C Chart Calculator

Example data table

Formula used

How to use this calculator

When a c-chart fits the process

A c-chart tracks defect counts when each sample has the same inspection opportunity. Examples include defects per panel or blemishes per roll. It assumes a Poisson count process where mean and variance are similar. If inspected area, time window, or unit definition changes, the chart can mislead. Keep sampling definitions fixed and log planned changes with the data. Use it for repeated comparable lots, shifts, or time blocks on purpose.

Center line and limits you can defend

The center line is c̄, the average count across n samples. For counts, the standard deviation estimate is √c̄, so limits are c̄ ± k√c̄, with LCL truncated at zero. If c̄ = 4.50, then √c̄ ≈ 2.12, UCL ≈ 10.86, and LCL = 0.00. Use limits to judge stability before comparing lots, shifts, or suppliers. Most teams start with k = 3 for routine monitoring as default.

Patterns that often signal special causes

A point above UCL suggests an unusual spike worth reviewing. Runs on one side of c̄ can indicate a sustained shift after setup changes or new material. Six-point trends may reflect gradual drift, such as tool wear or buildup. Warning rules using 1σ and 2σ bands increase sensitivity and false alarms. Apply rules consistently and document them in reports. If you enable multiple rules, review signals as a package, not individually.

Making the data trustworthy

Before interpreting signals, confirm that counting is consistent. Define defects, train inspectors, and audit for missed or double counts. Record product mix and environment because mixed conditions inflate variation. Avoid mixing rework with first-pass counts unless intended. For very low counts, group samples; for very high counts, verify one stable opportunity scale. Stratify by product family and recalc limits per family.

Turning chart output into action

Treat the chart as a decision aid: stabilize first, then improve. When a signal occurs, review sample context and check measurement, settings, and incoming material. After removing a special cause, keep original limits until performance is stable, then recalculate a new baseline. Use exports in audits by pairing limits with actions, dates, and verification results. Link investigations to CAPA records and confirm improvements.

FAQs

1) What does a c-chart measure?

It measures the number of defects per sample when each sample has the same inspection opportunity, such as the same area, unit, or time window.

2) How many samples should I use to set limits?

Use at least 20 to 25 samples when possible. More samples improve the estimate of c̄ and reduce the chance that early special causes distort the limits.

3) Why is the lower control limit sometimes zero?

Counts cannot be negative. When c̄ − k√c̄ is below zero, the limit is truncated to zero so the chart remains meaningful for low-defect processes.

4) What if the inspection size changes between samples?

A c-chart is no longer appropriate. Use a u-chart that normalizes defects by opportunity, or redesign sampling so each sample has the same inspection size.

5) Should I use k = 2 or k = 3?

k = 3 is the common default for stability monitoring. k = 2 is more sensitive but can trigger more false alarms, especially with short datasets.

6) Do run rules always mean the process is bad?

Not always. Run rules indicate non-random patterns that deserve investigation. Confirm with context, repeatability checks, and evidence of a plausible special cause.