Walsin MLCC Datasheet Search: 7 Symbols Written Two Ways in 14 Files, and a Typed Less-Than-or-Equal Sign Finds 18 Percent

A text search of a Walsin MLCC datasheet can miss most of what it should find. Across 252 pages from 14 Walsin PDFs, 7 symbols are written as two different characters. A typed micro sign finds 28.0 percent of the micro marks and a typed less-than-or-equal sign 17.8 percent of the limits. Only plus-minus is written one way.

Venatech Limited, an authorised Walsin distributor in Hong Kong, holds 28 datasheets from Walsin Technology (Walsin), 451 pages. This post is written from the first 14 PDFs alphabetically, 252 pages, run through pdftotext and counted character by character. We did not read the page images; we read the text layer, which is what a find-in-page box and a parsing script see. Every part number below is real and belongs to a series whose sheet is in the set. The question is the one every engineer asks after pressing Ctrl-F on a Walsin sheet and getting nothing back: is the value not there, or is the character different?

How many ways do Walsin’s MLCC datasheets write the same symbol?

Seven symbols have two spellings, and one has one. We counted every occurrence of each candidate character across the 14 files. The micro sign, greater-than-or-equal, less-than-or-equal, the ohm, the degree Celsius, the delta and the colon are each written as two different code points somewhere in the set; only the plus-minus sign, with 2,066 uses in all 14 files, is written one way.

We have not seen this counted anywhere. Eight of the 8 files that contain a micro sign contain both spellings. There is no file in the set where the micro sign is written one way. On top of those 4,922 marks, a further 375 capacitance and inductance values are written with a plain letter u, as in 0.1uF, in 6 files.

How much of a datasheet does a typed micro sign find?

Between 3.8 and 32.6 percent, depending on the sheet. The U+00B5 character is the one the common keyboard shortcuts produce, so it is the likeliest to be typed. Across the eight files with micro marks, a typed U+00B5 finds 1,378 of 4,922, which is 28.0 percent, and the share ranges from 3.8 percent in the middle-to-high-voltage sheet to 32.6 percent in the soft-termination ST sheet.

The pattern is by sheet, not by series: the two automotive sheets, MT and ST, hold 960 of the 1,378 U+00B5 marks between them. The RF, RH, UF, RT and HH sheets hold none because every value is in picofarads. A search for the micro sign in 03_Large_Size_1825_2220_2225 returns 4 of 93 hits if typed as U+00B5 and 89 of 93 if typed as the Greek mu.

Why does a search for the less-than-or-equal sign miss four limits in five?

Because the sheets mostly use a different character. The U+2266 form ≦ appears 1,163 times and the U+2264 form ≤ only 252 times; in the General Purpose sheet the split is 222 to 14, and in the MT sheet 382 to 15. A typed ≤ therefore finds 6 percent of the limits in the General Purpose sheet and 4 percent in MT.

The greater-than-or-equal sign splits more evenly, 2,299 to 2,104, but unevenly by file. In 01_General_Purpose the U+2267 form wins 469 to 170. In seven sheets (02, 03, 05, 07, 08, 09 and 10) only the U+2265 form appears, and sheet 06 holds one U+2267 against 32 U+2265, so a search for ≧ written against sheet 01 finds nothing in seven files. The insulation resistance and dissipation-factor limits are written in these characters, as in the insulation resistance count.

Which of these differences does a search box remove?

Four of the seven, in a viewer that folds compatibility characters; three survive. We ran the Unicode compatibility normalisation (NFKC) on each pair. Micro sign and Greek mu fold, ohm sign and Greek omega fold, the degree-Celsius character folds to °C, and the full-width colon folds to ASCII. The two comparison pairs and the delta pair do not fold.

If a viewer does fold, the only searches still blind are the limits and the delta. Those are the characters in the pass criteria, such as “Cap≧10μF” and “∆T ≦ 150℃”.

Do the datasheet characters match the export?

No, and that is a second mismatch, with 1,378 against 0. The micro sign U+00B5 appears 1,378 times in these 14 datasheets. Our published count of the parametric export, in the BOM import check, found the Greek mu U+03BC in 2 export cells and U+00B5 in none. Folding datasheet text to the export spelling works for the micro sign; matching an export cell against datasheet text spelled with U+00B5 does not, unless both sides are normalised.

What should an extraction script or a search do?

Fold before comparing: NFKC handles 4 of the 7 pairs and explicit maps cover the other 3. Three rules follow from the counts.

1. Normalise to NFKC, then map by hand. NFKC handles 4 of the 7 pairs; add explicit maps for ≧ to ≥, ≦ to ≤ and the delta.
2. Search for both spellings of every limit. A search written against one file finds 6 of 14 files for ≧.
3. Do not read a zero as an absence. The five microwave sheets have no micro sign because they have no microfarad values, not because the search failed.

We hold 28 Walsin datasheets, 451 pages, and send them on request. We hold none for thin film, inductor, varistor, disc capacitor, RF or antenna lines, so this post is confined to MLCC. Paste your incumbent into our free cross-reference tool and read the sheet for the series it returns. What is in Hong Kong is on ready stock, 1,125 part numbers as at 23 September 2026, and the range sits on the MLCC page.

The short version

Seven symbols are written two ways in Walsin’s first 14 datasheets, and every file that has a micro sign has both spellings. A typed U+00B5 finds 28.0 percent of the micro marks; a typed ≤ finds 17.8 percent of the limits and 6 percent in the General Purpose sheet. Four of the seven pairs fold under NFKC and three do not. Only the plus-minus sign, 2,066 uses in 14 files, is written one way. Qualification stays with you.

Send us the limit you were searching for and the part number, and we will read the sheet and return the line. And a question back: when your search found nothing in a Walsin sheet, did you conclude the limit was not there?

Frequently asked questions