Frequently Asked Questions

Frequently Asked Questions

Using and Dimensioning Drivers

Can the gate charge specified in the data sheet of the power semiconductors be used to dimension the driver?

The driver performance is generally calculated from:
P = fsw * Qc * dVge
fsw: Switching frequency
Qc: Gate charge
dVge: Gate voltage swing

The gate charge depends strongly on the gate voltage swing. So a check must be made to see for which gate voltage swing it applies. Typically, SCALE drivers have a gate voltage swing of 30V (from –15V to +15V) and SCALE-2 drivers of 25V (from –10V to +15V) or 23V (from -8V to +15V).

If the gate charge is specified in the data sheet for a different gate voltage swing (e.g. 0V...+15V Mitsubishi, -15V/+15V Infineon, -5V/+15V Semikron), the actual gate charge can be determined from a gate charge curve (if given in the data sheet) or it can be measured with a digital oscilloscope by integration of the actual gate current.
Please refer also to Application Note AN-1001.

Do Power Integrations drivers offer short-circuit protection for the IGBT/MOSFET module? Do Power Integrations drivers offer Vce / desaturation detection?

All Power Integrations drivers contain a Vce detector that measures the voltage across the IGBT or MOSFET after turn-on of the power component. If the power component is subject to excessive current (which is significantly higher than twice the nominal current of the power switch) or a short circuit, the collector-emitter/drain-source voltage rises in response. This voltage rise as a function of the current is detected by the Vce detector, which then effectively protects the power component.

However, Vce detection is unable to protect the load reliably from short circuits and overcurrents. This is because an overcurrent level that is still permissible for the IGBT can already be critical for the connected motor or similar load. Moreover, a very strong rise of the IGBT collector-emitter voltage occurs only as a result of the desaturation effect. If the current remains below the desaturation threshold, no true short circuit occurs.

As a rule, the Vce detection should be considered as protection for the power component, and the load protection in the application should be assured via dedicated load-current measurement and regulation or other suitable measures. Please refer also to Application Note AN-1101.

Are Power Integrations drivers overload proof?

Most Power Integrations drivers do not limit the input current respectively the output power of the driver module. This means that they are not permanently protected from short circuits at their outputs. The advantage of this design is that the drivers may be briefly overloaded. The corresponding data sheets show the relevant limitations of this mode.

A typical application case would be burst operation (see also FAQ on this topic) when high-frequency switching phases alternate with suitable pauses. Another case is represented by pulsed-power topologies, which also require very high peak loads although the average power drawn is significantly lower.

The advantages of these special operating modes have led to us at Power Integrations to dispense with a general hard short-circuit protection of the drivers and to configure the input load limit in a flexible way.

Thermal measurements show that the temperature of certain driver parts exceeds 85°C (e.g. 95°C). Is this permissible or is the driver thermally overloaded?

As a rule, Power Integrations drivers are designed for maximum ambient temperatures of +85°C (see corresponding data sheets). The driver components then heat up above the ambient temperature due to the absorbed energy. So it is quite normal for certain parts to become warmer than the ambient temperature. The driver is then not necessarily overloaded.

Are short-term high-frequency bursts above the maximum switching frequency specified in the data sheet of plug-and-play drivers permissible?

The maximum continuous or mean switching frequency of gate driver cores and plug-and-play drivers is in the first place determined by the resulting component temperature due to the mean load of the components. In view of the transient thermal impedance of the components, higher switching frequencies are briefly permissible.

Due to the large thermal time constants, it can be assumed that the mean power dissipation can be fully exploited for a burst duration of less than a millisecond and a duty cycle of up to 0.1, i.e. bursts with about 10 times the mean switching frequency are permissible.

If this means that the absolute maximum ratings specified in the data sheet are exceeded, please contact the Power Integrations support team for further clarification.