From a probe contact to repeatable data
How to structure repeated I–V sweeps so you can tell device variation from measurement variation.
Repeat before moving the probe
When two sweeps disagree, it helps to know what changed between them. Start by repeating the sweep with the probe in the same position. Then lift and re-land on the same pad and repeat it again. Finally, measure another device under the same settings. These comparisons separate variation within one contact, variation between contacts, and differences across devices.
Log the conditions with the trace
Save device and pad identifiers, sweep direction, voltage step, delay, current range, and compliance events. Keep the raw exports alongside your analysis. A plot labeled only with a sample name makes it difficult to tell whether an odd point came from the device, a range change, or a run that reached compliance.
Make the analysis easy to audit
Overlay repeated runs in Python before averaging them. Mark any exclusions and explain why they were excluded. If you report leakage or breakdown, calculate the value for each run using the same rule, then summarize the spread. A trace that disagrees with the others is worth keeping. It tells you where to look when the next measurement behaves differently.