Thin Film Precision Resistor Tolerance: One Value on 52,808 of Walsin’s 58,314 WF Rows

We re-measured Walsin’s thin film precision resistor tolerance column ourselves, file by file. 52,808 of 58,314 WF rows read exactly ±0.01% and nothing else; the eighth file, 0201, carries three different values instead of one. A prior note flagged this as unconfirmed. We counted it from scratch, and the count holds.

Venatech Limited, an authorised Walsin distributor in Hong Kong, holds all 20 of Walsin Technology (Walsin)’s thin film parametric export files. This post is measured from the eight non-empty Thin_Film_Precision_Chip_Resistor_WF_Series_*.xlsx files that make up the WF precision line, 58,314 rows in total. Every part number below is real and came out of those files, and every count is a count of part numbers, not pieces. A claim about this exact column had been flagged in our own notes as unconfirmed before this post. We opened the files and counted the column ourselves rather than repeat it.

How many of Walsin’s 58,314 WF thin film rows actually carry one tolerance value?

Fifty-two thousand, eight hundred and eight, out of 58,314. We read the Tolerance column in each of the seven case-size files other than 0201 — WF_Series_0402_100, 0603_160, 0805_201, 1206_321, 1210_322, 2010_502 and 2512_643 — and every one of the 52,808 rows in those seven files prints ±0.01%. No exceptions, no second value, in any of the seven files.

We have not seen this counted file by file anywhere else. The flagged claim said “±0.01% on all 52,808 rows of seven case-size files while the 0201 file carries three tolerances.” That is exactly what our own count returns. The claim is confirmed, not assumed.

Does the 0201 file really carry three tolerances, and what are they?

Yes, and we counted the split. The 5,506 rows in WF_Series_0201_060.xlsx divide into ±0.5% (1,916 rows), ±1% (1,916 rows) and ±0.1% (1,674 rows). None of the three matches the ±0.01% that fills every other WF file. The smallest WF footprint is also the only one where the WF series does not offer its own flagship tolerance.

That is the shape of the finding: the column is not “mostly ±0.01% with some noise.” It is two completely separate populations — one case size locked at three loose-to-mid tolerances, and seven case sizes locked at one tight tolerance, with no overlap between the two sets at all.

If the letter in a WF part number is not a tolerance code, what is it?

TCR, measured consistently across all seven uniform-tolerance files. Walsin’s general numbering convention normally puts the tolerance letter in this position (B = ±0.1%, F = ±1%, and so on, in the thick film line). Here it cannot be a tolerance code — the Tolerance column is frozen at ±0.01% regardless of which of eight letters sits in the part number. So we checked what the letter does correlate with, and it is TCR: B → ≦±3 ppm/°C, Z → ≦±5, W → ≦±10, F → ≦±15, and R/U and Q/T both → ≦±25 and ≦±50 respectively, split instead by a one-step-lower power rating at the same case size.

We checked all 52,808 rows across the seven files for this, not a sample. The mapping holds in every file with zero exceptions.

Does that letter-to-TCR decode survive at 0201, where tolerance is not frozen?

No, and that is itself informative. At 0201 the same four letters that behave cleanly elsewhere — Q, R, T, U — mix across two TCR grades depending on which tolerance the row carries. R rows read ≦±25 ppm/°C at ±0.1% tolerance but ≦±50 ppm/°C at ±0.5% and ±1%; Q reads ≦±25 ppm/°C at every tolerance it appears with. We grouped all 5,506 rows in the 0201 file by letter, tolerance and TCR to check this, and it does not resolve to one letter, one TCR.

That tells us the clean decode in the seven other files is a side effect of tolerance being constant there, not a tolerance-independent part-numbering rule. We would not have found this without checking the one file where the assumption breaks.

Why does one flat tolerance column matter for a precision cross?

Because a buyer filtering a BOM on tolerance alone cannot tell seven WF case sizes apart on that field. If every row in 0402 through 2512 reads ±0.01%, tolerance stops being a discriminator the moment you leave 0201, and TCR, case size and rated power become the only fields doing any selection work. A cross built purely on “match the tolerance” against this export will return every size-0402-and-up part as an equal candidate, which is correct on tolerance and wrong on everything else that actually differs.

The practical reading: never cross a WF part on tolerance alone outside 0201. State the TCR grade and case size you need in the same breath, because the tolerance column will not help you choose between them.

What should a buyer sourcing WF thin film resistors do with this?

State TCR and case size up front, because tolerance will not separate anything for you above 0201. Thin film carries zero datasheets in our file — none of the 28 Walsin PDFs we hold covers thin film at all — so the parametric export is the whole of our evidence here, and we say so before you read a figure from it.

1. Write the TCR grade you actually need, not just the tolerance. ±0.01% appears at six TCR grades across the seven case sizes; naming only the tolerance leaves the TCR to chance.
2. Cross the whole BOM in one pass. Paste up to 500 part numbers or drop an Excel or CSV into the free cross-reference tool. It decodes 18 manufacturers’ part numbers and matches against 800,000+ crossable Walsin part numbers, stating every difference rather than hiding it behind the word “equivalent.”
3. Check what is physically in Hong Kong on ready stock before committing to a production slot — that page is a dated snapshot as at 23 September 2026, not a live feed.
4. Ask us for the WF datasheet. We hold 28 Walsin datasheets, 451 pages, and none of them is a thin film sheet. A request from you is what puts it on the list to Walsin.

For the catalogue-wide TCR and tolerance grid across all three thin film series, see The thin film TCR and tolerance grid. For what the other nine empty thin film exports were meant to hold, see What is actually in Walsin’s thin film catalogue. The full line is on the Walsin thin film range page.

The short version

Walsin’s WF precision thin film tolerance column is not a spectrum — it is two populations. 52,808 of 58,314 rows, across seven case sizes, read exactly ±0.01% with zero exceptions, which we confirmed by reading the column ourselves rather than repeating an unconfirmed note. The 0201 file alone carries ±0.1%, ±0.5% and ±1%, and no other file carries any of those three. The single letter that looks like a tolerance code in the part number is not one — it tracks TCR instead, cleanly in the seven uniform-tolerance files and messily at 0201, which is the proof that the clean version is a side effect, not a rule.

Tell us the TCR grade and case size your circuit needs, not just the tolerance, and we will come back with the WF candidates field by field — then tell you which ones were in Hong Kong on 23 September 2026. If you have seen a Walsin document that confirms what that letter actually stands for, we would like to see it before we publish it as a rule rather than a correlation.

Frequently asked questions