The seven measurements most often automated with an ATE system are DC voltage and current, resistance and continuity, frequency and timing, digital and protocol signals, analog and mixed-signal performance, temperature and environmental response, and RF. Together they cover most of what production and validation tests need to prove about an electronic or electromechanical product. Automating them makes each measurement faster, repeatable and recorded, but only if the setup avoids a handful of well-known errors.
For how these measurements fit into a complete test station, see our complete guide to automated test equipment (ATE). Below, each measurement is covered with the instruments used, what to watch for, and the numbers that matter.
| Measurement | Typical instruments | Most common source of error |
|---|---|---|
| 1. DC voltage and current | DMM, SMU, power supply readback, DAQ | Settling time, ranging, lead and switch resistance |
| 2. Resistance and continuity | DMM (2- and 4-wire), LCR meter, micro-ohmmeter | Contact and lead resistance; thermal EMF |
| 3. Frequency, timing and phase | Counter/timer, oscilloscope, digitizer | Gate time vs resolution; trigger noise |
| 4. Digital and protocol | Digital I/O, pattern generator, protocol analyzers | Timing margins, logic levels, bus loading |
| 5. Analog and mixed-signal | Digitizer, AWG, audio analyzer | Sample rate, noise floor, grounding |
| 6. Temperature and environmental | Thermocouple and RTD inputs, chambers | Soak time, sensor placement, cold-junction error |
| 7. RF | Power meter, spectrum and vector network analyzers | Calibration to the fixture, shielding |
1. DC voltage and current
Supply rails, reference voltages, quiescent and operating current, and sensor outputs are the backbone of most functional tests. A digital multimeter (DMM) gives the best accuracy; a source-measure unit (SMU) can apply a voltage and measure current, or the reverse, down to nanoamps for leakage and sleep-current tests.
The usual mistake is reading too soon. After a relay switches or a supply steps, the signal settles exponentially through the source resistance and the capacitance of cables and inputs.
Worked example: how long to wait before measuring
A sensor output with 1 MΩ source resistance drives 100 pF of cable and input capacitance. The time constant τ = 1 MΩ × 100 pF = 100 µs.
After 5τ the signal is within e−5 ≈ 0.67% of its final value, fine for a rough check. To settle within 0.01% (1 part in 10,000) takes ln(10,000) ≈ 9.2τ ≈ 0.92 ms.
Measure the real settling time on the station, then add a delay with margin. Settling delays that are too short are a leading cause of measurements that drift with timing and intermittent false failures.
Also fix the DMM range and integration time rather than auto-ranging every reading, and choose the integration time as a whole number of power-line cycles to reject 50/60 Hz hum.
2. Resistance and continuity
Winding and coil resistance, shunt resistors, harness continuity, contact resistance and insulation resistance are all resistance measurements, at very different scales.
Worked example: why 4-wire (Kelvin) measurement matters
A 0.5 Ω coil is measured through test leads, a relay matrix and fixture probes totaling about 0.2 Ω.
A 2-wire reading shows 0.7 Ω, a 40% error (0.2 ÷ 0.5), and the error changes as probes wear.
A 4-wire measurement forces current through one pair of wires and senses voltage on a separate pair at the part, so lead and contact resistance drop out. Use it for anything under roughly 100 Ω where accuracy matters, and use offset compensation to cancel thermal EMFs at low resistances.
At the other end of the scale, insulation resistance and hipot tests apply hundreds or thousands of volts and need interlocked, guarded fixtures. Our article on test fixtures covers Kelvin probing and contact design.
3. Frequency, timing and phase
Oscillator frequency, PWM duty cycle, propagation delay, pulse width and phase between channels are measured with a counter/timer, an oscilloscope or a digitizer. The trade-off is resolution against time: a simple counter with a 1 s gate resolves 1 Hz, and a 0.1 s gate only 10 Hz. Reciprocal counters measure the period and achieve much finer resolution in a short gate, which matters when test time is tight.
For timing between edges, check the trigger level and hysteresis against noise on the signal, and remember that every cable and relay adds delay that must be calibrated out for nanosecond-level measurements.
4. Digital and protocol signals
Most modern products contain microcontrollers and buses, so ATE drives digital patterns, reads responses and communicates over I2C, SPI, UART, CAN, LIN, Ethernet and USB. Tests confirm logic levels, timing margins, communication, firmware version and configuration. The ASIC test system we built, for example, generates high-speed digital input patterns while recording digital and analog outputs for correct timing and amplitude.
Test at the specified logic thresholds and timing limits, not just at nominal, to catch marginal parts. Keep the tester from loading the bus: long cables and extra capacitance can make a good board fail or, worse, make a marginal board pass.
5. Analog and mixed-signal performance
Audio paths, sensor interfaces, ADCs and DACs, filters and power converters need dynamic measurements: gain and frequency response, noise, total harmonic distortion (THD), signal-to-noise ratio, ripple, and converter linearity (INL and DNL). An arbitrary waveform generator supplies the stimulus and a digitizer captures the response.
Sample at well above twice the highest frequency of interest, and filter to prevent aliasing; our article on sampling rate and anti-aliasing explains the rule. Grounding matters as much as the instrument: a ground loop through the fixture can add more noise than the product produces. See ground loops and galvanic isolation.
6. Temperature and environmental testing
Products are tested across temperature, humidity, vibration or pressure to prove they meet specification in service and to screen early failures. The ATE controls an environmental chamber, logs many thermocouples or RTDs, and runs functional tests at each condition. The environmental chamber test system we built controls two chambers and logs over 60 thermocouples alongside GPIB instruments and programmable loads.
The usual error is not waiting for the unit, as opposed to the chamber air, to reach temperature. Define soak times from measurements on the product’s thermal mass, place sensors on the part, and account for cold-junction compensation accuracy on thermocouples; our thermocouple data acquisition article covers it.
7. RF testing
Wireless modules, radar and RF front ends are tested for output power, frequency accuracy, spectral mask, return loss or VSWR, sensitivity and, for digital modulation, error vector magnitude (EVM). Instruments include power meters, spectrum analyzers, vector signal analyzers and vector network analyzers.
RF measurements are only as good as the calibration at the reference plane: cables, switches and fixture probes add loss and mismatch that must be measured and removed. Shielded fixtures keep other stations and nearby transmitters from corrupting results.
Getting automated measurements right
- Size instrument accuracy to the tolerance: at least a 4:1 test uncertainty ratio on critical limits, with guardbands.
- Measure settling times on the real station and set delays from data.
- Use 4-wire connections for low resistance and guarded paths for low current.
- Fix ranges and integration times; avoid auto-ranging in production.
- Validate each critical measurement with a gauge R&R study before release.
- Log every value, not just pass or fail, so drift is visible before it causes failures.
These measurements show up across ATE use cases from semiconductors to aerospace, and in every end-of-line test.
Need these measurements automated on your product?
Sciotex designs custom ATE systems that combine the right instruments, switching, fixtures and software for your tolerances and takt time, built and supported in the US since 1997.
Frequently asked questions
Common ATE measurements include DC voltage and current, resistance and continuity, frequency and timing, digital and protocol communication, analog and mixed-signal performance, temperature and environmental response, and RF parameters.
In a 4-wire (Kelvin) measurement, current flows through one pair of wires and voltage is sensed on another pair at the part, so the resistance of leads, relays and fixture probes does not add to the reading. It is essential for low-resistance measurements.
Long enough for the signal to settle to the required accuracy. For an RC settling path, about 5 time constants reaches 0.67% and about 9.2 reaches 0.01%. Measure real settling on the station and add margin.
A common rule is a test uncertainty ratio of at least 4:1: the total measurement uncertainty should be no more than a quarter of the tolerance being checked, with limits guardbanded by that uncertainty.
Yes. Most ATE systems combine DMMs or digitizers for analog measurements with digital I/O and protocol interfaces, switched to the unit under test through a relay matrix and fixture.
Related articles
- What Is Automated Test Equipment (ATE)? The Complete Guide
- Test Fixtures for Automated Testing: Bed of Nails, Pogo Pin and Functional Fixtures
- In-Circuit Test vs Functional Test: ICT, Flying Probe, Boundary Scan and FCT Compared
- Automated Test Software: Test Executives, TestStand, LabVIEW and Python
- Automated Test Equipment Use Cases: Semiconductor, Automotive, Aerospace and More
- End-of-Line Testing: How EOL Test Systems Work and How to Specify One