How to Choose a Wire Harness Manufacturer

Qualification compliance directly determines the bottom line for cooperation. I have dealt with a small-to-medium-sized manufacturing enterprise whose goods were detained by customs when exporting to North America because their partner lacked UL certification, resulting in a single-batch loss of over 800,000 RMB and a delivery delay of 3 months. Another medical device manufacturer had their product registration rejected by the medical products administration because the supplier failed to pass ISO 13485 certification, wasting nearly 2 million RMB in R&D costs.

Data shows that in the domestic wiring harness industry, only 65% of small and medium-sized enterprises hold complete industry certifications, whereas among the partners of leading enterprises, over 90% require the provision of original certificates and online verification.

Industry Certifications

I once chose a wiring harness factory offering 15% lower prices, only to have the first batch of goods returned when exporting to Europe because VDE testing revealed failure to meet high-temperature resistance standards. The entire batch of 8000 sets of wiring harnesses was sent back, costing 280,000 RMB just in return shipping and re-testing fees, and the client canceled 30% of the following year's order. When asked why they didn't check the certification, the reply was "the other party said having ISO 9001 was enough."

But ISO 9001 is just the foundation. Automotive wiring harnesses require IATF 16949, exports require VDE – these are the real thresholds. Let the data speak: Among domestic wiring harness enterprises, only 42% of Tier 1 automotive suppliers can fully provide valid IATF 16949 certificates; the remaining 58% either lack the certificate or have expired certificates. When medical device manufacturers select wiring harness suppliers, 90% will directly filter out those without ISO 13485 certification, even if their prices are lower.

General Quality System Certifications: Having them isn't enough; industry suitability matters

1.ISO 9001: The basic foundation, but insufficient on its own

ISO 9001 is the "elementary school diploma" of quality management systems, proving the enterprise has basic quality control processes. However, relying solely on ISO 9001 results in only 38% acceptance from automotive industry customers – because automotive wiring harnesses must withstand vibration, temperature differences, and voltage fluctuations, which ordinary quality systems don't adequately control. For example: a factory had ISO 9001 but not IATF 16949. When supplying an automotive customer, the height deviation of crimped terminals fluctuated significantly (±0.1mm), causing poor contact after 1000 sets were installed in vehicles, resulting in 120,000 RMB in rework costs. IATF 16949 is different; it mandates APQP and FMEA. Enterprises holding this certificate have an average mass production defect rate 35% lower than non-certified enterprises (2023 data from China Society of Automotive Engineers).

2.Medical industry requires ISO 13485; even one letter difference is unacceptable

Medical wiring harnesses may contact the human body or be implanted in devices, requiring infection prevention, aging resistance, and traceability. ISO 13485 specifically addresses this, mandating traceability for each harness's raw materials (e.g., insulation layer), production batch, and operating personnel. I know of a medical device company that chose a wiring harness factory claiming "good quality," but the supplier lacked ISO 13485. During product registration, the medical products administration found no traceability records and directly rejected the application, wasting the entire 1.8 million RMB initial R&D investment. This is not alarmist: 85% of Class III medical device manufacturers explicitly require wiring harness suppliers to have valid ISO 13485 certification. More critically, suppliers without this certification cannot enter the medical industry's "approved supplier list," even if their product performance meets standards.

Export Certifications: The "hard currency" of the customer's country; missing one can prevent customs clearance

1.European and American markets: UL and VDE are "entry tickets," not "bonus points"

The US market recognizes UL, Europe recognizes VDE. Taking UL as an example, wiring harnesses must pass the UL 1581 standard, including tests for temperature resistance, flame retardancy, and aging resistance. I dealt with a wiring harness factory exporting to the US without UL 1581 certification. The goods were detained by customs at the Port of Los Angeles. The customer urgently needed the goods, so they had to spend 350,000 RMB on expedited third-party testing and paid the customer a 5% penalty. VDE is stricter, also testing EMC. Harnesses without VDE certification used in German industrial equipment might interfere with PLC control systems, causing entire production lines to shut down. The data is clearer: Wiring harness factories with UL/VDE certification have 62% lower customer complaint rates and 40% higher repurchase rates for exports to Europe and America compared to uncertified enterprises (2023 Customs General Administration Export Quality Report).

2.Asian markets: "Smaller certificates" like PSE, KC must also be obtained

Japan requires PSE, South Korea requires KC. Last year, a harness factory received an order from a Korean automaker but hadn't obtained KC certification in advance. During sample testing, insulation resistance was found substandard; re-sampling delayed the process by 2 months, and the customer switched to a competitor. Note that KC certification is a mandatory item during initial factory audits by Korean automakers for new suppliers; without it, you can't even enter the workshop. Southeast Asian markets seem to have lower barriers, but countries like Indonesia and Vietnam are promoting local certifications. Harnesses without local certification may face 10%-15% additional tariffs at customs – money that could cover profits for three more batches.

Special Industry Qualifications: The "invisible thresholds" for military, new energy; you really can't enter without them

1.Military wiring harnesses: Confidentiality certificate is the "lifeline," Level 2 is the starting point

Military harnesses are used in radar, missiles, armored vehicles, requiring anti-eavesdropping, impact resistance, and tolerance for extreme environments. To undertake military orders, you must first have the "Confidentiality Qualification Certificate for Weapons and Equipment Scientific Research and Production Units," with Level 2 being the minimum. I asked a military supplier: Private enterprises without a Level 2 confidentiality certificate cannot even enter the military procurement network's supplier database, disqualifying them from bidding.

2.New energy high-voltage wiring harnesses: UL 2594 is the "dividing line"; certified companies can earn 30% more

New energy vehicle high-voltage harnesses must withstand high voltage, prevent short circuits, and resist high temperatures. UL 2594 specifically tests these: arc resistance, temperature resistance, flame retardancy. Harness factories without UL 2594 certification cannot enter the supplier lists of companies like Tesla or NIO. One factory specializing in new energy harnesses spent 800,000 RMB to obtain UL 2594, leading to a direct 40% increase in order volume and a unit price increase from 80 RMB/set to 105 RMB/set – the extra profit fully covered the certification cost. More realistically, OEMs have a sampling inspection rate for high-voltage harnesses 3 times that of ordinary harnesses. Non-certified products have a failure rate of up to 22% during sampling; one rework can wipe out half a year's profit.

Business License and Permits

To save 12% on costs, he chose a small wiring harness factory in Jiangsu – the salesperson promised "we can produce harnesses, the license includes 'wires and cables'." After placing an order for 3000 sets of industrial robot wiring harnesses, market regulation authorities conducted a surprise inspection. Checking the business license, they found the "Scope of Business" only included "Wire and Cable Manufacturing," with no mention of "Wiring Harness Production".

The production line was sealed. The batch couldn't be delivered on time. The customer deducted 180,000 RMB in penalties per the contract, and they had to compensate an additional 150,000 RMB for prepared materials, totaling a loss of 330,000 RMB. He told me, "How was I to know? I thought 'wires and cables' covered harnesses" – but according to commercial regulations, "wires and cables" refers to bare wire processing, while "wiring harness" involves finished product assembly; they are completely different things.

More revealing data: In 2023, nationwide market supervision systems investigated over 1200 cases of unlicensed or ultra vires wiring harness operations, accounting for 18% of industrial product violation cases, with an average fine of 96,000 RMB per case, not including customer compensation.

Business License: Don't just look at the "header," scrutinize every word in the "Business Scope"

The former can handle entire orders from design to production, the latter can only handle the final assembly stage. I've seen the worst case: A new energy vehicle company chose a factory whose license stated "Wiring Harness Assembly." It turned out the factory used semi-finished harnesses purchased externally without secondary inspection. After installation, 100 sets experienced short circuits. The automaker recalled 500 vehicles, directly losing 20 million RMB, and had to pursue the supplier's joint liability.

More critically, a wiring harness factory operating ultra vires cannot issue formal invoices correctly – you pay, but they can only issue invoices for "Wires and Cables." During tax audits, this constitutes "issuing false invoices," leading to tax recovery, late fees, or even transfer to public security. In 2022, in the Yangtze River Delta region, 32% of wiring harness contract disputes originated from suppliers operating ultra vires, 80% of which involved "Wire and Cable" factories expanding into "Wiring Harness" production and being sued by customers, requiring not only refunds but also 3x the penalty.

Special Permits: "Small certificates" like explosion-proof, export permits; lacking them is "illegal," not just "troublesome"

Not all harnesses require extra permits, but those involving hazardous scenarios or crossing borders must have them. For example, explosion-proof harnesses for chemical equipment require the "Production License for Explosion-proof Electrical Equipment" – this certificate tests the harness's explosion-proof enclosure, insulation class, temperature group. I have a client in chemical equipment who previously chose a cheaper factory without the Ex certificate. When the harness was used on a reactor, the insulation aged and short-circuited, causing a minor explosion. Fortunately, there were no casualties, but the safety supervision department fined them 500,000 RMB and mandated a 1-month operational suspension.

Another example: The EU requires the Low Voltage Directive under CE marking. It's not about just sticking a CE label – you need the number of a Notified Body, proving compliance with standards like EN 60529, EN 60335. Last year, a harness factory received an order for shelf wiring harnesses from a German supermarket. Without the LVD certification, the goods were detained at Rotterdam port. The customer demanded expedited third-party testing, costing 220,000 RMB, plus a 5% penalty. Remember, the EU, for products without CE marking, can impose fines up to 50% of the goods' value besides detention.

Permit Expiry: Not a matter of "renewal," but a signal to "cease production immediately"

There's a more hidden pitfall: Permits expiring without renewal. For example, a factory had the "Production License for Explosion-proof Electrical Equipment," but it expired in 2023 and wasn't renewed. The client didn't check. This year, safety inspectors conducted a spot check and sealed the production line. The client urgently needed the goods, forcing them to find a subcontractor. The subcontracting fee was 20% higher than the original price, causing a 2-week delay. The client deducted 10% of the payment, resulting in a total loss of 800,000 RMB.

More alarming data: Annually, 15% of wiring harness enterprises have cooperation terminated by clients due to expired permits, with an average loss of 40% of monthly sales – because you can't prove you are producing compliantly, and clients won't take the risk. I know a military harness factory whose Level 2 confidentiality certificate expired and wasn't renewed for 3 months. The military directly terminated their qualification. A potential 30 million RMB project was lost, and they were blacklisted for 5 years from military orders.

Technology and Design Capability

A new energy vehicle's high-voltage wiring harness total length might exceed 5 kilometers, involving over 2000 connection points; an industrial robot's control harness needs to operate stably for 10 years in environments from -40°C to 85°C, with vibrations from 10-2000 Hz.

Internal data from a leading automaker shows that 30% of early after-sales faults are directly attributable to wiring harness design defects.

A manufacturer's design capability must penetrate scenario requirements: understanding both the "physical constraints" of mechanical space and calculating the "numbers" of electrical performance. Choosing a manufacturer with the right design capability can directly reduce project rework rates from 15% to below 3%.

R&D Team

When wiring harness design has problems, 80% of the blame lies with the R&D team – this is not an exaggeration. A third-party statistic shows that projects independently completed by harness designers with less than 5 years of experience have an average late-stage rework rate of 18%; projects led by engineers with over 10 years of experience, especially from OEMs or Tier 1s, can control the rework rate below 3%.

A real case: A new automaker initially engaged a young team to design battery pack harnesses. The engineers lacked experience with 400V high-voltage systems and underestimated the creepage distance between busbars and connectors. After 3 months of mass production, 3 short-circuit incidents occurred, costing 20 million RMB in recall expenses.

Engineers from OEMs/Tier 1s are the "main force for risk mitigation"

Don't be fooled by "our team has 10 engineers"; ask how many came from OEM chassis/body electronic control departments or international Tier 1s. In top harness factories we've collaborated with, 60% of the design team engineers have 8-15 years of OEM experience. For instance, one lead engineer previously responsible for main harness development for 3 fuel vehicle models at FAW knows that "engine bay harnesses must avoid the heat radiation zone of the turbocharger, requiring XLPE insulation." Their protégés are also capable – 30% of the team are senior assistants with over 5 years of experience, able to independently complete 80% of conventional vehicle harness designs.

DFMEA is not about filling forms; it's about "marking pitfalls in red" beforehand

A good team does DFMEA at the "millimeter level." For a commercial vehicle harness, regarding the risk of "chassis harness being pierced by stones," they specified not just material but protective sleeve thickness ≥2mm and installation hole positions avoiding high-impact areas. Furthermore, they link DFMEA with actual road test data – after a light truck harness ran 200,000 km, they found accelerated aging near the exhaust, promptly adding a new requirement to the DFMEA. Their DFMEA covers over 92% of potential risk points, compared to an industry average of 75%.

Not "one person responsible for one model," but "the whole team deeply understands one scenario"

A common mistake of novice teams is engineers working in silos. OEM-experienced engineers know harness design is a system engineering. For a new energy vehicle high-voltage harness, the lead designer involves thermal management and EMC engineers in meetings. The thermal engineer provides temperature data dictating wire insulation choice; the EMC engineer dictates shielding requirements. The most rigorous manufacturer we've seen forms a dedicated team of 5-8 people for a vehicle project, holding 3 cross-departmental meetings per week. Just the waterproof structure for the charging port harness was discussed over 4 rounds until it passed IP67 immersion tests. This "whole team focuses on one scenario" approach reduces their design error rate by 60% compared to the industry.

Newcomer training depends on "how many design changes they've followed," not "how much software they've learned"

Good teams have newcomers follow at least 3 design change projects from start to finish: first, disassembling old harnesses to note failure points; second, modifying designs under guidance; third, observing road tests to validate fixes. One manufacturer's training manual states: "After 3 change projects, you should be able to state blindfolded that 'the bend radius at the A-pillar must be ≥10x the wire diameter, otherwise it will break within six months'." Under this "hands-on mentoring" model, harnesses independently designed by newcomers within 1 year have a rework rate below 5%, compared to 20% for those without such training.

Customization Depth

90% of manufacturers claim they can customize, but less than 30% truly deliver. Data from an automotive parts procurement platform shows that custom harness projects without data validation have a mass production design change rate as high as 45%, whereas projects with complete testing have only an 8% change rate.

A medical device manufacturer chose a supplier claiming "can produce sterile harnesses" but didn't request test data. After mass production, 10% of samples had pull-off force at crimps of only 45N (standard 50N). The entire batch was scrapped, losing 1.2 million RMB.

Customers don't want "can do," but "have done similar and know where it might fail"

Quality manufacturers we've engaged with match each custom requirement with at least 3 similar past cases, each accompanied by complete test reports. For a robot customer requiring cable carrier harnesses with a bending life ≥10 million cycles, the manufacturer immediately referenced a similar project for an automotive welding line: initial PVC insulation cracked at 5 million cycles; after switching to TPU, it passed 12 million cycles.

Every requirement must be tested to yield "specific values," not "close enough"

"Anti-interference" isn't a slogan; it requires measuring radiated emission values from 100MHz-1GHz. "High temperature resistance" requires specifying the long-term operating temperature of the wire insulation and tensile strength retention after aging. For a battery pack harness, the customer required terminals not to become brittle at -40°C. The manufacturer conducted pull-off tests at -40°C: ordinary terminals failed; switching to PA66+GF30 material met the ≥60N standard.

Sample testing is not a formality; it's about turning "potential errors" into "certainty"

Providing samples isn't the end; check how many "mass production risk points" the sample testing covers. Reliable manufacturers conduct at least 3 rounds: functional, environmental, and limit testing. For an industrial control harness, limit testing revealed chafing leading to short circuits under specific vibration; the routing was modified accordingly.

Data must be "traceable to the source," not just "tested"

The most rigorous manufacturers can explain the "origin and journey" of test data. For an IP67 test report, you can find not just the pass result but details like the specific seal model, adhesive thickness, and curing time used.

Sample Capability

Sample testing isn't about "all pass"; it's about "scrutinizing every parameter to the decimal point"

Don't trust verbal assurances; examine the specific numerical deviations in the test report. For a battery pack harness pilot run, tests included:

• Crimp Height: Standard ±0.05mm, first sample ±0.03mm (excellent), but second sample measured ±0.07mm (out of spec).
• Pull Test: Standard ≥80N, first sample avg. 85N, but third sample measured 78N (near limit).
• Wire Sequence Check: Using optical scanning, first sample 0% error, but fifth sample found one mislabeled wire.

Pilot production must "simulate mass production," not "make a few separately"

Some manufacturers fool with "handmade samples" – using old equipment, skilled workers, inventory materials. Looks good, but fails upon mass production automation. Reliable manufacturers use pilot lines identical to mass production lines: same automatic cutting machines, same servo crimpers, same MES tracking. One manufacturer's pilot process:

• Quantity: 50 sets (50% of the minimum mass production batch of 100 sets).
• Equipment: Shares 3 automatic crimpers, 2 electrical testers with the mass production line.
• Personnel: Assigned skilled workers from the mass production line, not a temporary "sample team."

Result: The pilot run revealed that the crimper speed was too high, causing a deformation rate increase after 30 sets.

Samples aren't "done once delivered"; they must allow tracing back to production parameters. Each sample has an "ID":

With this table, if a crimp height anomaly occurs during mass production, you can immediately trace which wire batch, which machine, which operator was involved, reducing problem identification from 3 days to 2 hours. One client used this to identify an out-of-spec insulation thickness within 3 days, avoiding batch rework.

Sample issues must be "corrected thoroughly," not "close enough"

Pilot runs don't end with identifying problems; assess the manufacturer's corrective action intensity. For a harness failing vibration due to wire chafing, correction wasn't just "adding a sleeve":

• Step 1: Analyze wear location, confirm routing too close to metal part.
• Step 2: Redesign, increase bend radius.
• Step 3: Remake 3 samples, test for 100 hours (double the standard), confirm no wear.
• Step 4: Update DFMEA, incorporate improvement into design specs for similar projects.

We've统计ed, manufacturers implementing thorough corrections have a ≤2% recurrence rate for similar issues; those with perfunctory corrections have ≥30% recurrence.

Strict Quality Control System

J.D. Power's 2022 Global Vehicle Dependability Report indicated that wiring harness faults account for 18% of vehicle recall causes. A single large-scale recall can cost up to 1.5 billion RMB.

A Tier 1 supplier we previously worked with faced a 2.2 million RMB loss plus a 5% penalty because a terminal crimp height exceeded tolerance by 0.15mm, causing an 8-hour client production line stoppage.

Raw Materials

In a 2023 quality review for a new energy vehicle plant, we found that 41% of wiring harness faults in the past 12 months were rooted in raw materials: wires labeled 125°C whose insulation softened at 105°C; connectors with specified insertion force of 50N measuring only 42N, leading to loosening. One supplier's terminal tin plating was specified at 5μm, but sampling found an average of 3.2μm.

In our own raw material warehouse, every spool of wire, every connector, every box of terminals comes with 3 test reports + 5 sets of dynamic data. This isn't for inspection compliance; it's genuine fear that one unchecked value leads to downstream line stoppages and customer claims.

Wires: Temperature and voltage resistance aren't slogans; every spool undergoes 3 rounds of testing

Wires constitute 40% of harness cost, but their issues are most hidden – insulation aging, core breakage. We subject each wire batch to 3 rounds of "destructive testing":

First, Temperature Resistance Test. For automotive-grade wires, beyond the -40°C~125°C cycle standard, we intensify to -50°C~150°C, baking for 168 hours. Post-test tensile strength retention must be ≥85%. Last month, a domestic wire batch measured 78% after baking; though not scrap level, we rejected it – needing a safety margin for battery pack temperatures reaching 130°C.

Second, Continuity Test. Using Fluke DSX-5000, testing 300 lines per minute, focusing on hidden opens and insulation resistance. Hidden opens are insidious. One rejected batch had 10 wires with insulation resistance of only 50MΩ at 200V (standard ≥100MΩ).

Third, Mechanical Properties. Testing elongation at break ≥25%. For new energy vehicle harnesses requiring frequent bending, low elongation leads to breakage. We once received Japanese wires with 22% elongation (meeting but near lower limit); we worked with the supplier to improve the formula to a stable 27%.

Connectors: Insertion/Extraction Force and Contact Resistance measure "unbreakability"

Connectors are harness "joints," prone to issues. We test connectors, not by sampling, but with full inspection items:

First, Insertion/Extraction Force. Using MTS CTS-2000, insertion force 40-60N, extraction force ≥50N. Variation must be ≤5N. Why? Lower force leads to loosening, increased contact resistance, heating. We once rejected Molex connectors measuring 42N extraction force; the client later reported 15% loosening after 3 months, costing 800,000 RMB rework.

Second, Contact Resistance. Using German HAEFELY micro-ohmmeter, each terminal measured 3 times, average must be ≤1mΩ. High resistance causes heating. For new energy high-voltage harnesses, we require ≤0.5mΩ. A batch of Aptiv connectors measured 1.2mΩ; we insisted on mold change, new batch stabilized at 0.4mΩ.

Waterproof Test. IP67 isn't just immersion; we test 2 extra hours + measure leakage current (must be ≤0.5mA). High current corrodes terminals. A TE connector batch showed 0.7mA leakage; disassembly revealed insufficient seal compression. After correction, leakage dropped to 0.3mA.

Terminals: 1μm plating thickness difference means 3x lifespan difference

We test terminals using X-ray fluorescence thickness gauges, sampling 20 per batch, measuring 3 points each, taking the minimum:

Automotive tin-plated copper terminals require plating thickness ≥5μm. Thin plating oxidizes under salt spray/heat, increasing contact resistance. A domestic batch averaged 4.1μm; after rejection, the supplier increased plating time, now stable at 5.5μm.

Industrial gold-plated terminals require thickness ≥1μm. Gold is expensive but durable. For a 5-year requirement in 95% humidity, 1.2μm gold plating showed minimal resistance increase after 5 years, while 0.8μm increased significantly.

Also, Terminal Crimp Zone Hardness, measured by microhardness tester, must be HV≥80. Low hardness causes deformation during crimping, reducing contact area. A batch with HV72 showed 30% higher contact resistance; after heat treatment adjustment, HV stabilized at 85.

Auxiliary Materials: Tapes, Ties – these "small items" also have strict numbers

Often overlooked, but loose tape abrades wires, broken ties cause harness failure. Our auxiliary material inspection is equally strict:

• Insulation Tape: Temp resistance 130°C, tensile strength ≥15N/cm, peel adhesion ≥10N per segment. A batch with only 8N peel strength came loose from vibration, abrading wires and causing a short circuit. The supplier was blacklisted.
• Nylon Cable Ties: Tensile strength ≥200N, temp -40°C~85°C. For new energy vehicle chassis ties, we require ≥300N. A batch measuring 250N was rejected – chassis vibration is high; broken ties cause harness sagging, contacting the exhaust pipe.

Production Process

Last year, during consultation at an automotive harness factory, we found their final assembly line had 120 monthly reworks due to "harness routing deviation" – the problem originated in pre-assembly, where workers fixed harnesses by feel, leading to 2mm misalignment. A more serious issue was crimping: occasional 0.1mm height deviation by experienced workers caused an 8-hour client line stoppage, losing 2.2 million RMB. In our workshop, crimpers self-check every 10 terminals, pre-assembly fixture positioning accuracy is ±0.1mm, welding force variation ≤1N. Machines turn "feel" into "scale," "occasional mistakes" into "almost no mistakes."

Crimping Process: Machines are 10x more stable than hands, tolerance held to 0.05mm

We use Komax automatic crimpers with crimp height tolerance ±0.05mm. Machines have 3 sensors: pressure, displacement, vision. Every 10 crimps, the machine pauses 0.5s for laser height measurement. For a client requiring 5.20mm±0.05mm, our machines didn't deviate once in 3 months, reducing their line defect rate from 1.2% to 0.1%.

We are meticulous about crimp dies: each die is 3D scanned for wear; wear exceeding 0.02mm means scrap. One die used 50,000 times had 0.03mm wear, producing terminals at 5.28mm height; we replaced it immediately. Die set readiness is 100%, changeover time reduced from 45 to 10 minutes.

Pre-assembly: Fixture positioning accurate to 0.1mm, eliminating final assembly rework

We use 3D positioning fixtures with CNC-machined locator pins and wire channels, repeatability ±0.1mm. For a commercial vehicle harness with 12 wires in one channel, the channel width allows 0.2mm clearance, but pins control lateral deviation within 1.5mm. Workers just place wires in the channel; a clip fixes the position. Rework rate dropped from 5% (manual) to 0.3%.

For complex harnesses, we use laser projection assistance to project routing and bend radii onto the fixture. For a 90° bend, laser lines specify radius ≥8mm; workers bend accordingly, measuring post-bend radius with a tube bender, error ≤0.5mm.

Welding Process: Parameters tuned to micrometer level, force stable above 9N

Aluminum wire welding is delicate. We use ultrasonic welders with parameters precise to 0.1mm amplitude, 50Hz frequency. Post-weld, we test 3 items: ① Weld pull force ≥8N, ② Weld nugget diameter ≥1.2mm (50x microscope), ③ Internal voids (X-ray, 5μm resolution). A batch with 6N force was traced to low amplitude (180μm vs. standard 200μm); after adjustment, force stabilized above 9N.

Copper wire welding is stricter: Weld resistance ≤50mΩ. For new energy high-voltage harnesses, we require ≤30mΩ. Using high-frequency welders at 300kHz, 1.5MPa pressure, resistance stabilizes at 25mΩ.

Bundling & Securing: Tie tightness, tape wrapping – all by the numbers

Harness bundling is often overlooked but crucial. We use pneumatic tie-guns set to 250N force (ties rated ≥300N break force). Tightness is measured with a force gauge, maintaining 80-100N (too loose risks detachment, too tight cuts wires). Tape wrapping uses automatic machines, overlap ≥30%, thickness ≥0.5mm per cm. A batch with 20% overlap came loose during vibration testing, abrading wires; increasing overlap to 35% solved the issue.

Abnormality Response: Quick stop, accurate troubleshooting, preventing error accumulation

The production line has an ANDON system. Any process issue, the worker presses a button, machine stops within 3 seconds. For a crimping deviation, the system auto-alarms, isolates the last 50 harnesses, uses AOI vision (10μm resolution) to scan terminals, locating the issue (e.g., die wear 0.03mm) in 5 minutes, replaced with a spare die in 10 minutes – average downtime ≤15 minutes vs. 2 hours previously.

We also have a real-time data dashboard updating hourly: crimp height, weld force, tie tension from each machine. Last week, a crimper pressure spike to 900N triggered an alarm; logs revealed air pressure fluctuation; stabilizing pressure brought it back below 800N.

Finished Product Testing