Key points for reverse voltage testing of discrete components

Reverse Voltage Withstand Testing for Discrete Components: What Actually Matters

Most people think reverse voltage testing is just "apply voltage, check for breakdown, done." It is not. The way you ramp the voltage, how long you hold it, what current you limit it to, and even the temperature of the part during the test all change the result. Get any of those wrong and you either pass a bad part or fail a good one. This is especially critical for diodes, transistors, and MOSFETs where reverse breakdown is a hard failure mode.

What Reverse Withstand Testing Is Really Measuring

When you apply reverse bias to a discrete semiconductor, you are stressing the junction. Ideally, almost no current flows until you hit the breakdown voltage, where the current spikes catastrophically. But in reality, there is always some leakage current even below breakdown. The test is not just about finding that breakdown point — it is about confirming the device can handle its rated reverse voltage without excessive leakage, without premature breakdown, and without damage that will show up later in the field.

The datasheet gives you a maximum reverse voltage rating, but it also gives you a maximum reverse leakage current at a specified reverse voltage and temperature. That leakage spec is just as important as the breakdown voltage. A part that breaks down at the right voltage but leaks ten times the specified current at half that voltage is not a good part, even though it "passed" a naive breakdown test.

Setting Up the Test Circuit Correctly

Current Limiting Is Non-Negotiable

The single biggest mistake in reverse withstand testing is using a power supply without current limiting. When a device breaks down, the current can go from nanoamps to amps in nanoseconds. Without a current limit, you destroy the part and possibly the supply. Set the current limit to a value well below the device's maximum rated reverse current — typically 1mA to 10mA for small signal diodes, higher for power devices but still bounded.

A series resistor works in a pinch for low-voltage tests, but it is not precise. The voltage drop across the resistor changes as current changes, so you do not actually know what voltage is across the device. A proper current-limited source gives you a flat current ceiling and lets the voltage rise naturally until breakdown, which is exactly what you want.

Ramp Rate Changes the Outcome

How fast you increase the reverse voltage matters more than people think. A fast ramp can cause avalanche breakdown to occur at a higher voltage than a slow ramp because the junction does not have time to heat up and trigger thermal runaway. This means a part might pass a fast ramp test but fail in the field where voltage is applied continuously.

For qualification testing, use a ramp rate of 1V per second or slower. For production screening, 5V per second is acceptable if you have correlated it against the slower ramp data. Never use a step function — applying the full voltage instantly is the fastest way to destroy a marginal part and get a false pass rate.

Temperature Is the Silent Variable

Test at the Worst-Case Temperature

Reverse leakage current roughly doubles for every 10°C increase in junction temperature. A diode that leaks 1uA at 25°C might leak 16uA at 85°C. If you test at room temperature and the part operates in a hot environment, your test tells you nothing about real-world performance.

For any meaningful reverse withstand test, you need to test at the maximum rated junction temperature. This usually means mounting the device on a temperature-controlled stage or chuck and letting it soak at temperature before applying voltage. If you cannot do that, at minimum note the ambient temperature and apply a derating factor to your leakage readings.

Thermal Runaway During the Test Itself

Here is a trap that catches even experienced engineers. As you increase reverse voltage, the leakage current increases. That leakage current generates heat. The heat increases the leakage current further. This positive feedback loop is thermal runaway, and it can destroy a good part during the test itself — not because the part is bad, but because you held the voltage too long.

The fix is to use a pulse test instead of a DC hold. Apply the reverse voltage for 100 milliseconds, measure the leakage, then remove the voltage and let the part cool. Repeat. This gives you the same information as a DC test without the self-heating distortion. Most curve tracers and SMUs support pulsed reverse voltage mode.

How to Read the Results Without Getting Fooled

Leakage Current Is the Real Pass/Fail Criterion

Breakdown voltage is easy to measure — you watch for the current spike. But leakage current at rated voltage is what determines long-term reliability. A part that breaks down at 100V but leaks 50uA at 50V when the spec says 5uA max is a field return waiting to happen.

Set your pass/fail threshold based on the datasheet leakage spec at the test temperature. Do not use a generic "less than 1mA is fine" rule. That rule works for power rectifiers but will let marginal small-signal diodes through.

Soft Breakdown vs. Hard Breakdown

Not all breakdown is the same. A hard breakdown is a sharp knee where current goes from near-zero to high current almost instantly. This is normal for Zener diodes operated in their breakdown region. A soft breakdown is a gradual increase in current with no sharp knee — this indicates junction damage or contamination.

If you see soft breakdown during a reverse withstand test, reject the part even if the voltage is above the rated maximum. Soft breakdown means the junction is already degraded and the part will fail prematurely under normal operating conditions.

Test Duration and Repetition

How Long Should You Hold the Voltage

For qualification, hold the reverse voltage at the rated maximum for at least one second. For production, 100 milliseconds is usually enough if you are using pulsed mode. Holding for longer than necessary does not give you more information — it just heats the part and risks inducing thermal runaway.

Do not cycle the voltage on and off rapidly during a single test. Each cycle injects charge into the junction and can shift the apparent breakdown voltage. Apply once, measure, remove. If you need to verify, wait 30 seconds between applications to let the junction discharge.

Repeated Testing Degrades the Part

Every reverse breakdown event causes some damage to the junction, even if the current is limited. A part that passes reverse withstand testing ten times in a row may not pass the eleventh time because you have been slowly degrading it with each test.

For production screening, test each part only once. For characterization, use a fresh sample for each data point. If you must retest the same part, increase the voltage in smaller increments and watch for any shift in leakage current between runs. A rising leakage curve across repeated tests means the junction is taking cumulative damage.