	≥ 150 V ≤ 300 V		1500	2250
Basic	> 300 V ≤ 1000 V		2UR + 1000	2.8 UR + 1400
Double or reinforced	≥ 150 V ≤ 300 V		3000	4500
Double or reinforced	> 300 V ≤ 1000 V		2(2UR + 1000)	2(2.8 UR + 1400)
1NOTE: UR is the rated a.c. voltage, but the test is V d.c. (EX.: UR = 300Va.c.: Utest = 1290 V d.c. All a.c. test voltages are a.c. and 50 Hz or 60 Hz, unless otherwise specified in the detail specification.

The test is performed for a period of one minute at power line frequency or at DC at levels of 2 to 4kV depending on requirements found in the end-product standard. The test is successful if no breakdown or flashover occurs. Hi-pot test generators have an upper current limit that will be reached if there is a problem with the filter’s insulation. This test is highly stressful on the components of the EMI filter including any capacitors tied to ground. Repeated testing will degrade filter components.

The second hi-pot test that is performed is one that is applied to 100% of production units. Because this test stresses the insulation, a reduced test duration of 1 to 2 seconds is allowed by most standards, but at a slightly higher (~10%) stress level. The test is successful if no breakdown or flash- over occurs.

Insulation Resistance: Insulation resistance (IR) is a type of quality control check similar to a hi-pot test. The test is carried out without discharge resistors in the circuit. An applied voltage of usually 500 VDC (table is given below) is used to determine the resistance of filter components including wiring, inductors and capacitors. Low IR indicates a problem that will only get worse over time. Careful filter component layout and selection will ensure low values of IR are avoided.

DC Voltage for Insulation Resistance

Voltage Rating of the Filter	Measuring Voltage
UR < 10 V	UR ± 10%
10 V ≤ UR < 100 V	(10 ± 1) V
100 V ≤ UR < 500 V	(100 ± 15) V
500 V ≤ UR < 1000 V	(500 ± 50) V
1000 V ≤ UR < 1500 V	(1000 ± 100) V
When it can be demonstrated that the voltage has no influence on the measuring result, or that a known relationship exists, measurement can be performed at voltages up to the rated voltage (10 V shall be used in case of dispute).

CONCLUDING REMARKS

Additional guidelines to examine while designing an EMI filter include:

• Climatic classifications and environmental aspects including temperature and humidity extremes, in addition to the pollution degree expected of the installed environment
• The variety of capacitors used
• Reliability, mean-time-between failure (MTBF)
• Transportation and installation requirements for vibration, shock, bump, and seismic environments

Entire articles could be devoted to each one of these topics. In the meantime, if you have any questions or need help designing your own EMI filter please do not hesitate to contact us. Our contact information is provided below:

REFERENCES

1. MIL-PRF-15733J, Performance Specification: Filters and Capacitors, Radio Frequency Interference, General Specification for (06-Jul-2010).
2. Ozenbaugh, R.L., Pullen, T.M., EMI Filter Design, Third Edition. CRC Press,
3. In Compliance Magazine. (2018, November 12). What Every Electronics Engineer Needs to Know About: Filters. Retrieved from https://incompliancemag.com/article/filters/
4. In Compliance (2019, November 12). Let’s Talk About Why Filters Fail. Retrieved from https:// incompliancemag.com/ article/lets-talk-about-why-filters-fail/
5. Interference Technology. (2019, June 17). Basics of Passive Filters for EMC Compliance. Retrieved from https://interferencetechnology. com/basics-of-passive-filters-for-emc-compliance/
6. Williams, , EMC for Product Designers, Fifth Edition. Newnes, 2017.
7. UL 60939-3, Standard for Passive Filter Units for Electromagnetic Interference Suppression - Part 3: Passive Filter Units for Which Safety Tests are Appropriate (22-Jul- 2016).