Quality & InspectionFree worksheetPDF, Excel, Word exportSemiconductors

Wafer Sort Test Yield Pareto Worksheet

A test yield Pareto shows which fail bins cost you the most good die. This worksheet takes bin counts from wafer sort or final test and returns the loss and DPM for each bin, plus the total test yield.

Download:

Bin	Failure mode	Test stage	Failing die	Tested die	Loss (% of tested)	DPM
					2.55	25,500
					6.1667	61,667
					9.625	96,250
					3.9167	39,167
					2.1667	21,667
					1.5208	15,208

Total failing die

6,227

Max tested die

24,000

Largest single bin

2,310

Total yield loss

25.9%

Failing die across listed bins over tested die.

Test yield

74.1%

Top bin share of all loss

37.1%

Shows how concentrated the loss is.

Sort the rows by failing die from largest to smallest to read the Pareto, and focus on the bins that reach about 80 percent of cumulative loss. Enter the same tested die count on every row for one lot or wafer sort period. Bin numbers are examples; use your own bin map.

Yield loss across systems

Want sort bins and ERP yield postings reconciled automatically? Ask for an optional expert review.

Optional. The tool and every export work without this. No spam.

How to use the worksheet

Enter one row per fail bin. Add the bin number from your test program, a short failure mode, the test stage, the number of failing die and the tested die count. The sheet calculates the loss as a percentage of tested die and the failures in parts per million. The summary adds the total yield loss, test yield and how much loss sits in the largest bin. Sort the table by failing die and export to Excel to build a cumulative line and a Pareto chart.

The Pareto method

A Pareto separates the vital few causes from the trivial many. In test yield, a handful of bins usually account for most lost die, so fixing the top two or three has the highest return. Calculate loss per bin, rank them, and accumulate. Look at the Pareto by lot, by wafer position, by tool or by product to find where one bin concentrates, since a bin that clusters on edge die or one lot points to a different root cause than a bin spread evenly.

Rank by failing die, not by bin number
Break down by lot, wafer slot and probe card
Separate repairable from non repairable fails
Review recurring bins after every process change

Common mistakes and next steps

Mistakes include mixing different tested counts across rows, adding bins from different test stages into one total, and treating retests as new failures. If one bin jumps, check the test program revision, probe card contacts and tester calibration before blaming the process. Link persistent bins to wafer maps, inline defect data and your OCAP. In ERP, scrapped die and downgraded bins feed yield reporting and inventory, so make sure the bin mapping is consistent with what is booked.

Frequently Asked Questions

How do I read the Pareto?

Sort bins by failing die from largest to smallest and add up the loss. The first few bins usually cover most of the yield loss. Those are where process, design or test fixes pay back the most.

What is DPM here?

DPM is failing die per million tested die in that bin. It lets you compare bins and lots with different volumes on one scale.

Should I include retests?

Count the final disposition of each die. If a die fails and passes on retest, record it as a pass, but track the retest rate separately because it signals probing or contact problems.

Can I use it for final test?

Yes. Choose the test stage per row. Keep wafer sort and final test in separate sheets if you want a clean total for each stage.

Netray can bring sort bin data and ERP yield postings together in a private assistant that lets engineers ask which bin is moving and which lots are affected.

Talk to an Expert Explore More Free Tools

Related Tools

Quality & Inspection

More Quality & Inspection tools

See all

Back to All Tools