Copyright © 2020-2021 Shenzhen CDTech Electronics LTD. All rights reserved. Site Map Powered by iwonder.cn
PPAP arrives in industrial and medical display programmes by inheritance rather than requirement. A quality engineer moves across from automotive, or a customer's supplier manual carries the eighteen-element package over from another product line, and a custom display module is asked for DFMEA, capability studies and an appearance approval report. The module is often built around a panel the supplier did not design, at volumes far from automotive. The result is a package that consumes weeks and changes nothing on the line. This article separates the elements that do work outside automotive from the ones that are ceremony, and sets out a reduced package a quality team can defend.
The automotive PPAP manual assumes a particular production model: a high-volume part, a customer who owns the design record, a tool dedicated to that part, and requalification for the life of the vehicle programme. The elements prove that a process can repeat a design the customer defined, at a rate the customer will buy, from a tool the customer paid for. Non-automotive display programmes break most of those assumptions at once.
The panel is usually a catalogue component from a display maker, so the design record is the supplier's drawing rather than the OEM's. Volumes are lower, tooling is often shared, and the governing quality system is ISO 9001 or, for medical devices, ISO 13485. None of that makes the discipline worthless: it means elements are selected by function rather than submitted by index. The AIAG manuals remain the reference for what each element contains, and the automotive package is set out in PPAP requirements for automotive TFT LCD displays.
Holding every element to the same standard produces a file nobody reads. The test that matters is whether an element changes a decision on the line, at incoming inspection or in a change review. Five do that on a display programme.
| Element | What it prevents | How to right-size it |
|---|---|---|
| Control plan | Undocumented process drift | Keep in full, written for the line that runs |
| MSA, gauge or attribute | Inspection that cannot repeat | Study the two checks that gate shipment |
| Dimensional report | Fit failures at the chassis | Report to the module outline, not the panel |
| Material and performance evidence | Unverified claims to a customer | To the declared specification, method named |
| Capability on the key characteristic | A process that passed once | One characteristic, not a wall of indices |
Everything else earns its place only where the programme owns a design decision or a cosmetic grade. A DFMEA on a module assembled from a bought-in panel and driver board describes the supplier's assembly, not the design, and a short risk table often carries the same information.
Most PPAP documents are evidence about a moment: this sample measured this value on this date. The control plan is different. It is the only document in the package that a production supervisor, an inspector and a second-source engineer will open next year, because it describes what happens at each step rather than what happened once.
A display control plan is short and specific. It names the operation, the characteristic, the specification with tolerance, the evaluation method, the sample size and frequency, the control method, and the reaction plan when a result falls outside. On a bonded module that means incoming panel inspection, bonding, alignment and cure, optical and electrical test, and packing.
Each row needs a reaction plan an operator can act on without finding an engineer. Notify supervisor is a delay with a signature; quarantine the last hour of output is a reaction plan.
Yes, but as an attribute study, not a gauge R&R.
Cosmetic inspection produces attribute data, so the study asks whether two inspectors agree on the same panel, not whether an instrument repeats to a micron.
Attribute agreement analysis is the right tool: several inspectors, panels containing known good, known bad and borderline units, and repeated rounds so each inspector sees the same panel twice without knowing it. The output is agreement within an inspector, between inspectors and against the reference, which is why the reference panels must be graded by someone whose judgement the programme trusts.
Conditions matter as much as the statistics. Viewing distance, illumination, background colour, inspection time and panel angle all change the answer, so the study should fix them and the control plan repeat them. For luminance, chromaticity and contrast, MSA is a variable-data exercise that belongs with the instrument, where gauge R&R exposes a worn meter.
The dimensional results element assumes the customer's drawing defines the part. On a display module the opposite is often true: the panel comes from a display maker with its own drawing, and the OEM's specification governs the assembly around it. Verifying the panel's internal dimensions verifies someone else's design record.
What the report should cover is everything the OEM's assembly depends on: outline dimensions and tolerances, mounting hole positions, active area position relative to the bezel, overall thickness including cover glass, flex tail length and bend radius, connector position and orientation.
Report against the revision the samples were built to, since a later revision proves nothing about the parts shipped, and keep the measurement method with the numbers: a profile projector, a vision system and a caliper produce different confidence at the same tolerance.
Level two is the useful shape for a non-automotive programme: samples and a warrant submitted to the customer, supporting data retained at the supplier. It keeps evidence where it can be produced and the submission to documents a customer will read.
Ten items is not a shortened automotive package. It is a package built for a programme where design record, tool and volume sit differently, and the quality and reliability guide covers how the evidence is produced.
Waivers are where reduced packages fail, and not because the wrong elements were dropped. They fail because the drop was never written down, so an auditor finds elements missing and cannot tell a decision from an omission.
A waiver should name the element, the part number and drawing revision, the reason, and any substitute evidence. DFMEA is the easiest to waive on a module the supplier did not design, provided the design record and the panel maker's qualification are identified. Appearance approval drops with no cosmetic grade, and the bulk material checklist drops where the part is not a bulk material. Annual layout requalification needs care: the cadence may be unnecessary, the change triggers are not.
Who signs is a quality decision rather than a purchasing one. The customer's quality function signs, the supplier's quality manager countersigns, and the waiver is filed with the PSW so the two are read together. A waiver agreed by email and absent from the file is treated as missing. Product documentation duties for a display sold in the EU sit under the General Product Safety Regulation.
First article inspection and PPAP are often used as synonyms, and the confusion costs time. FAI verifies that a first part matches its drawing: every dimension, every note, every material callout. PPAP verifies that the process can repeat that part, which is why the package carries a control plan, a capability result and a warrant rather than a measurement sheet.
The golden sample sits between them: a unit signed by both parties and retained as the reference for cosmetic and functional judgement where a drawing cannot express the difference between acceptable and not. For a display that means a unit at a defined brightness and inspection condition, because appearance shifts with age.
Three rules keep it useful. Label it with part number, revision, date and the settings it was approved at. Store it where temperature and light do not change it. Replace it deliberately when a change is approved, retiring the old unit. Document control is the mechanism, and the ISO standards portal sets out the requirement in outline; which certificates a supplier holds is described on the quality certifications page.
The package describes a process at a moment. Anything that changes the process makes part of it stale, and the practical question is which part. A tool refurbishment or move, a change of panel source, driver IC, backlight, adhesive or flex supplier, a new laminator or a relocated line all qualify.
A full re-run is rarely necessary. A delta submission is usually enough: the elements the change touches, plus the control plan revision that records it. Moving a tool changes dimensions and fixtures, so the dimensional report and first article return. Changing an adhesive changes cure and bond strength, so material evidence returns. A panel source change alters almost everything downstream of incoming inspection, changes power consumption under the EU Ecodesign requirements for displays, and changes the declaration under RoHS.
The clause that makes this work is a notification duty with named triggers, agreed while the programme is healthy. Without it, the first news of a material change is usually the shipment that contains it. The custom sourcing guide covers where that clause sits among the terms of a display order.
If a reduced PPAP package is being argued over, start from the control plan and work outwards. CDTech can supply the process evidence a level two submission needs for a custom display module and confirm which elements apply. Write to [email protected] or use the contact page.
Which PPAP level is normal for industrial displays?
Level two or level three in most programmes, with level two the better fit where the customer wants a warrant and samples but does not want to receive the full evidence file. Level three adds the supporting data to the submission, which is heavier but sometimes demanded by a customer's supplier manual. Level one, a warrant only, is defensible for a catalogue module with no custom tooling. The level should be named in the purchase agreement rather than assumed, because it decides what is submitted and what is merely retained.
Is MSA relevant to a visual inspection?
Relevant, but as an attribute agreement study rather than a gauge repeatability exercise. You are testing whether inspectors classify the same panel the same way, so the study needs known good, known bad and borderline units, repeated rounds, and fixed viewing conditions. It will usually expose a gap in limit samples rather than a gap in eyesight, which is the useful outcome. For brightness, colour and contrast, run a conventional gauge study on the instrument instead.
Can PPAP be done once per family?
Often yes, if the family shares a panel, a process route and a control plan, and only the outline or the interface differs. The submission then covers the worst case in the family and lists the variants by part number, with the differences documented. It stops being valid when a variant adds a step the control plan does not cover, such as bonding or cover glass, or when the panel source changes. Family logic is a reasonable simplification as long as the boundaries are written down.
How does PPAP relate to first article inspection?
First article inspection is one input to a PPAP submission, not a substitute for it. FAI confirms that a first part matches its drawing, dimension by dimension. PPAP confirms that the process can repeat that part, which is why it carries a control plan, a capability or yield result and a signed warrant alongside the dimensional data. A programme that does FAI only has evidence about one part. A programme that does both has evidence about a process, which is what a customer is buying.
By continuing to use the site you agree to our privacy policy Terms and Conditions.