How first pass yield is calculated
First pass yield = units that pass the first time ÷ units that entered the step
A unit counts in the numerator only if it went through the step once and passed. Anything that was reworked, repaired, retested after a failure or scrapped is left out, even if it ships in the end.
For example, 1,000 boards enter functional test. 940 pass on the first attempt, 45 fail and pass after repair, and 15 are scrapped.
| Measure | Calculation | Result |
|---|---|---|
| First pass yield | 940 ÷ 1,000 | 94.0% |
| Final yield | (940 + 45) ÷ 1,000 | 98.5% |
| Scrap rate | 15 ÷ 1,000 | 1.5% |
Both figures are correct. A factory that reports 98.5% is describing what it shipped. FPY describes how well the process ran.
Rolled throughput yield: good steps add up to a weaker line
A product passes through several steps, each with its own FPY. Rolled throughput yield (RTY) multiplies them to give the chance that a unit gets through every step without any rework:
| Step | FPY |
|---|---|
| SMT and automated optical inspection | 98% |
| In-circuit test | 97% |
| Functional test | 95% |
| Final assembly and inspection | 99% |
| Rolled throughput yield | about 89.4% |
Each step looks healthy on its own, yet roughly one unit in ten needs attention somewhere. That rework loop uses operators, test-station time and line capacity, so RTY is a better guide to real cost and capacity than any single-station figure.
Why buyers should care
- Cost. Rework labor, retest time and scrap are paid for somewhere, usually in the unit price or in the next price discussion.
- Reliability. A reworked unit is not always as good as a first-pass unit. Hand soldering, BGA rework and repeated heating can leave latent defects. Rework should follow agreed workmanship standards such as IPC-7711/7721, and some buyers limit how many times a board may be reworked.
- Capacity. A line with low FPY needs more hours to ship the same quantity, which matters most during ramp-up.
- Early warning. A falling FPY trend often shows a material, fixture or process problem before it reaches outgoing inspection.
Agree the definition before comparing numbers
Two factories can report different FPY for the same performance because they count differently. Settle these points in the quality plan:
- Retests. Whether a unit that fails, is retested without repair and then passes counts as first pass. If one retest is allowed, report the retest rate separately.
- Level. Per station, per line or per finished product. Ask for per-station figures; a line-level number hides where the losses are.
- Source. Test logs or production system data, not manual tallies.
- Scope. Per build, per SKU and per shift, so a bad shift or material lot shows up.
- Targets. Mass-production targets per station, with the action expected when one is missed.
Judge FPY as a trend across builds rather than as a single figure. A low number in an engineering build with well-understood failure modes is less worrying than a flat number nobody can explain, and the pilot run is where the per-station targets should first be met.
What to ask a factory for
- FPY by station for recent builds of your product, or a comparable one
- A Pareto chart of the top failure modes at each station
- Retest and rework rates, reported separately from FPY
- Rework records linked to serial numbers
- The corrective actions open for the top failures, normally tracked through CAPA or an 8D report
FPY per station is one of the main exit criteria for moving from a pilot run into mass production, alongside process capability measures such as Cpk.
Where Zefinity helps
Reading yield data with the factory, chasing the top failure modes and judging when a line is ready to run faster fall under QA and corrective action follow-up and production readiness and ramp-up, both part of Zefinity’s Launch service.
Frequently asked questions
What is the difference between first pass yield and final yield?
First pass yield counts only units that passed without any rework, repair or retest. Final yield counts every unit that eventually passed, including reworked ones. A line can show 98% final yield and 90% first pass yield at the same time. The gap is the rework the factory is doing, which costs time and capacity and can affect reliability.
What is a good first pass yield for electronics?
There is no single good number. It depends on product complexity, test coverage, where in the line it is measured and how mature the product is. Early engineering builds run well below mass-production levels. Agree a target per station in the quality plan, check that it is measured the way you expect, and look at the trend through the pilot run.
Does a retest count against first pass yield?
It depends on the definition you agree. Many lines allow one automatic retest for fixture contact problems and still count a pass as first pass. That can be reasonable, but only if the rule is written down and the retest rate is reported separately. A high retest rate usually points to a fixture or test-limit problem.