Note: General engineering and procurement information, not insulation-coordination design advice. Clearance and creepage for a given board are set by the design authority against the standard that product is certified to. Confirm any dimension with the engineer who owns the drawing.
A customer specification arrives with a part number on it and "CTI 600 V" underneath. Is that line negotiable? It depends on a document the buyer has usually never seen.
Three branches run from that document, and they give opposite answers for the same material at the same voltage. If an automotive OEM standard governs the board, CTI 600 is not a negotiating position: the OEM's own insulation-coordination table is written on the premise of material group I. If the board is dimensioned from Annex F of IEC 60664-1 at a clean pollution degree 2 (a board’s expected contamination class under IEC 60664-1), the printed-wiring columns already give it group I spacing, so CTI 600 buys nothing further. If the board sees pollution degree 3, there is no printed-wiring column at all and the material group decides directly.
One correction first, because it is the commonest confusion we meet. IEC 60112, the method behind the comparative tracking index, states in its scope that it is designed for test voltages up to 600 V AC. CTI 600 is not a material keeping pace with an 800 V battery system. It is the top of a scale that ends at 600, and asking for "CTI 800" is asking for a mark that does not exist on the ruler.
This article is about the specification. FR-4 is the most common grade of copper-clad laminate (CCL), and high-CTI FR-4 is also harder to buy in 2026, because capacity in standard glass cloth and circuit-grade copper foil moved to higher-margin AI-server material and, by trade-press accounts, some suppliers introduced allocation quotas; that story has its own numbers and we treat it separately.
This article works through six specification traps and ends with a checklist you can drop into an RFQ.
What CTI Is, and What It Is Not
IEC 60112, now in its sixth edition (2025), drips 50 drops of electrolyte between two chisel electrodes set 4 mm apart and reports the voltage the material survives. The reference electrolyte, solution A, is 0.1% by mass ammonium chloride. The scope caps the method at 600 V AC. Two scales are built on those values: UL's performance level category, which is how the number reaches a datasheet, and the material groups of IEC 60664-1 clause 4.8.1.3, which is how it reaches a creepage table.
Read the right-hand columns against each other and the first trap appears: group IIIa straddles PLC 2 and PLC 3. Neither scale can be derived from the other, so ask for the raw CTI value alongside whichever class a supplier quotes. The second trap is in clause 4.8.1.3 itself, which defines the groups on values measured using solution A. A number from the contaminated solution B — the M-suffixed value measured with solution B — is not the value the groups are written against.
The standards are frank about what the index is worth. Clause 4.8.1.2 says a direct correlation between the material groups and actual performance "is not practical", only that materials ranking higher in relative performance rank in approximately the same order by CTI; clause 4.8.1.4 calls the test a qualitative comparison under test conditions. The standard's own wording is the warning. A CTI value ranks materials under one defined set of test conditions — platinum electrodes, one electrolyte, one spacing, 50 drops — and it is a position on that ranking, not a voltage the finished board will withstand. Treat it as the index it is named as.
Hence the confusion behind much of the over-specification here. CTI 600 V is a material rating, the top band of a tracking test. 400 V and 800 V are system working voltages, what the DC link runs at. The only connection between them runs through creepage distance, and that connection does not hold on every table.
Which Table Decides Whether CTI 600 Buys Anything
The shorthand is that a higher CTI lets conductors sit closer together, so CTI 600 keeps a high-voltage board small. On the table usually cited for it, it does not.
We read IEC 60664-1:2007, Table F.4, from the standard itself. It carries two sets of columns: one for printed wiring material, one general and split by material group, which that table prints as I, II and III. The printed-wiring heading reads "All material groups" at pollution degree 1 and "All material groups, except IIIb" at pollution degree 2.
Ordinary FR-4 is normally group IIIa, which falls inside that group III column; the finer split into IIIa and IIIb lives in the clause 4.8.1.3 definitions, not in the table's headings, and note 2 says only that group IIIb is not recommended in pollution degree 3 above 630 V.
Two things fall out. In pollution degrees 1 and 2 the printed-wiring column does not ask what material group the laminate is in — the only exclusion is group IIIb, CTI 100 to below 175, a threshold every FR-4 datasheet we read clears. And the printed-wiring pollution-degree-2 values are identical to the group I values. A board dimensioned on that column already has group I spacing. CTI 600 cannot shorten it further.
The arithmetic confirms that rather than leaving it to the eye. Above 400 V creepage is linear in voltage — first order, not squared — with pollution-degree-2 slopes of 0.005 mm/V for group I, 0.007 for group II and 0.010 for group III. Check: 2.0/400 = 4.0/800 = 5.0/1,000 = 0.005, and the printed-wiring pollution-degree-2 column carries the same slope. (Below 400 V the relationship is not linear.) IPC arrives at the same figure independently — IPC-2221B, as tabulated by a semiconductor vendor's application note, Table 6-1, category B2, external conductors uncoated, gives 2.50 mm plus 0.005 mm/V above 500 V, so 4.00 mm at 800 V.
Two qualifications matter more than the tidy conclusion.
Edition. We read the 2007 edition, where this is Table F.4. IEC 60664-1:2020 renumbers it Table F.5, adds 1,500 V DC rows and an altitude-correction table (F.10); consolidated edition 3.1 (May 2025) folds in Amendment 1. We did not confirm that the row values above are unchanged in Table F.5 — the current edition is paid and we did not obtain it. Cite the edition and table number whenever you use these numbers.
Text and practice diverge, and some published figures are wrong. A semiconductor vendor's application note describes Table F.4 as defining creepage by pollution degree and "material group of printed circuit board", and a second note from the same vendor prints an extract without the printed-wiring columns even though its own text describes them. We take the standard text as controlling, but many practising designers do dimension from the group-dependent columns — which affects what your customer and your supplier each assume. Further down the quality scale, one widely syndicated PCB blog publishes a creepage-by-voltage table roughly six times smaller than the IEC minimum, apparently lifted from IPC-2221 Table 6-1's internal-conductor column rather than the external uncoated one. We do not reproduce it: a wrong number is harder to retract than to repeat. The same write-up calls IPC-2221C values proposals with updates "expected in Q3 2026". IPC-2221C was published in December 2023 and superseded IPC-2221B; its table of contents, which we obtained, carries clause 6.3.3 on creepage, 6.3.4 on CAF growth, 6.3.5 on the comparative tracking index, Tables 6-1 and 6-2, and clause 4.3.1 on high-Tg laminates. We did not obtain the body and quote no value from it.
Where a CTI 600 Requirement Actually Comes From
None of that makes a CTI 600 specification overspecified. It means the justification is not in the printed-wiring column, and a buyer looking for an equivalent has to find out which document the number came from.
Not every standard grants the printed-wiring exemption. UL's published engineering Q&A on IEC 62368-1 answers whether to use "Other Materials" for outer layers and "Printed Boards" for inner layers like this: per IEC 62368-1:2023, "there is no material group designated as 'Printed Boards' and 'Other Materials'. The material groups are based on the CTI and there are material groups I, II, IIIa, and IIIb only" — that standard's own four-way grouping, finer than the three columns Table F.4 prints. Where a standard is written that way, the laminate's CTI group decides PCB creepage with no printed-wiring relief at all, and on the IEC 60664-1 numbers above group III needs 8.0 mm at 800 V in pollution degree 2 against 4.0 mm for group I: exactly half. IEC 62368-1 is an audio/video, information and communication technology equipment standard. We did not obtain its text and did not verify its scope of application, so we draw no conclusion about which products it reaches. The actionable point stands anyway: establishing which standard governs the board is a prerequisite, not a detail, and whether that standard carries a printed-wiring exemption is what decides the CTI requirement.
On the vehicle side, the OEM table is written for material group I. A high-voltage inverter insulation-coordination worksheet posted publicly by an engineer works an 880 V DC maximum supply, assigns pollution degree 2 from the IP class, and dimensions basic and reinforced insulation from a German OEM-consortium standard (LV 123, cited there as MBN 20123:2020-01 clause 7.7.6.3) under a heading stating the premise as "material group I [600CTI]". In the same document the IEC 60664-1 route — written there as "material group I-IIIa, PCB", which is the printed-wiring column — appears as an exception, available only "with noted explanation and customer agreement". Read that the right way round. The CTI 600 on a customer specification is often not an arbitrary margin but what the OEM table assumes, and the route that makes material group irrelevant is the exception needing agreement, not the default. The caveat is real: this is a third-party engineer-authored worksheet, not the OEM standard, which is a German OEM joint specification not available for public reading. We did not obtain LV 123 or MBN 20123 and quote none of its dimensions.
Pollution degree 3 has no printed-wiring column at all. Table F.4 provides those columns for pollution degrees 1 and 2 only. At pollution degree 3 the group-dependent columns apply: 10.0 mm for group I against 12.5 mm for group III at 800 V — group I about 20% shorter — less dramatic than the halving at pollution degree 2, still substantial on a board. Which one applies is a definition, not a preference. Pollution degree 2 is non-conductive pollution with occasional temporary conductivity from condensation; pollution degree 3 is conductive pollution, or dry pollution that becomes conductive because condensation is to be expected.
A mixed creepage path is governed by its worst segment. Clause 5.2.2.7: a creepage distance may be split into portions of different materials or pollution degrees if one portion withstands the whole voltage, "or if the total distance is dimensioned according to the material having the lowest CTI and the highest pollution degree". Real board paths cross solder mask, adhesive, component bodies, connector housings, potting and routed edges, so one low-CTI segment can set the dimension for the whole path. That makes the laminate CTI a floor you cannot design away.
And one route that makes CTI 600 unnecessary
IEC 60664-3:2016 covers coated printed boards, inner-layer surfaces included, and defines two protection types. Type 1 improves the microenvironment so that the IEC 60664-1 requirements for pollution degree 1 apply under it — on the table above, 2.4 mm at 800 V, all material groups. Type 2 is treated as similar to solid insulation: the solid-insulation requirements apply, spacings must be at least those in its Table 1, and "the requirements for clearances and creepage distances in IEC 60664-1 do not apply". Table 1 gives 0.15 mm above 0.8 kV up to 1.0 kV peak.
So for an 800 V board at pollution degree 2: uncoated group III, 8.0 mm; uncoated group I, 4.0 mm; qualified type 2 coating, 0.15 mm — roughly 27 times shorter than what CTI 600 buys (4.0 / 0.15 = 26.7, a ratio of lengths, not of squared quantities). This is not an argument for a cheaper laminate. Both types must be qualified by test, and type 2 additionally passes a scratch-resistance test before the sequence, and IEC 60664-1 applies in full to every unprotected part. Coating and material CTI are substitutes for each other, so changing the grade without knowing which one the design was dimensioned on quietly changes a premise.
On the vehicle side the chain ends in delegation. ISO 6469-3:2021, which we obtained in full, requires at clause 6.4.5 that clearance, creepage and solid insulation of voltage class B components be "designed according to the applicable sections of the IEC 60664 series", with normative references listing "IEC 60664 (all parts)" — which pulls the coating part in too. Its own text contains no occurrence of "material group", none of "pollution", no CTI and no numeric creepage table. (ISO 21498-1 is sometimes cited as specifying creepage. It does not, and says of itself that it does not cover electrical safety.)
Six premises decide the answer
The short version: a CTI 600 requirement has a basis, and a buyer who treats it as padding will be wrong more often than right. But that basis is almost never the printed-wiring column of IEC 60664-1 — it is a table layered on top of that standard, or a pollution degree, or the materials the creepage path crosses. So what is open to change is not the grade but the premise, and the premise belongs to the designer, not the mill. Drop to CTI 400 without establishing which premise applies and nothing visible happens: the board is simply no longer dimensioned on the document it was certified against.
Tg, Halogen-Free and the Rest of the Specification
A Tg figure — Tg is the glass transition temperature — without its test method does not support a comparison. From one Japanese laminate maker's published catalogue, for one grade on one page: Tg by TMA, method A, about 180 °C; Tg by DMA, method A, about 220 °C. Same material, same datasheet, roughly 40 °C apart, and the methods are not convertible. Write the method number: 2.4.25 for DSC (older datasheets: 2.4.25c), 2.4.24 for TMA, 2.4.24.1 for T260/T288, with Td under either 2.4.24.6 or ASTM D3850 and DMA usually quoted as a maker's own "method A". Related trap: RTI and MOT are not Tg. We saw a relative thermal index of 130/140 °C (electrical/mechanical) on that same grade, and, from a different datasheet, a grade quoting a maximum operating temperature of 150 °C with Tg 175 °C.
CTI and Tg are independent axes. A buyer hunting an equivalent to a high-CTI, mid-Tg part number usually assumes moving up in thermal class is the safe direction. It is not a direction at all. One Japanese maker's published UL recognition table gives counter-examples in both directions on a single page: a halogen-free FR-4.1 grade at CTI PLC 0 (≥ 600 V) with RTI 130/140 °C and Tg about 180 °C by TMA; its near-neighbour, same ANSI designation and same RTI, at PLC 2; and a grade with Tg about 200 °C by TMA sits at PLC 4, the second-lowest band — while the row with the highest RTI on the page is PLC 0.
High CTI and halogen-free pull the same way, not against each other. A technical paper from a developer of high-CTI laminates explains why: conventional brominated bisphenol-A epoxy includes easily carbonized aromatic groups, and the weak C–Br bond readily generates HBr at high temperature, accelerating resin carbonization, so that chemistry does not reach a good CTI value. High-CTI formulations cut the brominated content, while keeping flame retardancy, in favour of harder-to-carbonize resins, and add aluminium hydroxide filler to raise the index. What narrows the field is the formulation window: more filler raises CTI but destabilizes the resin system, and because the reinforcement is pre-impregnated, the combination of glass fabric and resin content that still works turns out to sit in a narrow band. That constrains thin dielectrics and multilayer use, and it explains the small number of qualifying part numbers better than any story about brominated chemistry.
Halogen-free: write the standard, and know what it covers. IEC 61249-2-21, Edition 1.0, published 12 November 2003, covers non-halogenated epoxide woven E-glass reinforced laminated sheets, copper-clad, 0.05 to 3.2 mm thick, at Cl ≤ 900 ppm, Br ≤ 900 ppm and Cl + Br ≤ 1,500 ppm. It is a laminate standard, not a board standard, and we tried to trace the relationship to the IPC/JEDEC JS709 family for a board-level declaration and could not confirm the connection. It is also a commercial consensus rather than a regulation: by one drafter's published account the limits were settled around 2001 in an IEC working group, between a Japanese proposal of roughly 450 ppm and a US proposal of 1,200 ppm.
And CTI itself has to be written with its method, for the same reason. One publicly available datasheet we read states its CTI as Grade 0 (>600) measured to ASTM D3638, not IEC 60112. The two are not guaranteed to agree — the summary we found states that a later IEC 60112 edition (it names the fourth) can yield values differing significantly from D3638 — while the IEC material groups are defined on IEC 60112 with solution A. So a specification saying only "CTI ≥ 600 V" can be satisfied with a D3638 number and read as an IEC 60112 number. Of every trap in this article it is the one most likely to pass unnoticed through a qualification file.
CAF: The Test Method Stops at 100 Vdc
CAF (conductive anodic filament) growth is the other high-voltage failure mode, and we obtained IPC-TM-650 Method 2.6.25, Revision C (February 2021) in full.
And then, in the method's own text: "This IPC test method has not been designed for use with voltages exceeding 100 Vdc bias, however higher voltage bias testing is being routinely required to be performed, such as for electric vehicle electronics." The standard test for the other high-voltage failure mode has the same shape of gap as the tracking test. It stops below the voltage the board runs at, and the standards body wrote that down rather than leave it to be discovered.
Two consequences for an RFQ. "We run 2.6.25" is not an answer. Ask the bias voltage, the duration, the coupon's hole wall-to-wall spacing, the chamber condition used, and the reinforcement orientation. Without those, two suppliers' CAF claims are not comparable with each other, let alone with the stress your board sees; one datasheet we read advertises "high voltage CAF resistance" as a headline feature without stating a bias voltage anywhere. And the failure criterion is logarithmic. One decade is a factor of ten, so do not let a report restate it as a percentage, and note that the sensitivity to a 5% humidity error is large against the pass/fail margin itself.
How Narrow the Field Really Is
CTI ≥ 600 — material group I, PLC 0 — is one band above where most FR-4 sits, and the published record shows how thin that band is.
Isola publishes a product guide with a UL CTI class column across its portfolio. Across the eighteen rows of that table there is no PLC 0. The highest is Class 1, 400 to below 600 V. The same table shows how little thermal class says about tracking: at Tg 200 °C it lists grades at Class 1, at Class 2 and at Class 3. The company's grade marketed specifically for tracking resistance, DE104 KF, is CTI 400 — Tg 135 °C, UL 94 V-0, IPC-4101C /21 — and its datasheet gives dishwashers and washing machines as example applications, framing its market as damp, dusty or corrosive environments. A dedicated tracking-resistant product from a major Western supplier stops one group below material group I. A semiconductor vendor's application note says the same from the design side: "Most FR4 materials used in PCB fabrication are rated for material group IIIa", and its own isolation products are material group I specifically to reduce package size and PCB footprint.
Grades at group I do exist, with the other requirements attached. The datasheet discussed above claims, in a single grade, halogen-free, high Tg (170 °C minimum, 175 typical, by DSC), CTI 600 V, UL 94 V-0, high-voltage CAF resistance, availability on E-glass 7628, 2116 and 1080, and three IPC-4101E slash sheets. So CTI 600 plus high Tg plus halogen-free is not an impossible combination — it is a short list, stated to a method you have to read for.
The oldest market for tracking-resistant grades appears to be appliances — which is what the dishwasher example records; vehicle electrification arrived on top of it, and IPC-TM-650 2.6.25C — a test method, not a marketing document — names electric vehicle electronics as the reason higher-bias CAF testing is routinely demanded. We found no confirmed CTI requirement for any individual application category. The requirement comes from the standard the product is certified to and the six premises above, not from the application name.
What a Buyer Should Do
The aim is not to argue the customer's specification down. It is to write the requirement so more than one mill can meet it, and to know which parts are load-bearing. The discipline we apply to converting a brand grade into measured properties in our map of Japanese materials companies applies here, with laminate-specific traps.
- CTI, with method and solution:
CTI ≥ 600 V, IEC 60112, solution A — plus the raw value, not only the PLC or class, since group IIIa straddles two PLC bands.
- Tg, with its method:
Tg ≥ 150 °C (IPC-TM-650 2.4.25, DSC). If what you need is a continuous-use temperature, specify RTI or MOT by name instead.
- Halogen-free, with the standard and the measurement:
IEC 61249-2-21 (Cl ≤ 900 ppm, Br ≤ 900 ppm, Cl + Br ≤ 1,500 ppm), measured per IPC-TM-650 2.3.41. Say separately if you need a board-level declaration, because that standard covers the laminate.
- CAF as conditions, not a method number: bias voltage, duration (500 hours biased, 596 total, or longer), coupon hole wall-to-wall spacing, chamber condition, reinforcement orientation relative to the via rows.
- UL through the file, not the datasheet: ask for the recognition file number and the row in it for that grade, and look it up in UL's product database yourself. Its ANSI column is also the halogen cross-check — UL 746E splits FR-4 into FR-4.0, brominated, and FR-4.1, non-halogen, the latter requiring halogen-content verification.
- Prepreg qualified with the laminate, as a stack: UL's CCIL and MCIL programmes recognize laminate and prepreg combination as a set, and blended-resin FR-4 runs into the "generically similar" requirement of UL 796 clause 16.2, where the slide lists a new type designation as required. Specify the glass styles (7628, 2116, 1080) and the resin content, not just the grade — a prepreg substitution here is a material change, not a logistics one.
- CTI on its own line, whatever slash sheet you cite. We could not confirm whether CTI is a requirement line in any IPC-4101 slash sheet — the standard is paid and we did not obtain it. Every datasheet we read carries CTI under a UL or ASTM heading instead, so our reading, offered as a reading rather than a fact, is that a slash sheet alone probably does not specify it.
- The board requirement as IPC-6012E Class 3 plus IPC-6012EA, the automotive addendum of October 2021, rather than as IATF 16949 — which is a supplier quality-management requirement, not a board acceptance criterion. The addendum itself says procurement documentation takes precedence over it, so your own specification can be stricter where it needs to be. For environmental requirements, ISO 16750-4 sets climatic loads by mounting location; we confirmed its bibliographic record but not its body, and the "−40 to +125 °C thermal cycling" often quoted from it is not a general requirement of the standard.
Then put the first three premises above to whoever owns the drawing — which table set the dimension, what pollution degree was assigned and on what basis, and whether the coating is qualified to IEC 60664-3 as type 1 or type 2. (The vehicle worksheet we read assigns pollution degree from the IP class rather than by preference, so a basis should exist on record.) With those answers a CTI 400 alternative is either obviously acceptable or obviously not, before samples are ordered rather than after.
Availability questions belong in the same round, because by trade-press accounts, some laminate makers introduced allocation quotas and purchasing restrictions during 2026, commonly rationed against last year's volume: ask whether the grade is allocated, what the cut-to-size and minimum-order position is, and for lead time by grade rather than the mill's general figure. On qualifying what arrives, our sample evaluation protocol covers how many lots prove what and our vendor due-diligence checklist what to verify about the mill; the notice clocks that decide a claim on a delivered lot are in our claim-handling guide. That specification block follows the same discipline — a numeric band, a method, a tolerance — that our RFQ specification builder enforces for powders, and our Other Materials page asks for the same: send the specification with your expected annual quantity, and we answer on feasibility first, before samples or a quote. In an allocation market the payment term carries weight beyond its financing cost — T/T, L/C and usance compared sets out what each side holds at each date.
Frequently Asked Questions
Is there a CTI 800 grade for an 800 V system?
No. IEC 60112's scope states the method is designed for test voltages up to 600 V AC, so 600 is the top of the scale and nothing above it exists to be measured. CTI 600 V is a material rating; 400 V and 800 V are system working voltages. They are different axes, and a specification asking for CTI 800 cannot be satisfied.
Does a higher CTI always allow tighter spacing on a PCB?
No. In IEC 60664-1:2007 Table F.4 the printed-wiring-material columns apply to all material groups at pollution degree 1 and to all except IIIb at pollution degree 2, and the pollution-degree-2 values equal the material group I values — so a board dimensioned on that column already has group I spacing. Where CTI does decide the dimension: pollution degree 3, which has no printed-wiring column; a governing standard that grants no printed-wiring exemption; an OEM table written for material group I; and a mixed creepage path under clause 5.2.2.7. We read the 2007 edition; the current 2020 edition renumbers this as Table F.5 and we did not verify that the row values are unchanged.
Can a halogen-free laminate reach CTI 600?
Yes, and the chemistry favours it. Brominated bisphenol-A epoxy carbonizes readily and releases HBr at temperature, which is why brominated FR-4 struggles to reach a high CTI; high-CTI formulations use harder-to-carbonize non-brominated resins with aluminium hydroxide filler. One datasheet we read combines halogen-free, Tg above 170 °C, CTI 600 V and UL 94 V-0 in one grade. What narrows the field is the formulation window on glass style and resin content, not the halogen requirement.
Does a high Tg grade imply a high CTI?
No, in either direction. One maker's published UL recognition table shows a halogen-free FR-4.1 grade at CTI PLC 0 (≥ 600 V) with RTI 130/140 °C, and on the same page a grade with Tg about 200 °C by TMA sits at PLC 4, the second-lowest band — while the row with the highest RTI on the page is PLC 0. In one Western supplier's published product guide there is no PLC 0 at all, and grades at Tg 200 °C appear at three different CTI classes.
Is IEC 61249-2-21 a halogen-free standard for the whole board?
No. Its scope covers copper-clad laminated sheets of 0.05 to 3.2 mm. The limits — Cl ≤ 900 ppm, Br ≤ 900 ppm, Cl + Br ≤ 1,500 ppm — came from a 2001-era IEC working-group compromise, so they are a commercial consensus rather than a regulation, and a board-level halogen-free declaration needs a separate basis that we could not trace to a primary document.
References (Public Sources)
- IEC 60664-1:2007, full text: clauses 4.6.2, 4.8.1.2 to 4.8.1.5, 5.2.2.7, 5.2.4 and Table F.4; official previews of the 2020 third edition and the 2025 consolidated edition 3.1 for scope, contents and change summary only. IEC 60664-3:2016 official preview: clauses 4.1, 4.3, 4.4 and Table 1. IEC 60112:2025 webstore scope and change record. IEC 61249-2-21:2003 webstore records, with Doug Sober in I-Connect007, 12 November 2015, on the origin of the 900/900/1,500 ppm limits. ISO 6469-3:2021, full text, clause 6.4.5 and normative references; ISO catalogue record for ISO 21498-1:2021.
- IPC-TM-650 Method 2.6.25 Revision C (February 2021), full text including Warning 1. IPC-2221C (December 2023) and IPC-6012EA (October 2021) tables of contents, with the latter's scope clauses 0.1, 0.1.2 and 0.1.4. UL 746E material published by UL Japan for a 2013 UL forum, on the FR-4.0 / FR-4.1 split and the CCIL/MCIL programme; UL's published engineering Q&A on IEC 62368-1.
- Isola product guide (2022) and Isola DE104 KF datasheet. Texas Instruments application notes SLUP419 (March 2024) and SLUAAR5 (August 2023); a vehicle high-voltage inverter insulation-coordination worksheet posted in Texas Instruments' engineering community (May 2021) referencing LV 123 / MBN 20123:2020-01 clause 7.7.6.3 — an engineer-authored document, not the OEM standard.
- A Japanese laminate maker's published catalogue, for its UL recognition table and grade-by-grade Tg by TMA and DMA; a technical paper by a developer of high-CTI laminates, for the resin chemistry; a publicly available laminate datasheet (2024) for a halogen-free, high-Tg, CTI 600 V grade stating CTI to ASTM D3638. None is identified here. A widely syndicated PCB trade blog, consulted and not relied on for any figure: its creepage-by-voltage table and its characterization of IPC-2221C values as proposals.
What we could not confirm: that the row values of Table F.5 (2020) match Table F.4 (2007); the text of IEC 62368-1, its scope of application and its Table 17 values; the values in IPC-2221C Tables 6-1 and 6-2; whether CTI is a requirement line in any IPC-4101 slash sheet; the bodies of IEC 60112:2025, IEC 61249-2-21 and IPC-6012EA; the text of ISO 16750-4, where the file we retrieved proved not to be the standard; the text of LV 123 or MBN 20123, not publicly available; the link between IEC 61249-2-21 and the IPC/JEDEC JS709 family; industry practice for CAF bias and duration on vehicle high-voltage boards; an authoritative source for automotive multilayer boards running Tg 150–170 °C; and a CTI requirement for any individual application category. Where each arises above, we have said so rather than fill the gap.
Nami Tech Solutions (NTS) works project by project, not from standing inventory. On a laminate enquiry of this kind our work starts by rewriting the specified part number as measured properties with their test methods — CTI with method and solution, Tg with method, halogen limits with the measurement standard, CAF with its conditions — and putting that written specification to mills in their own language, to find out which of them holds a grade that meets it and whether they will cut and ship a small quantity. We do not represent a grade as equivalent until the buyer's own tests say so.
