Walsin Thin Film Part Number: Position Five Separates Thin From Thick on 196,363 Rows, and 10 Resistance Codes Disagree

Every Walsin thin film part number is exactly 12 characters. We decoded all 93,914 of them. The prefix does not tell you the film type — position five does, and its letter set is disjoint from the thick film line’s across 196,363 rows. Ten resistance codes disagree with the column beside them.

---

A part number is the only field that survives a BOM export intact, so engineers decode it and trust the result. On this line that trust is misplaced in four specific places, and we can name all four. Venatech Limited, an authorised Walsin distributor in Hong Kong, holds all 203 parametric export files published by Walsin Technology (Walsin). This post is measured from the eleven non-empty thin film files — the eight Thin_Film_Precision_Chip_Resistor_WF_Series_*.xlsx sheets, Tantalum_nitride_Thin_Film_Chip_Resistor_MF_Series_1.xlsx and two Anti_Sulfuration_Thin_Film_Chip_Resistor_SF_Series_*.xlsx sheets — against the 137 files in Walsin_ChipR_Product_Data. Every part number below is real and sits in our extract. We hold no thin film datasheet, so the decode below is reverse-engineered from the export’s own columns, not read off a Walsin ordering key — and we flag it as such.

What do the twelve characters of a Walsin thin film part number encode?

Six fields, in a fixed order, on all 93,914 rows. We split every part number character by character and cross-tabulated each position against the column beside it. The length never varies: 12 characters, every row, no embedded spaces and no hyphens — unlike the thick film line, where 39,576 of 234,815 rows — 39,162 distinct part numbers — carry an embedded space.

Position	Field	Values we counted	Note
1–2	Series prefix	WF 58,314 · MF 20,520 · SF 15,080	precision, tantalum nitride, anti-sulfuration
3–4	Case size	8 codes	not a simple truncation — see below
5	TCR grade	8 letters for 6 grades	two letters are ambiguous
6–9	Resistance	3 significant digits plus a decade digit, R as the decimal point	10 rows disagree with the column
10	Tolerance	7 letters for 6 tolerances	one letter appears 15 times
11–12	Packaging	TL on all 93,914 rows	carries no piece count

The case size codes take the first two digits of the imperial size at seven of the eight footprints — 02, 04, 06, 08, 12, 20, 25. The eighth breaks the rule: 1210 is coded 10, its last two digits, because 12 was already taken by 1206. So WF12U1002TTL is a 1206 part and WF10U1002TTL is a 1210 part, and the two differ by 11% in board area. We verified this on all 93,914 rows — 29,217 rows coded 12 are 1206 and 7,216 coded 10 are 1210, with no exceptions. For the footprint names themselves see imperial and metric case sizes.

Does the prefix tell you whether a part is thin film or thick film?

No, and that trap is already established in our own published work: 102,449 thick film rows begin WF, MF or SF, which is more parts than the entire thin film catalogue holds. What has not been shown before is that the next character resolves it completely. We read position five on all 93,914 thin film rows and on all 102,449 thick film rows carrying one of those three prefixes — 196,363 rows in total — and the two letter sets do not overlap by a single character.

Line	Rows read	Letters at position five	Overlap
Thin film	93,914	B F Q R T U W Z	—
Thick film, WF/MF/SF prefix only	102,449	A E G H J K L M N P S V	zero shared letters

So WF12U1002TTL is thin film and WF12H1002BTL is thick film, at the same 1206 footprint and the same 10 kΩ, and the only character that says so is the fifth. We have not seen this counted anywhere. Two cautions before you build a filter on it: we measured it on today’s extract of Walsin’s published catalogue, not from an ordering key, and nothing stops Walsin from issuing a new letter. Treat it as a check you run, not a rule you trust blind.

Does the fifth character give you one TCR grade, everywhere?

Six of the eight letters, yes. Two invert at one case size. We cross-tabulated position five against the TCR column on every row: B is ≦±3 ppm/°C on all 3,064 rows, Z is ≦±5 on all 6,007, W is ≦±10 on all 11,812, F is ≦±15 on all 11,808, U is ≦±25 on all 14,507 and T is ≦±50 on all 14,534. Then there are Q and R.

Letter	Grade at 0402, 0603, 0805, 1206, 1210, 2010, 2512	Grade at 0201	Match / mismatch across footprints
U	≦±25 ppm/°C	≦±25 ppm/°C	Match — 14,507 rows, one grade
T	≦±50 ppm/°C	≦±50 ppm/°C	Match — 14,534 rows, one grade
R	≦±25 ppm/°C on 16,907 rows	≦±25 on 411 rows, ≦±50 on 1,064	Mismatch — the letter carries two grades at 0201
Q	≦±50 ppm/°C on 12,736 rows	≦±25 on all 1,475 rows	Mismatch — the meaning inverts at 0201

At every footprint except 0201, Q means ≦±50 ppm/°C and R means ≦±25. At 0201 Q means ≦±25 on every one of its 1,475 rows, and R means ≦±25 at ±0.1% and ≦±50 at ±0.5% and ±1%. A decoder built on the 88,408 rows above 0201 reads every 0201 Q part two grades tighter than it should and every 0201 R part at ±1% two grades looser. That is the smallest footprint in the line and the one most likely to be bought in volume. For what the grades buy you, see tolerance, TCR and dielectric codes decoded.

How many thin film part numbers are parametric twins of another?

826 rows, in 413 pairs, and no group larger than two. We built the five-field signature — case size, resistance, tolerance, TCR, rated power — for all 93,914 rows and counted collisions. 93,501 distinct signatures, 413 of them held by two part numbers each. Every pair is a direct consequence of the two anomalies above.

Pair	Case size	Resistance	Tolerance	TCR	Rated power	Match / mismatch
WF02Q1002BTL · WF02R1002BTL	0201	10 kΩ	±0.1%	≦±25 ppm/°C	1/20 W	Match on all five published fields — one of 411 such pairs at 0201
WF04T98R8TTL · WF04T1000TTL	0402	98.8 Ω	±0.01%	≦±50 ppm/°C	1/16 W	Match on all five, but the second part’s code reads 100 Ω
WF04U98R8TTL · WF04U1000TTL	0402	98.8 Ω	±0.01%	≦±25 ppm/°C	1/16 W	Match on all five, same code conflict
WF12U1002TTL · WF12R1002TTL	1206	10 kΩ	±0.01%	≦±25 ppm/°C	1/8 W vs 1/4 W	Mismatch on rated power — these are not twins

411 of the 413 pairs are the 0201 Q/R collision at ±0.1%: two orderable codes, five identical published fields, nothing in the export saying what differs. The other two are at 0402 and are a different problem. A cross tool that returns both members of a pair is not wrong — it is telling you the export cannot separate them. See one specification, 13 Walsin parts.

Which resistance codes disagree with the Resistance column?

Ten of 93,914. We decoded positions 6 to 9 on every row — three significant digits, a decade digit, R standing in for the decimal point — and compared the result against the Resistance column. 93,904 agree. Ten do not, and they fall into two groups that look like two different slips.

Part number	Code	Code decodes to	Resistance column says	Match / mismatch
WF04T1000TTL	1000	100 Ω	98.8 Ω	Mismatch — and WF04T98R8TTL already holds 98.8 Ω
WF04U1000TTL	1000	100 Ω	98.8 Ω	Mismatch — same at the ≦±25 ppm/°C grade
WF12U2210TTL	2210	221 Ω	22,100 Ω	Mismatch by a factor of 100
WF12U2610TTL	2610	261 Ω	26,100 Ω	Mismatch by a factor of 100
WF12U3010TTL · WF12U3610TTL · WF12U5110TTL	3010 3610 5110	301 Ω · 361 Ω · 511 Ω	30,100 Ω · 36,100 Ω · 51,100 Ω	Mismatch by a factor of 100
WF12U6810TTL · WF12U7410TTL · WF12U9310TTL	6810 7410 9310	681 Ω · 741 Ω · 931 Ω	68,100 Ω · 74,100 Ω · 93,100 Ω	Mismatch by a factor of 100

All ten sit in two files: Thin_Film_Precision_Chip_Resistor_WF_Series_0402_100.xlsx and Thin_Film_Precision_Chip_Resistor_WF_Series_1206_321.xlsx. The eight 1206 rows are the instructive ones, because the sibling grade in the same workbook does it correctly. The ≦±50 ppm/°C family holds both WF12T2210TTL at 221 Ω and WF12T2212TTL at 22,100 Ω — two codes, two values, no ambiguity. In the ≦±25 ppm/°C family, WF12U2212TTL does not appear in the export at all, and there is no WF12U part at 221 Ω either. So either the column is wrong on those eight rows or the part number is, and we cannot tell you which, because we hold no Walsin thin film datasheet. A two-decade error on a divider is not a tolerance problem, it is a wrong circuit.

Why does one tolerance letter appear on only 15 of 93,914 parts?

Because position ten is one-to-one on 93,899 rows and carries a second, unexplained symbol on 15. We read it on every row: T is ±0.01% on 52,808 rows, A is ±0.05% on 6,364, B is ±0.1% on 8,483, C is ±0.25% on 7,469, D is ±0.5% on 9,385 and F is ±1% on 9,390. Then W appears on 15 rows, also at ±1%, and nowhere else in the catalogue.

All 15 are the same shape: tantalum nitride MF series, 1206, ≦±10 ppm/°C, 2/5 W, and a resistance code ending 12. MF12W1012WTL at 10.1 kΩ is one of them. At those 15 resistance positions the four tighter tolerances are coded normally — MF12W1012ATL is the ±0.05% part — and F never appears. At the other 1,011 positions in that family, ±1% is F without exception.

Look at MF12W1012WTL once more. The W at position five means ≦±10 ppm/°C. The W at position ten means ±1%. The same letter carries two unrelated meanings inside one twelve-character code, and a positional decoder that loses its offset by one character will read a 10 ppm part as a 1% part. That is the whole argument for decoding by position and verifying against the column, which is what a part number that will not paste was about.

What should you check before pasting a thin film part number into a BOM?

Four checks, and all four come from the 93,914 rows above. Run them in this order and you will catch every defect named here.

1. Read position five before the prefix. WF, MF and SF sit on 102,449 thick film parts as well; the fifth character settles it.
2. Re-read the case size digits. 12 is 1206 and 10 is 1210.
3. Decode the resistance and compare it to the column. Ten of 93,914 disagree, eight by two decades.
4. At 0201, do not trust Q or R for TCR. The grades invert there.

Then stop decoding and ask. Paste up to 500 part numbers or drop your Excel or CSV into the free cross-reference tool and read every stated difference rather than a claimed match. Check what is physically in Hong Kong on ready stock — 1,125 part numbers as at 23 September 2026, a dated snapshot and not a live feed. The full line sits on the thin film page. Unit 207, 2/F, Sun Ling Plaza, 30 On Kui Street, Fanling, N.T., Hong Kong. (852) 2628 4777.

The short version

A Walsin thin film part number is twelve characters: two for the series, two for the case size, one for TCR, four for resistance, one for tolerance and TL. Position five, not the prefix, separates thin film from thick film, and the two letter sets are disjoint across 196,363 rows. 1210 is coded 10 while 1206 is coded 12. Q and R invert their TCR meaning at 0201 and nowhere else. 826 rows in 413 pairs are identical on all five published fields. Ten resistance codes disagree with the column. We hold no thin film datasheet, so every line of this decode is inferred from the export. We will not tell you a part fits. We will show you where it differs — qualification stays with you.

Send us the thin film lines off your bill of materials and we will decode each one character by character, tell you which field we could not confirm and why, and say which of them was in Hong Kong on 23 September 2026. And a question back: your selector decoded that part number too — did it tell you the fifth character, or did it stop at the prefix?

---

Frequently asked questions

<!– wp:rank-math/faq-block {"questions":[{"id":"faq-question-01","title":"How do you read a Walsin thin film resistor part number?","content":"Twelve characters, six fields: two for the series prefix, two for the case size, one for the TCR grade, four for resistance as three significant digits plus a decade digit, one for tolerance, and TL for packaging. We verified this split on all 93,914 rows. No Walsin ordering key confirms it; we inferred it from the export.”,”visible”:true},{“id”:”faq-question-02″,”title”:”Does a part number beginning WF mean the part is thin film?”,”content”:”No. 102,449 thick film part numbers begin WF, MF or SF. The fifth character settles it: thin film uses B, F, Q, R, T, U, W or Z and thick film under those prefixes uses A, E, G, H, J, K, L, M, N, P, S or V. We read all 196,363 rows and the sets share no letter.”,”visible”:true},{“id”:”faq-question-03″,”title”:”Which case size code is the exception on Walsin’s thin film line?”,”content”:”1210, coded 10. Seven of the eight codes are the first two digits of the imperial size — 02, 04, 06, 08, 12, 20, 25 — while 1210 uses its last two because 12 already means 1206. We confirmed it on 29,217 rows coded 12 and 7,216 coded 10, with no exceptions.”,”visible”:true},{“id”:”faq-question-04″,”title”:”Does the TCR letter mean the same thing at every case size?”,”content”:”Not for two of the eight letters. Q means 50 parts per million per degree Celsius at seven footprints and 25 at 0201, on all 1,475 of its 0201 rows. R means 25 at seven footprints and splits at 0201, reading 25 on 411 rows and 50 on 1,064. The other six letters are consistent.”,”visible”:true},{“id”:”faq-question-05″,”title”:”Are any Walsin thin film part numbers duplicates of each other?”,”content”:”No part number repeats, but 826 rows in 413 pairs are identical on all five published fields — case size, resistance, tolerance, TCR and rated power. 411 pairs are the 0201 Q and R collision at plus or minus 0.1 percent. The export does not say what distinguishes the members of a pair.”,”visible”:true},{“id”:”faq-question-06″,”title”:”Do any Walsin thin film resistance codes disagree with the stated resistance?”,”content”:”Ten of 93,914. Two 0402 parts carry a code reading 100 ohm against a column reading 98.8 ohm. Eight 1206 parts in the 25 parts per million per degree Celsius grade carry a code two decades low — 221 ohm against 22,100 ohm, and seven more of the same shape. We cannot say which field is wrong.”,”visible”:true},{“id”:”faq-question-07″,”title”:”Can Venatech confirm this decode against a Walsin datasheet?”,”content”:”No. We hold 28 Walsin datasheets, 451 pages, covering 26 multilayer ceramic capacitor series and 2 chip resistor series. None covers thin film. The whole decode is inferred from the parametric export, flagged for confirmation, and on our request list with Walsin. Nothing has been requested in writing yet.”,”visible”:true}],”listStyle”:””,”titleWrapper”:”h3″,”titleCssClasses”:””,”contentCssClasses”:””,”sizeSlug”:””} –>

---