Test method for forward characteristics of discrete semiconductors

How to Test Forward Characteristics of Discrete Semiconductors the Right Way

Pulling a diode or transistor off the shelf and slapping a multimeter on it tells you almost nothing useful. The forward voltage you read on a cheap meter is a single data point at one arbitrary current. Real forward characteristic testing means mapping the entire V-I curve under controlled conditions. That curve tells you if the part is healthy, if it matches the datasheet, and whether it will survive in your actual circuit.

What Forward Characteristics Actually Mean

The forward characteristic of a discrete semiconductor is the relationship between the voltage across the device and the current flowing through it when it is biased in the conducting direction. For a diode, this is the classic exponential curve. For a BJT, you are looking at the base-emitter junction behavior. For a MOSFET, it is the body diode or the channel conduction depending on how you set up the test.

The key parameters people extract from this curve are the forward voltage drop at a specified current, the dynamic resistance (the slope of the curve at a given operating point), and the knee voltage where the device starts conducting significantly. A healthy silicon diode should show roughly 0.6 to 0.7V at 10mA. If you measure 0.4V, the part is either a Schottky or it is damaged. If you read 0.9V, something is wrong.

But here is the thing most people miss: forward voltage is not a fixed number. It changes with current, temperature, and even the history of the device. A single measurement at room temperature with a multimeter does not capture any of that.

Setting Up a Proper Forward Curve Test

Using a Curve Tracer for Full Characterization

A curve tracer is the gold standard for this job. It sweeps the voltage across the device while measuring the resulting current, and it plots the curve in real time. You can see the knee, the exponential region, and any abnormal behavior like soft breakdown or leakage all on one screen.

Connect the device under test to the curve tracer terminals with proper polarity. For a diode, anode to the positive terminal, cathode to negative. For a BJT, you test the base-emitter junction the same way. Set the current range so the sweep covers from microamps up to the maximum rated forward current. Start low and work up — you do not want to blow the part during setup.

The curve tracer also lets you do comparison testing. You can overlay a known good device against the one you are testing. If the curves do not match within a reasonable tolerance, the part is suspect even if the multimeter reading looked fine.

Source Measure Unit Method for Precision Work

When you need accurate numbers at specific current levels — say you are characterizing a device for a precision analog circuit — a source measure unit (SMU) gives you better control than a curve tracer. You program the exact current you want, the SMU forces that current through the device, and it reads back the voltage drop with high resolution.

Step the current in logarithmic increments: 10uA, 100uA, 1mA, 10mA, 100mA. At each step, let the reading settle before recording. This gives you a clean data set you can plot later. The advantage over a curve tracer is repeatability and the ability to test at exact operating points rather than a continuous sweep that can smear out fine details.

Temperature Effects You Cannot Ignore

Why Room Temperature Testing Lies to You

Forward voltage drops by roughly 2mV per degree Celsius for silicon devices. That sounds small until you realize a 50-degree temperature swing shifts Vf by 100mV. If you are testing at 25°C and the device will operate at 85°C in the field, your room-temperature curve is basically useless for predicting real-world behavior.

For any serious characterization, you need to test at multiple temperatures. A hot plate or a temperature-controlled chuck under the device works. Take the forward curve at 25°C, then at 75°C, then at 125°C if the part is rated for it. Plot all three on the same graph. The spread between them tells you how much thermal drift to expect in your circuit.

Self-Heating Skews Your Readings

When you push current through a semiconductor, it heats up from the inside. At high currents, the junction temperature can be significantly higher than the ambient temperature you think you are testing at. This means the forward voltage you measure at 100mA includes the effect of self-heating, which makes Vf look lower than it actually is at the nominal temperature.

The fix is to use short pulses instead of DC current. A pulse width of 100 microseconds to 1 millisecond gives the device enough time to conduct but not enough time to heat up. Most curve tracers have a pulse mode for exactly this reason. If you are using an SMU, configure it for pulsed output with a low duty cycle.

Common Mistakes That Ruin Test Data

One of the most frequent errors is testing with long leads. If you are using a breadboard or clip leads, the resistance of those leads adds to your forward voltage measurement. At 10mA, even 0.5 ohms of lead resistance adds 5mV to your reading. That does not sound like much, but when you are comparing parts or validating against a tight spec, it throws everything off.

Use four-wire (Kelvin) connections whenever possible. The curve tracer does this natively. With an SMU, you need to wire the sense leads directly to the device pins, separate from the force leads. This eliminates lead resistance from the measurement entirely.

Another mistake is testing right after the part has been sitting in a humid environment. Moisture on the package can create leakage paths that distort the forward curve, especially at low currents. Dry the part and let it equilibrate to room temperature before testing.

Interpreting the Curve When Something Looks Wrong

A healthy diode curve is smooth and exponential. If you see a knee that is too soft, the junction may be degraded. If there is a sudden jump in current at a lower-than-expected voltage, the device could have a partial short. If the curve does not reach the expected current at the rated voltage, the series resistance is too high — possibly from a damaged bond wire or contaminated contact.

For BJTs, the base-emitter forward curve should track a diode curve closely. Any deviation at higher currents indicates base resistance problems or emitter degradation. The forward beta (hFE) at a given current can be cross-checked by measuring the collector current while forcing a known base current, but that is a separate test.

Do not trust a single data point. Trust the shape of the curve. A part can read the correct forward voltage at one current and still be failing at another. The full curve is the only thing that tells the whole story.