Control Charts in SPC: Types, Rules, and How to Read Them

1. Center Line (CL)

The center line represents the process average (mean). It reflects the typical performance of the process when it is stable.

2. Upper Control Limit (UCL)

The upper control limit is the highest value the process is expected to reach under normal conditions. It is calculated statistically, not based on specifications.

3. Lower Control Limit (LCL)

The lower control limit is the lowest value the process is expected to reach under normal conditions.

The control limits are usually set at ±3 standard deviations (3σ) from the process mean, which captures approximately 99.73% of normal process variation.

It is important to note that control limits are not the same as specification limits. Specification limits are defined by customer or design requirements, while control limits are determined by process behavior.

Common Cause Variation

• Inherent to the process
• Always present
• Caused by normal factors such as machine variation, material differences, or environmental conditions
• Can only be reduced by improving the process itself

Special Cause Variation

• Unusual and unexpected
• Caused by specific events such as equipment failure, operator error, or incorrect setup
• Indicates that the process is out of control
• Requires investigation and corrective action

Control charts are designed specifically to identify special cause variation.

1. X-bar and R Chart

Purpose:

Used to monitor the process mean and variability when data are collected in subgroups.

When to Use:

• Subgroup size between 2 and 10
• Measurements are continuous

How It Works:

• X-bar chart tracks changes in the subgroup averages
• R chart tracks changes in the range within subgroups

Typical Applications:

Machining processes, dimensional measurements, and manufacturing operations.

2. X-bar and S Chart

Purpose:

Monitors process mean and variability using standard deviation instead of range.

When to Use:

• Subgroup size greater than 10
• Large data sets where standard deviation is more accurate

Advantages:

Provides a better estimate of process variation than the R chart for large subgroups.

3. Individuals (I) and Moving Range (MR) Chart

Purpose:

Used when data are collected one observation at a time.

When to Use:

• No logical subgroups
• Low production volume
• Long cycle times

How It Works:

• I chart tracks individual measurements
• MR chart tracks the difference between consecutive observations

Typical Applications:

Service processes, batch production, and laboratory measurements.

4. P Chart (Proportion Defective)

Purpose:

Monitors the proportion of defective units in a sample.

When to Use:

• Data are pass or fail
• Sample size may vary

Typical Applications:

Final inspection results, supplier quality data, audit findings.

5. NP Chart (Number of Defectives)

Purpose:

Tracks the number of defective units in each sample.

When to Use:

• Sample size remains constant
• Data are pass or fail

6. C Chart (Count of Defects)

Purpose:

Monitors the number of defects per unit.

When to Use:

• Defects can occur multiple times on a single unit
• Area of opportunity remains constant

Typical Applications:

Surface defects, cosmetic flaws, documentation errors.

7. U Chart (Defects per Unit)

Purpose:

Tracks defects per unit when sample size varies.

When to Use:

• Variable sample sizes
• Counting defects rather than defective units

Rule 1: One Point Beyond Control Limits

A single data point outside the UCL or LCL signals an out-of-control condition and requires immediate investigation.

Rule 2: Run of Points on One Side of the Center Line

Typically, seven or more consecutive points above or below the center line indicate a process shift.

Rule 3: Trend or Continuous Movement

Six or more consecutive points steadily increasing or decreasing suggest a trend that may lead to instability.

Rule 4: Cyclic or Repeating Patterns

Regular up-and-down patterns may indicate external influences such as shift changes, temperature cycles, or machine wear.

Rule 5: Sudden Changes in Variation

A noticeable increase or decrease in spread may signal a change in measurement system, materials, or operating conditions.

Environmental conditions can significantly impact product performance and reliability. For electronics manufacturers, our article [Thermal & Environmental Testing in Electronics: Standards, Methods & Best Practices] explores how temperature, humidity, and stress testing help validate product durability and compliance.

Step 1: Check Control Limits

Verify that all points fall within the control limits.

Step 2: Look for Patterns

Identify trends, runs, or cycles that may indicate instability.

Step 3: Apply Control Chart Rules

Use standard SPC rules to confirm whether special cause variation exists.

Step 4: Investigate Causes

If a rule is violated, identify potential root causes such as equipment issues, operator changes, or environmental factors.

Step 5: Take Corrective Action

Remove the special cause and document the action taken.

Step 6: Continue Monitoring

Ensure the process returns to and remains in a state of statistical control.

• Adjusting a stable process unnecessarily
• Ignoring out-of-control signals
• Using insufficient data
• Mixing data from different processes or machines
• Failing to verify measurement system accuracy
• Recalculating control limits too frequently

Avoiding these mistakes ensures control charts provide meaningful insights.