How to Calculate the Cost of Poor Quality in SMT Manufacturing

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.

What Counts as Cost of Poor Quality on an SMT Line?

En American Society for Quality (ASQ) defines cost of poor quality 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.

For an SMT line, the categories translate into specific ledger items:

The Basic COPQ Formula for SMT Assembly

COPQ = Internal failure costs + External failure costs

Total cost of quality = Prevention + Appraisal + Internal failure + External failure

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.

Step 1: Set the scope and the unit of measure

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.

Step 2: Map defects to the process step that caused them

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.

Step 3: Calculate the cost per defect at each detection point

For each defect type, use:

Cost per defect = (rework minutes × loaded labor rate per minute) + materials consumed + repeated inspection or test time + machine time consumed + handling and administration

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.

Step 4: Add quality-related line stoppages

Many losses leave no defective board. A splice that jams a feeder, a nozzle clog, or a wrong-reel load stops a line that carries depreciation, labor, power, and floor space whether it is producing or not.

Stoppage cost = minutes lost × line cost per minute

Line cost per minute is the hourly cost of the line (equipment depreciation, direct labor, overhead) divided by 60. If the line is capacity-constrained and orders are being delayed, the true cost is higher because lost output has a margin attached. If the line has spare capacity, the depreciation-based figure is the more defensible one. State which basis you used.

Step 5: Add external failure costs

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.

Worked Example: An Illustrative Monthly COPQ Calculation

The numbers below are hypothetical and chosen to show the arithmetic. They are not industry benchmarks, and your costs will differ.

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).

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.

Why First-Pass Yield Alone Understates COPQ

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

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.

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.

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.

This does not mean first-pass yield is a bad metric. It means yield is the trigger for the cost calculation, not the result.

Where SMT Quality Losses Start, and What Prevention Looks Like

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.

Solder paste printing

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, 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.

Feeders, splicing, and reel handling

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.

• Feeder condition. Feeders with stored usage history and maintenance alerts make it easier to link stops to specific units. For Yamaha YSM-series lines, the Yamaha Smart Feeder SS type is one option. Check machine-model compatibility and feeder-ID workflow before ordering.
• Splicing. A splice that is too thick, stretches, or mismatches carrier tape width can jam a feeder or cause a mis-pick. The SMT double splicing tape for component reels is listed in standard tape widths from 8 mm to 44 mm. Ask for thickness, adhesive type, and tensile data, and trial it on your actual feeders.
• Reel storage. Organized storage reduces wrong-reel and damaged-reel events. A SMD component reel storage rack helps with organization and physical protection. It does not replace moisture control. IPC/JEDEC J-STD-033 covers the handling, packing, shipping, and use of moisture, reflow, and process sensitive devices, so moisture-sensitive parts still need the dry-pack and exposure-time controls that standard describes.

Placement accuracy and machine capability

Placement defects include offsets, rotated or missing parts, and pick errors on small components. These depend on the machine, nozzles, feeders, vision settings, and component mix together. When a machine cannot hold placement quality across your component range, the COPQ report will show recurring offsets on specific packages. Buyers comparing equipment can look at the Hanwha Decan pick and place machines listed as high-speed chip shooters. Ask for placement-accuracy data on your smallest component and fine-pitch parts, not only headline speed.

Control ESD

ESD damage can fail a part immediately, or it can leave a latent defect. The ESD Association explains that latent ESD damage may escape immediate detection but can cause premature failure. That makes ESD a driver of external failure cost, which is typically the costliest category per defect. Ionizers support a program but do not replace grounding, ESD-safe packaging, and training. If you are evaluating an ionizing air blower snake for assembly or repair stations, ask for decay-time and offset-voltage test data and the verification and cleaning schedule.

Preventive maintenance

Dry or contaminated mechanisms cause drift, jams, and intermittent stops, and those show up as stoppage minutes. High-temperature equipment such as conveyor chains and bearings in ovens needs lubricants rated for the temperature. The AS2 high-temperature industrial grease page lists oven, furnace, and bearing applications. Use only lubricants that are approved in your equipment manufacturer’s maintenance manual, and keep them away from stencils, boards, and other contamination-sensitive areas.

Common Mistakes When Calculating COPQ

How to Turn the Result into a Purchasing Decision

Rank defect types by monthly cost and by how controllable the cause is. Then compare the cost of each prevention option with the COPQ it can realistically avoid:

Simple payback (months) = prevention investment ÷ monthly COPQ avoided

Use conservative reduction estimates and measure results after a trial period. Low-cost consumables and storage changes are easy to trial. Capital equipment needs a stricter case: include the effect on throughput, changeover, spare parts, training, and support, and compare quotations on total cost of ownership, not only unit price. Lowest purchase price is not always lowest total cost. A cheaper consumable or feeder that causes more stops can cost more than a premium part within a few months, but that conclusion needs your own stop-log data and not a rule of thumb.

What Should You Gather Before Calculating COPQ or Requesting a Quote?

Before you contact suppliers or finance, collect the following:

• Monthly board volume, board types, and average placements per board
• First-pass yield and defect counts by defect type and detection point
• Repair-station logs showing real rework times
• Machine stop logs with causes, duration, and feeder or nozzle IDs
• Loaded labor rate and hourly line cost, with the basis stated
• RMA counts and the cost of each handling step
• Your machine models, feeder types, tape widths, and reel sizes, so suppliers can confirm compatibility
• ESD program requirements and moisture-sensitive component handling rules

With those inputs, a supplier can quote against a defined problem, and you can judge any proposal by the COPQ it is expected to remove.

Preguntas frecuentes

What is the formula for cost of poor quality in SMT?

COPQ equals internal failure costs plus external failure costs. Internal costs include scrap, rework, re-inspection, failure analysis, and quality-related line stops. External costs include returns, repair, freight, and penalties. Divide by boards, placements, or revenue to compare periods.

What is the difference between COPQ and cost of quality?

Cost of quality includes prevention, appraisal, internal failure, and external failure. COPQ covers only the two failure categories. Prevention and appraisal, such as training, SPI, AOI, and test, are spending intended to reduce COPQ.

How much does rework cost per board in SMT?

It depends on defect type, board complexity, and labor rate. Calculate repair minutes times loaded labor rate, plus materials, re-inspection, and retest. A touch-up on a chip component costs much less than BGA rework, so cost each defect type separately.

Should machine downtime be included in COPQ?

Include downtime when the cause is a quality problem, such as feeder jams, splice failures, or misloads. Multiply lost minutes by line cost per minute. Planned maintenance is prevention cost, and unrelated breakdowns can be tracked separately.

Is first-pass yield enough to measure SMT quality cost?

No. First-pass yield counts boards, not dollars, and does not show repair effort or escapes to customers. Use it as a trigger, then convert each defect type to cost per occurrence and add external failures.

How often should an SMT plant recalculate COPQ?

Monthly works for most plants, with a review after any change in product mix, equipment, supplier, or process. Use the same scope and denominator each time so that trends reflect real improvement and not changes in the method.