NEWS

Walsin Export Column Order: Column 8 Holds 14 Different Names, and a Tolerance Read by Position Is Wrong in 30 of 114 Files

Walsin export column order is not a schema. We read the header row of all 114 export files that hold parts: 89 distinct names, and only Part Number sits in the same column in every file. Column 8 carries 14 different names. A loader that reads column 6 as tolerance is right in 84 files and wrong in 30.


Venatech Limited, an authorised Walsin distributor in Hong Kong, holds all 203 parametric export files published by Walsin Technology (Walsin). This post is written from the 114 files that hold part numbers: 64 chip resistor, 18 MLCC, 11 thin film, 14 RF, 2 each for inductor, varistor and antenna, and 1 disc capacitor. We read each header with a spreadsheet library and recorded the column number of every name. Every part number below is real. If you read these files by column number, what are you reading?

Does the same column sit in the same place in every Walsin export file?

No. Only Part Number does. We read the header row of all 114 files and recorded the name at each column number. There are 89 distinct header strings. Column 1 holds one name in all 114 files. Column 2 holds two, columns 3 and 4 hold eight each, and column 8 holds 14. Of the 89 names, 62 sit at one column number and 27 at more than one.

Column number Distinct names Largest groups of files Match / mismatch for a loader written against the chip resistor files
1 1 Part Number 114 Match in 114 of 114 files
2 2 Series 50, Series Group 64 Match in 64 files; mismatch in 50
3 8 Series 64, Size 37, Application 4 Match in 64 files; mismatch in 50
4 8 Size 72, Dielectric 18, Resistance 11 Match in 72 files, 8 of them not chip resistors
5 10 Resistance 64, Capacitance 18, Tolerance 12 Match in 64 files; mismatch in 50
6 10 Tolerance 84, TCR 11, Insertion Loss (dB) 4 Match in 84 files; mismatch in 30
7 13 TCR 64, Rated Voltage 18, Rated Power 11 Match in 64 files; mismatch in 50
8 14 Rated Power 64, L 17, Packaging 11 Match in 64 files; mismatch in 50
9 13 Packaging 64, W 17, Termination 11 Match in 64 files; mismatch in 50
10 12 (in 103 files) Termination 64, T 17, Thickness (mm) 1 Match in 64 files; mismatch in 39; 11 thin film files have no column 10

We have not seen this counted anywhere. Of the first ten column numbers, nine carry between 2 and 14 different names, and the chip resistor layout is right for all 114 files at column 1 and for no more than 84 at any other. The files are 14 distinct ordered header lists: one for chip resistor, two for MLCC, one for thin film, six for RF and one each for the other four families.

What does a tolerance read by column number actually return?

Something other than a tolerance in 30 of the 114 files, covering 106,665 of 390,463 rows (27.3%). We took column 6, the position where the chip resistor, MLCC and varistor files all hold Tolerance, and recorded what the other files hold there. It is right in 84 files and 283,798 rows. In 30 files it returns a different quantity, in 11 of them a temperature coefficient: all 93,914 thin film rows. We counted rows after excluding the Test123999 QA row from the inductor files.

Name at column 6 Files Source files Rows An example part and its cell Match / mismatch against Tolerance
Tolerance 84 64 chip resistor files, 18 MLCC files including 00_ALL_MLCC_parts_combined.xlsx, Walsin_MOV_ALL.xlsx (twin pair) 283,798 SR180K05D: +/-10% Match
TCR 11 the 11 thin film files, e.g. Anti_Sulfuration_Thin_Film_Chip_Resistor_SF_Series_1.xlsx 93,914 SF12T1R00ATL: ≦±50 Mismatch — a temperature coefficient; its tolerance, ±0.05%, is in column 5
Inductance 2 Walsin_Inductor_ALL.xlsx (twin pair) 10,820 WLCW1005Z0J1N0TB: 1 Mismatch — a value in nanohenries; tolerance ±5% is in column 9
Frequency 2 Walsin_Filter.xlsx (twin pair) 470 KFBPF2012100C67B1U: 1125~1675 MHz Mismatch — no tolerance column exists in this family
Frequency 1 2 Walsin_Multiplexer.xlsx (twin pair) 232 AMDIP2004L157B1U: 2400 ~ 2500 MHz Mismatch — no tolerance column exists
Insertion Loss (dB) 4 Walsin_Coupler.xlsx, Walsin_Balun.xlsx (twin pairs) 178 RFBLN1005040C0T: 1 Mismatch — no tolerance column exists
Frequency 2 4 Walsin_SAW_Filter.xlsx, Walsin_BAW_Filter.xlsx (twin pairs) 16 DA18141747B3E1T: 1805~1880 MHz Mismatch — no tolerance column exists
Function 2 Walsin_Antenna_Switch.xlsx (twin pair) 18 RFASWA630PTF0E: SPDT Mismatch — a switch topology
Voltage Standing Wave Ratio 2 Walsin_Antenna_ALL.xlsx (twin pair) 88 RFANT6050110L0T: 2.0 (max.) Mismatch — no tolerance column exists
T.C. 1 Walsin_DiscCap_ALL.xlsx 929 SL0AH150J060: blank Mismatch — blank on all 929 rows; tolerance +/-5% is in column 5

Sixteen of the 30 files, 14 RF and 2 antenna, publish no tolerance at all. The other 14 publish one at column 5 in 12 files and column 9 in 2. A script that parses column 6 as a number fails on 1805~1880 MHz but succeeds on 1 and 0.32, and carries on.

Why do the chip resistor and thin film exports agree on one position out of nine?

Because the 64 chip resistor files carry a column the 11 thin film files lack. The families share nine header names, and we counted each position. Chip resistor files have ten columns and thin film files nine; the missing one is Series Group at column 2, so eight of the nine shared names sit one column to the left in thin film. Only Part Number stays put.

Shared name Chip resistor column Thin film column Match / mismatch
Part Number 1 1 Match
Series 3 2 Mismatch — shifted one column
Size 4 3 Mismatch — shifted one column
Resistance 5 4 Mismatch — shifted one column
Tolerance 6 5 Mismatch — shifted one column
TCR 7 6 Mismatch — shifted one column
Rated Power 8 7 Mismatch — shifted one column
Packaging 9 8 Mismatch — shifted one column
Termination 10 9 Mismatch — shifted one column

The consequence is concrete. A loader built on the 64 chip resistor files, 234,815 rows, and pointed at the 11 thin film files, 93,914 rows, reads column 6 expecting Tolerance and gets TCR, reads column 5 expecting Resistance and gets Tolerance, and finds no column 10. WF25H1R00BTL, a chip resistor, and SF12T1R00ATL, a thin film part, are both resistors with all nine shared fields filled, and they disagree about where each field lives. For the part-number side of the same boundary, see thick film or thin film: 6,836 pairs differ by one character.

Does a column name shared by several families keep its place?

Almost never: 14 of 15. Fifteen header names are used by more than one family, and we listed the column number each occupies in every family that uses it. One of the 15, Part Number, holds the same column number in every family. The other 14 move.

Header name Column number by family Match / mismatch
Part Number 1 in all eight families Match
Series 3 in chip resistor; 2 in the other seven Mismatch
Size 4 chip resistor; 3 inductor, MLCC, varistor, thin film; 3, 4 or 7 RF Mismatch
Tolerance 6 chip resistor, MLCC, varistor; 5 disc, thin film; 9 inductor Mismatch
Termination 10 chip resistor; 9 thin film; 11 or 12 MLCC Mismatch
Packing, SPQ 9, 10 disc; 8, 9 varistor; 12, 13 or 13, 14 MLCC Mismatch
Resistance, TCR, Rated Power, Packaging one column left in thin film than in chip resistor Mismatch
Dielectric, Capacitance one column left in disc capacitor than in MLCC Mismatch
MOQ 11 disc; 10 varistor Mismatch
Thickness (mm) 5 inductor; 10 MLCC Mismatch

Seventeen names also move inside one family: 13 in RF, whose six layouts differ, and four in MLCC. In the 173-row sample 00_ALL_MLCC_combined.xlsx, Remark is column 11 and Termination, Packing and SPQ sit one place later than in the 17 other MLCC files, where Remark is column 14.

Is one quantity ever called by several names?

Yes: 8 quantities carry two or three names, and the split between names often follows a split between filled and empty. We counted the non-blank rows in the files that declare each name.

Quantity Names used, files declaring each Rows non-blank Match / mismatch
Packaging Packaging 75 files (chip resistor, thin film) 328,729 of 328,729 reference
Packaging Packing 21 files (MLCC, disc, varistor) 0 of 49,912 Mismatch — same quantity, other name, never filled
Thickness T 17 MLCC files 47,358 of 48,052 reference
Thickness Thickness (mm) 3 files (inductor, MLCC sample) 10,993 of 10,993 Match on content, mismatch on name
Thickness Thickness 18 files (MLCC) 0 of 48,225 Mismatch — a column called Thickness that holds no thickness
Temperature coefficient TCR 75 files; T.C. 1; TCC 1 328,729 of 328,729; 0 of 929; 0 of 173 Mismatch — three names, one filled
Size Size 111 files; Size L*W (mm) 2; Dimension 2 all filled Mismatch — the inductor files carry both Size and Size L*W (mm)

The pattern is the useful part: the name that is filled is the one the chip resistor and thin film exports use, and the alternative is empty on every row. We counted that directly, not from the header. One variation is cosmetic but breaks a string match: 12 header strings put a space before a bracketed unit, as in Insertion Loss (dB), and 10 do not. Three names, Insertion Loss (dB) and Impedance for Frequency 1 (Ω) and 2 (Ω), each have a twin without the space, all six in the RF exports.

What should a loader key on instead of a column number?

The normalised header string and the product family, never a column number. Across 114 files, 89 names and 14 ordered layouts, a column number is dependable for exactly one name, Part Number. Four rules follow from the counts.

  1. Read by name, never by number. A column number is right for Part Number and, for a chip resistor layout, for nothing else across all 114 files.
  2. Key the schema on the file family before the column. The same name sits in different places across 14 layouts.
  3. Collapse the space-before-bracket variants before comparing, and keep the original string.
  4. Check that a column is filled. Packing, SPQ, MOQ, T.C., TCC and Thickness are declared and empty on every row we counted.

Our earlier join-key count of 90 header names came from a reader that renames the second Pin Description in Walsin_Multiplexer.xlsx to Pin Description.1. Read raw, the headers are 89 distinct strings and the multiplexer repeats one in columns 12 and 19. The correction changes none of the conclusions here. For the offset in that file, see the multiplexer export is offset by one column.

We hold 28 Walsin datasheets, 451 pages, and none for thin film, inductor, varistor, disc capacitor, RF or antenna, so for six families the export header is the only definition of a field we can show you. Our free cross-reference tool reads by name and family. What is in Hong Kong is on ready stock: 1,125 part numbers as at 23 September 2026, a dated snapshot and not a live feed. The resistor range is on the chip resistor page.

The short version

Across 114 Walsin export files there are 89 distinct names, 14 ordered layouts and one column that never moves. Column 8 carries 14 names; column 6 is Tolerance in 84 files and something else in 30, covering 106,665 of 390,463 rows. Chip resistor and thin film share nine names and agree on the position of one. Of 15 names used by more than one family, 14 move. A column called Thickness is empty on all 48,225 MLCC rows while T holds the thickness, and Packing is empty on 49,912 rows while Packaging is filled on 328,729. We will not tell you a field matches because the column number does. Qualification stays with you.

Send us the import script and one row from each family you load, and we will tell you which columns it reads from the wrong place. And a question back: when your thin film tolerance column came back as ≦±50, did you conclude the parts were loose?


Frequently asked questions

Is the column order the same in every Walsin export file?

No. Across the 114 files that hold parts, 89 distinct header names appear and 27 of them sit at more than one column number. Only Part Number holds the same column in all 114 files. Column 8 alone carries 14 different names, so a loader that reads by column number is wrong for most families.

What does column 6 hold in each Walsin file?

It holds Tolerance in 84 files, covering chip resistor, MLCC and varistor. In the other 30 files it holds a different quantity: a temperature coefficient in all 11 thin film files, inductance in the 2 inductor files, and frequency, insertion loss, function or standing wave ratio in the RF and antenna files.

Why do chip resistor and thin film exports have different column positions?

Because the chip resistor files carry a Series Group column at position 2 and the thin film files do not. The two families share nine header names, and eight of them sit one column to the left in thin film. Only Part Number is in the same position in both.

Are Packing and Packaging the same column in Walsin files?

They name the same quantity in different families. Packaging is used by the chip resistor and thin film files and is filled on all 328,729 rows. Packing is used by the MLCC, disc capacitor and varistor files and is empty on all 49,912 rows we counted, so only one of the two names carries data.

How many distinct column layouts do the Walsin exports use?

Fourteen ordered header lists across the 114 files that hold parts: one for chip resistor, two for MLCC, one for thin film, six for RF, and one each for inductor, varistor, disc capacitor and antenna. They range from 9 columns to 23, and the same name sits in different places across them.

Should I read Walsin exports by column name or column number?

By name, normalised, together with the product family. A column number is only safe for Part Number. We also check that a column is filled before relying on it, because Thickness, Packing, SPQ, MOQ, T.C. and TCC are declared and empty on every row we counted.


Share This, Choose Your Platform!