OPC UA Sampling Rate Calculator
Calculate the required OPC UA sampling rate and publishing interval based on signal dynamics and acceptable latency.
OPC UA
sampling rate
publishing interval
latency
SCADA
real-time
IIoT
Input values
⚙
Enter values and click Calculate to see results here.
Formula
f_sample ≥ 2 × f_signal (Nyquist theorem)
T_publish = T_sample × N_samples
Latency ≈ T_publish + network_delay
Worked Example
Given: A vibration signal at 500 Hz, acceptable latency of 10 ms.
Step 1: Min sampling = 2 × 500 = 1000 Hz (Nyquist). Use 5× for accuracy: 2500 Hz.
Step 2: If publishing every 10 samples: T_publish = 10/2500 = 4 ms
Step 3: Total latency = 4 ms (buffer) + ~2 ms (network) = 6 ms < 10 ms ✓
For slow-changing signals (temperature), 1 Hz sampling is sufficient.
Engineering Notes
- Nyquist theorem: The sampling rate must be at least 2× the highest signal frequency. In practice, use 5-10× for accurate waveform reconstruction.
- OPC UA sampling vs publishing: Sampling is how often the server reads the value. Publishing is how often it sends to the client. Publishing can be slower than sampling.
- Bandwidth impact: Higher sampling rates generate more data. A 1000 Hz sampling rate with 4-byte values = 4 KB/s per variable. Scale accordingly.
- Browse/Subscribe modes: For monitoring, polling (browse) at 1-10 Hz is usually sufficient. For control, subscription at 10-100 Hz is better.
- Typical rates: Temperature: 0.1-1 Hz. Pressure/Flow: 1-10 Hz. Vibration: 1000-10000 Hz. Motor current: 100-1000 Hz.
References
- IEC 62541 — OPC Unified Architecture
- OPC UA Part 8 — Data Access