Posted on 17 July 2019

Calculation Example: Ensuring EMC of Equipment Connected to 230V Lines

Varistor for electromagnetic compatibility








The following calculation example describes the selection procedure for a varistor in order to ensure that the electromagnetic compatibility (EMC) of a device is met in accordance with IEC 62019-4-5  for 230 V operating voltage and a test voltage of 4 kV.

Electromagnetic Compatibility test circuit

Figure 1. EMC test in accordance with IEC 62019-4-5 with Ri = 2Ω , charge voltage 4 kV on a 230 V AC line voltage

Line voltage: 230 VAC ±10%

Hybrid test generator: 4 kV, 2 ;Omega;

Number of repetitions: 10 (5 in each polarity)

Voltage endurance of equipment to be protected: 1 kV

Operating voltage

For European public AC power networks, IEC 60038 specifies that a line voltage tolerance of ±10% is to apply from the year 2003. In this selection example we shall assume this tolerance to apply. This means that a maximum required operating voltage of 253 VAC must be taken into account when selecting the varistor (the negative tolerance is of no significance for varistor selection).

To achieve the lowest possible protection level, the varistor types with the voltage class closest to 253 V must be chosen, i.e. in this case the voltage class “K275”.

Surge current

To determine surge current capability, the SIOV-S14K275 varistor will be taken as an example.

For a given surge voltage, the maximum current will always be observed if the varistor operates at the lower limit of the tolerance range.

For this reason the clamping voltage reduced by the tolerance band width must be inserted in


instead of the value reduced by the tolerance (K = ±10%).

The surge generator’s short circuit current would then be

i^*=\frac{4kV}{2 \Omega}=2kA

From the V/I characteristics of the SIOV-S14 varistor you derive a protection level of 1000 V for an S14K275. Reducing this by the tolerance (±10%) produces

V_{SIOV}=1000V\cdot \frac{0.9}{1.1}=820V\; .

By inserting this in


you obtain the surge current amplitude (worst case)

i^*=\frac{400V-820V}{2A}=1590A\; .

The selected varistors must be able to handle a surge current of this amplitude ten times consecutively, regardless of polarity.

In accordance with IEC 62019-4-5, the hybrid generator is designed to supply surge current waveforms of type 8/20 µs in the case of a short circuit. As the protection level of the varistor in this case is low in comparison to the no-load generator voltage, you can assume the 8/20 μs waveform to apply in this type of load as well. This waveform can be transformed into an equivalent rectangular wave with tr*= 20 μs, which is to be used.

Look up this type in the derating diagram to check whether or not an S14K275 can be subjected to the above surge current load. As a result of the investigation, a current of 1590 A (8/20 µs) is only permissible for two consecutive load cycles. For the required number of ten repetitions, the current imax would only be 1000 A.

Since i* > imax, S14K275 is not a suitable choice for the given application conditions. The type with the next highest surge current capability would be S14K275E2. The derating field yields imax (10 X) = 1500 A. For this reason, this type is not suitable either.

As a result the selection check procedure must be repeated for the type series having the next highest power dissipation capability. In this case the type in question is the varistor type S20K275:



results in

i^*=\frac{4000V-\big (950\cdot \frac{0.9}{1.1A}\big)}{2\Omega }=1610A\; .

For ten load repetitions (at tr* = tr = 20 μs) the derating field of the S20K275 shows

i_{max}=2500A\; .

With this value, the S20K275 meets the i* ≤ imax selection criterion.

Energy absorption

Since energy absorption, as calculated using

W=\int_{t0}^{t1} v(t)i(t)dt \; .

is directly correlated to surge current, the S20K275 also fulfils the selection criterion of W* ≤ Wmax.

Power dissipation

In order to determine power dissipation, you must calculate the energy absorbed by the S20K275 when conducting the surge current. According to

W^*=\hat{v}^*\cdot \hat{i}^*\cdot \hat{t}^*_r

we obtain

W^*=780V\cdot 1610A\cdot 20\times 10^{-6}s=25J \; .

As a pulse repetition rate, IEC 62019-4-5 specifies a maximum of one pulse per 60 s. Inserting this in

P^*=\frac{W^*}{T^*}=\frac{\hat{v}^*\cdot \hat{i}^*\cdot \hat{t}^*_r}{T^*}

results in

P^*=\frac{W^*}{T^*}\approx 0.4W \; .

From the product table, the maximum permissible periodic load, i.e. average maximum power dissipation of an S20K275, is found to be 1 W. With this the selection criterion of  P* ≤ Pmax is also met.

Protection level

The protection level is found to be 900 V (from the V/I characteristics for a value of 1610 A). In this case the 4 kV “overvoltage” is limited to 23%.

The protection level is lower than the voltage strength of the equipment to be protected, which is equal to 1000 V.

By fulfilling this final criterion, the Standard SIOV-S20K275 is found to meet all selection criteria and can thus be considered suitable for the application.

Comparison to PSpice

Selection of the varistors for table 3 was carried out using PSpice calculations. The results for S20K275 correlate well with the values calculated here.

Other suitable types of varistors

If the physical dimensions of the chosen component SIOV-S20K275 are too large, similar selection calculations show that the EnergetiQ varistor SIOV-Q14K275, which requires less headroom can also be used.


For more information, please read:

Selection Guide - Varistors

Design Notes - Varistors

Varistor Operation - Derating, Temperature, and Overload

V-I Characteristics - Varistors

Design-in Reccomendation: Application for 600 V in Accordance with UL

Calculation Example: Switching of Inductive Loads


VN:F [1.9.17_1161]
Rating: 0.0/6 (0 votes cast)

This post was written by:

- who has written 75 posts on PowerGuru - Power Electronics Information Portal.

Contact the author

2 Responses

  1. avatar SS says:

    Thank you, this is interesting. I think the timescales should be us (microseconds) not ms in the discussion of surge waveforms.

    VA:F [1.9.17_1161]
    Rating: 5.0/5 (1 vote cast)
  2. avatar PowerDahl says:

    Hello SS,

    I'm glad that you enjoyed the article. We always appreciate recieving feedback from our readers.

    You are absolutely correct. The timescales mentioned should be in microseconds. Thank you for pointing out our error. We made the necessary corrections.

    VN:F [1.9.17_1161]
    Rating: 0.0/5 (0 votes cast)

Leave a Response

You must be logged in to post a comment.