Key points for testing the trigger current of discrete semiconductors

Trigger Current Testing Essentials for Discrete Semiconductor Devices

You set the gate voltage, wait for the device to turn on, and nothing happens. The voltage climbs. The current stays at zero. Then suddenly, at some threshold you did not expect, the device fires. That threshold is the trigger current, and if you do not know it, your circuit does not work.

Trigger current — also called gate trigger current (IGT) for thyristors, or gate threshold current for MOSFETs and IGBTs — is the minimum current or voltage needed to switch a device from off to on. It sounds simple. It is not. The value shifts with temperature, with the rate of voltage rise, with the load current, and with the device's internal state. A test that ignores these variables gives you a number that is useless in the real world.

This article covers how to test trigger current properly, what makes the measurement drift, and why most test setups get it wrong.

What Trigger Current Actually Tells You

Trigger current is not just a switching threshold. It is a window into the device's internal gain and its sensitivity to noise.

For thyristors and SCRs, the gate trigger current (IGT) is the minimum gate current required to bring the device into conduction when the anode-cathode voltage is at its rated value. Below IGT, the device stays off no matter how long you wait. Above IGT, it latches on and stays on even after you remove the gate signal. This latching behavior is what makes thyristors useful in power control, but it also means that once you cross IGT, you cannot turn the device off through the gate. You have to commutate the anode current to zero.

For MOSFETs and IGBTs, the equivalent parameter is gate threshold voltage (Vgs(th)), which is the gate-source voltage at which the device just begins to conduct a small drain current, typically 250µA. Below Vgs(th), the device is off. Above it, the device turns on, and the drain current rises rapidly with further gate voltage increase.

The problem is that both IGT and Vgs(th) are not fixed numbers. They vary with temperature, with the anode-cathode voltage (for thyristors), and with the rate at which you apply the trigger signal. A test that measures IGT at 25°C with a slow voltage ramp gives a completely different result from a test at 125°C with a fast ramp. If your design uses the 25°C number, it will not trigger reliably at high temperature.

Testing Gate Trigger Current in Thyristors and SCRs

Thyristor trigger current testing is more involved than it appears. The device has three terminals, and the interaction between them during turn-on determines whether you get a clean trigger or a false one.

Setting Up the Anode-Cathode Bias Correctly

Before you apply any gate signal, the anode-cathode voltage must be at the correct level. IGT is specified at a particular Vak, usually the rated repetitive off-state voltage or a fraction of it. If you test at too low a Vak, the IGT will be higher than the datasheet value. If you test at too high a Vak, you risk damaging the device if the trigger fails.

Use a programmable DC supply for the anode-cathode bias. Set it to the specified Vak. Let it stabilize. Do not add series resistance unless the standard requires it. Series resistance drops voltage during turn-on and changes the effective Vak at the moment of triggering, which shifts the IGT reading.

For high-voltage thyristors rated above 600V, the off-state leakage current is small but not zero. At 125°C, this leakage can be several milliamps. If your gate current source is not stiff enough, the leakage current will shift the effective gate current and corrupt your IGT measurement. Use a gate current source with at least 10x the expected IGT value in compliance voltage headroom.

Applying the Gate Pulse and Measuring the Response

The gate pulse must be fast and clean. A slow rise time on the gate pulse allows the device to heat up during the trigger event, which lowers the IGT and gives you an optimistic reading. Use a pulse generator with a rise time of less than 1µs for most thyristors. For fast-recovery devices, aim for 100ns or better.

The pulse width must be long enough for the device to latch. A pulse that is too short will trigger the device, but the device will turn off again before it latches. You will measure a trigger event that does not actually result in conduction. The standard pulse width is typically 10µs to 100µs, depending on the device. Check the applicable standard for the exact value.

Measure the gate current at the moment the anode current rises above a defined threshold, typically 10mA or 100mA. That gate current is your IGT. Use a current probe on the gate lead, not a voltage measurement across a gate resistor. The voltage across a resistor includes the voltage drop from the gate-cathode junction, which varies with temperature and adds error to your current calculation.

Run the test at three temperatures minimum: 25°C, 85°C, and 125°C. IGT typically decreases with temperature for most thyristors. A device with IGT of 50mA at 25°C might have IGT of 20mA at 125°C. If your gate driver can only source 30mA, the device triggers at room temperature but fails at high temperature.

Testing Gate Threshold in MOSFETs and IGBTs

MOSFET and IGBT trigger testing is simpler in concept but tricky in execution. The parameter you care about is Vgs(th), the gate-source voltage at which the device just starts to conduct.

The Constant Current Method

The most common method forces a small constant drain current — typically 250µA — and measures the gate-source voltage at which that current flows. That voltage is Vgs(th).

Bias the drain-source voltage to a low value, typically 10V to 25V for MOSFETs, or to the rated Vces for IGBTs. The drain-source voltage affects Vgs(th) through the Miller effect and through drain-induced barrier lowering (DIBL). If you test at too low a Vds, your Vgs(th) will be higher than the value you would see in the actual circuit.

Use a current source on the drain, not a voltage source with a series resistor. A current source holds the drain current constant regardless of Vgs changes. A resistor lets the current drift as Vgs changes, which corrupts the threshold measurement.

The gate voltage must be ramped slowly. A fast ramp charges the gate capacitance and overshoots the actual threshold. The standard ramp rate is typically 0.1V/ms to 1V/ms. Slower is better for accuracy, but too slow lets leakage current affect the reading. Find the sweet spot and stay there.

The Constant Voltage Method and Why It Is Less Useful

Some engineers apply a fixed gate voltage and measure the resulting drain current. This tells you the on-state behavior, not the threshold. A device with Vgs(th) of 3V and a device with Vgs(th) of 5V can both conduct 1A at Vgs of 10V. The constant voltage method cannot tell them apart.

Use the constant current method for threshold characterization. Use the constant voltage method only for on-state resistance verification. Do not mix the two.

Temperature and Rate Effects on Trigger Current

Trigger current is not stable. It moves with every condition you change, and if you do not account for that movement, your design will fail at the worst possible time.

How Temperature Shifts the Trigger Point

For thyristors, IGT decreases as temperature rises. The gain of the internal transistor structure increases with temperature, so less gate current is needed to trigger conduction. This sounds good until you realize that your gate driver was sized for the high IGT at low temperature. At high temperature, the device triggers too easily, which can cause false triggering from noise.

For MOSFETs, Vgs(th) decreases with temperature, typically by 1mV/°C to 3mV/°C. A device with Vgs(th) of 4V at 25°C might have Vgs(th) of 3.2V at 125°C. That 0.8V shift is enough to turn on a device that should be off if your gate driver has poor noise immunity.

Test at the full temperature range your application sees. Do not test at 25°C only. The number you get is not the number you will use.

The dv/dt Trigger Problem in Thyristors

Here is the one that catches everyone off guard. A thyristor can turn on without any gate signal at all if the anode-cathode voltage rises too fast. This is called dv/dt triggering, and it is not a defect. It is physics.

When Vak rises rapidly, the displacement current through the internal capacitances acts like a gate current. If that displacement current exceeds the device's internal IGT, the thyristor fires on its own. No gate pulse needed. No control. Just sudden conduction.

The critical dv/dt rating tells you how fast you can rise the anode voltage without false triggering. If your circuit has a fast voltage transient — from a snubber, from inductive kickback, from a switching event — and that transient exceeds the critical dv/dt, the thyristor will fire when you did not want it to.

Test dv/dt triggering by applying a linearly rising voltage to the anode-cathode terminals with the gate open. Measure the voltage at which the device turns on. That is your dv/dt trigger voltage. Convert it to dv/dt by dividing by the ramp time. Compare it against the datasheet minimum. If your circuit has faster transients, you need a snubber network to slow the voltage rise.

Defining Pass/Fail Criteria and Sample Size

A trigger current test without clear criteria is just a data point. It does not tell you whether the device is good or bad.

Setting the Trigger Current Limits

The datasheet gives you a range, not a single number. IGT might be specified as 10mA to 50mA. For production testing, use the maximum value as your fail limit. Any device that requires more than 50mA to trigger fails.

For characterization, record the actual value and plot the distribution. You want to see whether the lot is centered in the range or drifting toward one edge. A lot that is centered at 45mA against a 50mA maximum has no margin. A lot centered at 15mA has plenty of margin.

For MOSFETs, use the maximum Vgs(th) from the datasheet as your fail limit. A device with Vgs(th) of 4.5V against a 4V maximum fails. Do not use the typical value. The typical value is for simulation, not for production.

How Many Devices Do You Need to Test

One device tells you nothing about trigger current variation. Five devices gives you a hint. Ten devices gives you a distribution you can trust.

For production, test five devices per lot minimum. For incoming inspection, test ten. For characterization, test at least twenty-five to build a meaningful statistical picture.

Record the temperature, the bias voltage, the pulse width, the ramp rate, and the raw waveform for every sample. Trigger current is sensitive to every one of these variables. If you do not record them, you cannot reproduce the test when a failure comes back.

Setup Errors That Ruin Trigger Current Measurements

The most common error is using a bench power supply for the gate current. Bench supplies have output impedance and current limit response times that are too slow for trigger testing. The current limit engages after the device has already triggered, and you measure the wrong value. Use a dedicated current source or a fast pulse generator with current monitoring.

The second error is ignoring lead inductance. Long gate leads add inductance that slows the current rise. The gate current reaches the device later than you think it does, and your trigger timing is off. Keep gate leads under 5cm. Use a direct connection to the gate pin, not a clip lead.

The third error is testing at the wrong anode-cathode voltage. IGT is specified at a particular Vak. If you test at half the rated voltage, your IGT will be 20% to 40% higher than the datasheet value. The device is not bad. Your test condition is wrong.

The fourth error is not preconditioning the device. Thyristors have a memory effect. If you test them immediately after a high-current pulse, the internal carrier distribution is disturbed, and the IGT reading shifts. Let the device rest for at least 10ms between tests. For MOSFETs, discharge the gate completely before each measurement. Residual gate charge changes the apparent Vgs(th).

Trigger current is the first thing your circuit sees when it tries to turn the device on. Get it wrong, and nothing else matters. The forward voltage, the leakage, the gain, the thermal resistance — all of it is irrelevant if the device will not trigger when you need it to. Test it at the right temperature, the right bias, the right speed, and the right current. Then trust the number.