About the Control Chart Limits Calculator
This control chart calculator turns raw sample data into the center line and upper and lower control limits (UCL and LCL) for the three most widely used Shewhart charts. For measured variables, paste one subgroup per line to get an X-bar chart and a matching R (range) chart. For counts of defective units, use the p chart; for counts of defects per inspection unit, use the c chart.
Quality engineers, Six Sigma green belts, warehouse and production supervisors and students use control charts to tell routine, common-cause variation from special causes that need investigation. The calculator plots every subgroup against its limits and counts the points that fall outside them, so you can see at a glance whether the process looks stable.
Limits use the standard 3-sigma Shewhart formulas and the published A2, D3 and D4 constants for subgroup sizes 2 to 10. Control limits describe what the process is doing, not what the specification requires — compare them with your tolerances using a process capability study.
With the default inputs, the upper control limit (ucl) is 10.3356. Change any value above to recalculate instantly.
How to use the control chart limits calculator
- 1Choose the chart type that matches your data: measurements, defective units, or defect counts.
- 2For X-bar and R, paste one subgroup per line with the same number of readings in each.
- 3For a p chart, list defectives per sample and enter the constant sample size.
- 4For a c chart, list the number of defects found on each inspection unit.
- 5Read the center line and limits, then check the chart for points outside them.
Formula and method
For the X-bar and R chart, each subgroup’s mean and range are calculated first. The center line is the grand mean X̿ and the limits are X̿ ± A₂·R̄, where R̄ is the average range and A₂ is a published constant that converts the average range into a 3-sigma band for subgroup means. The R chart limits are D₃·R̄ and D₄·R̄.
For a p chart with constant sample size n, p̄ is total defectives divided by total units inspected, and the limits are p̄ ± 3 standard errors from the binomial distribution. For a c chart, c̄ is the average defect count and the limits use the Poisson property that the variance equals the mean. Any negative lower limit is set to zero.
- X̿
- Grand mean of subgroup means
- R̄
- Average subgroup range
- A₂, D₃, D₄
- Shewhart constants for the subgroup size
- p̄
- Overall fraction defective
- c̄
- Average number of defects per unit
Worked examples
X-bar and R chart, six subgroups of five
The six subgroup means average 10.0567 and the ranges average 0.4833. With A₂ = 0.577 for n = 5 the X-bar limits are 10.0567 ± 0.2789, so 9.7778 to 10.3356, and the R chart UCL is 2.114 × 0.4833 = 1.0218. No point falls outside, so the process looks stable.
p chart for 12 samples of 200 units
57 defectives in 2,400 units gives p̄ = 0.02375. The standard error is √(0.02375 × 0.97625 ÷ 200) = 0.010767, so UCL = 0.02375 + 3 × 0.010767 = 0.05605 and the lower limit is negative, so it is set to 0.
c chart for defects per circuit board
72 defects across 12 boards gives c̄ = 6. The limits are 6 ± 3√6 = 6 ± 7.3485, so UCL = 13.35 and the negative lower limit is set to 0. The highest count, 10, is inside the limits.
Frequently asked questions
How do you calculate UCL and LCL?+
For an X-bar chart, UCL = grand mean + A₂ × average range and LCL = grand mean − A₂ × average range. For attribute charts the limits are the average ± 3 standard deviations of the binomial (p chart) or Poisson (c chart) distribution, with negative lower limits set to zero.
Which control chart should I use?+
Use X-bar and R for measured values taken in small subgroups, an individuals (I-MR) chart for one reading at a time, a p chart for the fraction of units that are defective, and a c chart for the number of defects on a fixed-size inspection unit.
Are control limits the same as specification limits?+
No. Control limits are calculated from process data and describe its natural variation. Specification limits come from the customer or design. A process can be in control yet still produce out-of-spec parts; process capability indices such as Cpk compare the two.
How many subgroups do I need?+
The common guideline is at least 20 to 25 subgroups collected while the process is running normally. Fewer samples give limits that move a lot as data is added, so recalculate once you have enough history.
What does a point outside the control limits mean?+
It signals a likely special cause — something unusual such as a tool change, new material lot or operator error. Investigate and fix the cause, then consider recalculating the limits without that point. Western Electric rules add other signals, such as runs of points on one side of the center line.