{"id":7130,"date":"2026-10-21T02:15:54","date_gmt":"2026-10-21T02:15:54","guid":{"rendered":"https:\/\/pickandplacemachine.com\/?p=7130"},"modified":"2026-10-21T02:15:55","modified_gmt":"2026-10-21T02:15:55","slug":"how-to-calculate-the-cost-of-poor-quality-in-smt-manufacturing","status":"publish","type":"post","link":"https:\/\/pickandplacemachine.com\/fr\/how-to-calculate-the-cost-of-poor-quality-in-smt-manufacturing\/","title":{"rendered":"How to Calculate the Cost of Poor Quality in SMT Manufacturing"},"content":{"rendered":"

To calculate the cost of poor quality (COPQ) in SMT manufacturing, add every cost your line incurs because a board, component, or process step did not conform the first time. That means scrap, rework, re-inspection, retest, failure analysis, quality-related line stoppages, and customer returns. Divide the total by boards built, placements made, or revenue. Do this by process step, not only as one plant-wide number. Most SMT plants track first-pass yield and scrap but miss rework labor, machine time lost to feeder or splice stops, and the engineering hours spent on failure analysis. This guide shows the formula, a step-by-step method, a worked example with illustrative numbers, and how to use the result to decide where prevention spending pays back.<\/p>\n\n\n\n

\"How<\/figure>\n\n\n\n

What Counts as Cost of Poor Quality on an SMT Line?<\/h2>\n\n\n\n

Le American Society for Quality (ASQ) defines cost of poor quality<\/a> as the costs of providing poor quality products or services. ASQ splits it into two groups. Internal failure costs come from defects found before the product reaches the customer. External failure costs come from defects the customer finds. Prevention and appraisal costs, such as training, process validation, SPI, AOI, and test, sit outside COPQ but belong in the wider cost of quality (COQ) picture.<\/p>\n\n\n\n

For an SMT line, the categories translate into specific ledger items:<\/p>\n\n\n\n

Cost category<\/th>SMT examples<\/th>Where to get the data<\/th><\/tr><\/thead>
Internal failure: scrap<\/td>Scrapped PCBs, components, solder paste, and stencils; boards that fail repair<\/td>ERP scrap transactions, bin counts<\/td><\/tr>
Internal failure: rework and re-inspection<\/td>Touch-up, BGA or QFN rework, cleaning, repeat AOI\/ICT\/FCT<\/td>MES repair stations, repair labor logs<\/td><\/tr>
Internal failure: line losses<\/td>Stops for feeder jams, splice failures, nozzle problems, misprints, wrong-reel loading<\/td>Machine stop logs, OEE records<\/td><\/tr>
Internal failure: analysis and handling<\/td>Failure analysis, MRB meetings, baking of moisture-exposed parts, expedited replacement material<\/td>Engineering timesheets, purchasing records<\/td><\/tr>
External failure<\/td>RMAs, field repair, replacement units, freight, customer penalties, lost orders<\/td>RMA database, finance, sales<\/td><\/tr>
Prevention (not COPQ)<\/td>Preventive maintenance, ESD program, operator training, feeder calibration<\/td>Maintenance and training budgets<\/td><\/tr>
Appraisal (not COPQ)<\/td>SPI, AOI, X-ray, ICT\/FCT, audits<\/td>Equipment depreciation, test labor<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

The Basic COPQ Formula for SMT Assembly<\/h2>\n\n\n\n

COPQ = Internal failure costs + External failure costs<\/p>\n\n\n\n

Total cost of quality = Prevention + Appraisal + Internal failure + External failure<\/p>\n\n\n\n

The formula is simple. The difficulty is allocation. A defect found at AOI costs far less than the same defect found at functional test, and far less again than one found at the customer. So the calculation needs a cost per defect at each detection point.<\/p>\n\n\n\n

Step 1: Set the scope and the unit of measure<\/h3>\n\n\n\n

Choose one period (a month works for most plants) and one denominator. Cost per board is the most practical for comparing product lines. Cost per placement or per million placements helps compare boards of different complexity. Percentage of revenue helps when presenting to finance. Keep the same unit when you re-measure after a process change.<\/p>\n\n\n\n

Step 2: Map defects to the process step that caused them<\/h3>\n\n\n\n

Record each defect by where it was created, not where it was found. A bridged fine-pitch pad found at AOI usually started in printing. A tombstoned 0402 might trace to paste volume, placement offset, pad design, or reflow profile. A missing component can trace to a feeder, a nozzle, or a reel that ran out. Without this mapping, COPQ only shows how much you lose. It does not show where to spend.<\/p>\n\n\n\n

Step 3: Calculate the cost per defect at each detection point<\/h3>\n\n\n\n

For each defect type, use:<\/p>\n\n\n\n

Cost per defect = (rework minutes \u00d7 loaded labor rate per minute) + materials consumed + repeated inspection or test time + machine time consumed + handling and administration<\/p>\n\n\n\n

Use a loaded labor rate that includes benefits and overhead, and use real repair-station times, not standard times. Add scrap value separately for boards that cannot be recovered. For rework on dense packages, include the risk that extra thermal exposure reduces reliability. That risk is hard to price, so flag it in your report instead of ignoring it.<\/p>\n\n\n\n

Step 5: Add external failure costs<\/h3>\n\n\n\n

For each customer return, add inbound freight, RMA administration, failure analysis, repair or replacement, outbound freight, and any contractual penalty or credit. Customer-side line-down charges and lost future orders are real costs, but they are often negotiated and uncertain. Report them as a separate line, so the core calculation stays traceable to documents.<\/p>\n\n\n\n

\"How<\/figure>\n\n\n\n

Worked Example: An Illustrative Monthly COPQ Calculation<\/h2>\n\n\n\n

The numbers below are hypothetical and chosen to show the arithmetic. They are not industry benchmarks, and your costs will differ.<\/p>\n\n\n\n

Assumptions: 20,000 boards built per month, average selling price $45 per board, loaded labor $24 per hour, line cost $180 per hour ($3 per minute).<\/p>\n\n\n\n

Objet<\/th>Basis<\/th>Monthly cost<\/th><\/tr><\/thead>
Rework (800 boards fail first pass)<\/td>12 min rework ($4.80) + 3 min re-inspection ($1.20) + $0.60 materials = $6.60 per board<\/td>$5,280<\/td><\/tr>
Scrap (80 boards)<\/td>$38 materials and value added per board<\/td>$3,040<\/td><\/tr>
Quality-related line stops<\/td>40 stops \u00d7 9 min = 360 min \u00d7 $3 per min<\/td>$1,080<\/td><\/tr>
Failure analysis<\/td>20 engineering hours \u00d7 $40<\/td>$800<\/td><\/tr>
Internal failure subtotal<\/strong><\/td><\/td>$10,200<\/strong><\/td><\/tr>
External failure (12 returns)<\/td>$25 inbound handling + $60 analysis and repair + $38 replacement credit = $123 per return<\/td>$1,476<\/td><\/tr>
Total COPQ<\/strong><\/td><\/td>$11,676<\/strong><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

That equals about $0.58 per board, or about 1.3% of $900,000 in monthly revenue. Two points stand out. Rework is the largest line, and stoppages look small in this case but may hide far larger capacity losses if the line is a bottleneck. The example also excludes expedited freight, overtime, lost customer goodwill, and long-term reliability risk, so the true figure is higher.<\/p>\n\n\n\n

Why First-Pass Yield Alone Understates COPQ<\/h2>\n\n\n\n

A common assumption is that a high first-pass yield means low cost of poor quality. It does not always hold, for three reasons.<\/p>\n\n\n\n

First, yield counts boards, not cost. A 96% first-pass yield on a simple board and on a dense board with BGAs gives very different rework bills.<\/p>\n\n\n\n

Second, repaired boards count as passes at final test but still consumed labor, inspection, and thermal exposure. Rework is not neutral. Each extra heat cycle on a component or pad is an additional stress, and some customers restrict the number of rework cycles allowed.<\/p>\n\n\n\n

Third, yield says nothing about escapes. A line with excellent in-house yield and weak inspection coverage can show low internal failure cost and high external failure cost. The right comparison is the cost of the whole detect-and-fix chain, which is why COPQ should be read together with appraisal cost and the escape rate from AOI and test.<\/p>\n\n\n\n

This does not mean first-pass yield is a bad metric. It means yield is the trigger for the cost calculation, not the result.<\/p>\n\n\n\n

\"How<\/figure>\n\n\n\n

Where SMT Quality Losses Start, and What Prevention Looks Like<\/h2>\n\n\n\n

Once the defect map is built, losses usually cluster in a few places. The sections below link each to the kind of prevention spending buyers evaluate. Treat product pages as starting points. Confirm specifications and compatibility with your own process and equipment documentation.<\/p>\n\n\n\n

Solder paste printing<\/h3>\n\n\n\n

Printing is the first process step where a defect can be created, and every downstream process inherits it. Insufficient paste, bridging, smearing, and misregistration are paid for in rework or in escapes to later stages. Printer repeatability, stencil alignment, under-stencil cleaning, and SPI feedback all affect the defect rate. Buyers evaluating printer capacity can review the GDK TSE printer listing<\/a>, which is filed under solder paste printers. Confirm the exact model, configuration, and intended use with the supplier before treating it as a like-for-like replacement.<\/p>\n\n\n\n

Feeders, splicing, and reel handling<\/h3>\n\n\n\n

Feeder and reel problems are among the easiest to measure, because they appear in machine stop logs. Typical causes are worn feeder mechanisms, tape that advances inconsistently, splices that fail in the feeder path, and the wrong reel loaded at the wrong position.<\/p>\n\n\n\n