MLCC height tolerance changes the answer to a height budget. We parsed the thickness field of Walsin’s MLCC export: 4,653 rows at 0603 state a 0.80 mm nominal, and they name four different maximum heights, 0.87, 0.95, 1.00 and 1.10 mm. A nominal-height filter treats all four as one.
Venatech Limited, an authorised Walsin distributor in Hong Kong, holds the parametric export of Walsin Technology (Walsin) for all multilayer ceramic capacitors: 00_ALL_MLCC_parts_combined.xlsx with 24,026 rows, plus 22 per-series files. This post is written from the T column, which holds the thickness as printed text. We read the nominal and the upper tolerance from every string and added them. Every part number below is real. The question is what a design with a hard height limit should trust: the nominal, or the nominal plus its tolerance.
Does one nominal height in Walsin’s MLCC export mean one maximum height?
Usually, but not at the case sizes that matter most. We grouped every parseable row by case size and nominal thickness: 54 pairs. Seven of the 54 pairs carry more than one maximum height, and those seven cover 13,540 of the 23,552 parseable rows, 57.5 percent. The seven are 0201 at 0.30 mm, 0402 at 0.50, 0603 at 0.80, 0805 at 1.25, 1206 at 1.60, 1210 at 2.50 and 1812 at 2.50.
Of the 24,026 rows, 23,552 hold a thickness we could read as nominal plus tolerance. 127 hold the text ≤ 1.78 and 347 hold nothing. We have not seen this counted anywhere.
| Case size and nominal | Rows | Distinct maximum heights | Range of maxima | Series carrying it | Match / mismatch of “nominal only” against “nominal plus upper tolerance” |
|---|---|---|---|---|---|
| 0201 at 0.30 mm | 2,511 | 3 | 0.33 to 0.39 mm | 6 | Mismatch — 26 rows exceed 0.33 |
| 0402 at 0.50 mm | 3,944 | 3 | 0.55 to 0.70 mm | 10 | Mismatch — 138 rows exceed 0.55 |
| 0603 at 0.80 mm | 4,653 | 4 | 0.87 to 1.10 mm | 10 | Mismatch — 743 rows exceed 0.87 |
| 0805 at 1.25 mm | 1,258 | 3 | 1.35 to 1.55 mm | 9 | Mismatch — 250 rows exceed 1.35 |
| 1206 at 1.60 mm | 718 | 3 | 1.80 to 2.10 mm | 7 | Mismatch — 331 rows exceed 1.80 |
| 1210 at 2.50 mm | 291 | 4 | 2.70 to 3.10 mm | 7 | Mismatch — 288 rows exceed 2.70 |
| 1812 at 2.50 mm | 165 | 2 | 2.80 to 3.00 mm | 6 | Mismatch — 36 rows exceed 2.80 |
The 1210 line is the starkest: of 291 rows at 2.50 mm nominal, only 3 have the smallest maximum, 2.70 mm.
How far does the 0603 at 0.80 mm nominal spread?
By 0.23 mm, which is 29 percent of the nominal. We split the 4,653 rows by the exact thickness string. 3,910 rows, 84.0 percent, state 0.80 plus or minus 0.07 mm, a 0.87 mm maximum. 614 state plus 0.15 and minus 0.10, a 0.95 mm maximum. 40 state plus or minus 0.20, 1.00 mm, and 89 state plus or minus 0.30, 1.10 mm.
| Part number | Series | Capacitance | Rated voltage | Thickness string | Maximum height | Match / mismatch against MT18B102K500 |
|---|---|---|---|---|---|---|
MT18B102K500 |
Automotive MT | 1,000 pF | 50 Vdc | 0.80 ± 0.07 | 0.87 mm | reference |
MT18B224J500 |
Automotive MT | 220 nF | 50 Vdc | 0.80 + 0.15 / − 0.10 | 0.95 mm | Match on case size and nominal. Mismatch on maximum, by 0.08 mm |
0603X106K100 |
General Purpose | 10 µF | 10 Vdc | 0.80 ± 0.20 | 1.00 mm | Match on nominal. Mismatch on maximum, by 0.13 mm |
ST18B473K500 |
Automotive soft termination ST | 47 nF | 50 Vdc | 0.80 ± 0.30 | 1.10 mm | Match on nominal. Mismatch on maximum, by 0.23 mm |
All four are 0603 parts with the same nominal. The thickness symbol letter does not separate them on its own: that mapping is covered in the Walsin MLCC thickness symbol, 26 letters for 16 thicknesses.
Which case sizes spread the most, and by how much?
The larger the case, the wider the spread. We took the largest and smallest maximum within each of the seven pairs. The spread is 0.06 mm at 0201, 0.15 mm at 0402, 0.23 mm at 0603, 0.20 mm at 0805, 0.30 mm at 1206, 0.40 mm at 1210 and 0.20 mm at 1812. At 1206 the 1.60 mm nominal can mean a part as tall as 2.10 mm.
| Part number | Case size | Thickness string | Maximum height | Match / mismatch against a 1.60 mm nominal request |
|---|---|---|---|---|
MT31B224K101 |
1206 | 1.60 ± 0.20 | 1.80 mm | Match on nominal, reference maximum |
MT31B105K500 |
1206 | 1.60 + 0.30 / − 0.10 | 1.90 mm | Mismatch — 0.10 mm taller at maximum |
ST31B225K500 |
1206 | 1.60 ± 0.50 | 2.10 mm | Mismatch — 0.30 mm taller at maximum |
1210B106M500 |
1210 | 2.50 ± 0.20 | 2.70 mm | Match on nominal, reference for 2.50 mm |
1210B224K401 |
1210 | 2.50 ± 0.60 | 3.10 mm | Mismatch — 0.40 mm taller at maximum |
Which series write the wider tolerances?
The soft termination lines. In every pair where it appears, the widest tolerance belongs to the soft termination SH, SG or automotive ST series. At 0603 the 89 rows of plus or minus 0.30 mm are all soft termination; at 0805 the 40 rows of plus or minus 0.30 mm are all soft termination; at 1206 the 58 rows of plus or minus 0.50 mm are all soft termination. The wider tolerance is consistent with the extra layer under the termination, but the export does not say so, and we do not state a cause.
General Purpose parts hold the widest tolerance in their own family at 0603: 40 rows at plus or minus 0.20 mm. For what soft termination is and where it applies, see soft termination MLCC SH, SG, ST and VT.
What does a nominal-height filter miss against a height budget?
Rows that pass on paper and fail in the worst case. We tested the 4,672 parseable 0603 rows against three budgets. A nominal filter at 0.80 mm passes all 4,672. Against a 0.87 mm maximum only 3,929 pass, a loss of 743; against 0.95 mm, 4,543 pass, a loss of 129; against 1.00 mm, 4,583 pass, a loss of 89.
The reverse risk is a part rejected for a tolerance it never needs. A 0603 with a 0.95 mm maximum fits a 1.00 mm limit, but a filter that reads the nominal alone cannot tell it from one that does not. The method for working out the worst case is in thickness and tolerance, the two fields every cross tool asks you to check.
How should an MLCC height budget be checked?
Add the upper tolerance to the nominal before comparing, and keep the original string. Compare the maximum height from the export with the datasheet outline for the exact part number: 347 of 24,026 rows publish no thickness at all and 127 publish only a ceiling of 1.78 mm. Our free cross-reference tool lists thickness and tolerance as fields to double check and states differences from your original, so the height is stated, not ticked.
We hold 28 Walsin datasheets, 451 pages, covering all 16 MLCC series that have export rows; ask for the one covering your part number and it will be in your inbox. Confirm against both datasheets and qualify in your own circuit. Hong Kong stock is on the ready stock page, a dated snapshot of 1,125 part numbers as at 23 September 2026 and not a live feed. The capacitor range is on the MLCC page.
The short version
We parsed the thickness text of 24,026 rows; 23,552 read as nominal plus tolerance. Seven of 54 size-and-nominal pairs carry more than one maximum height, covering 13,540 rows. The 0603 at 0.80 mm names 0.87, 0.95, 1.00 and 1.10 mm; the 1206 at 1.60 mm runs to 2.10 mm. Soft termination series write the widest tolerances. A nominal filter at 0603 passes 4,672 rows; a 0.87 mm worst case passes 3,929. Add the tolerance before you compare.
Send us the height limit off your drawing and the part numbers you filtered to, and we will tell you which maximum each one states. And a question back: when your board has a 0.9 mm limit, was the number you checked the nominal?
Frequently asked questions
What is the maximum height of a Walsin 0603 MLCC at 0.80 mm nominal?
It depends on the row. Of 4,653 rows with that nominal, 3,910 state a maximum of 0.87 mm, 614 state 0.95 mm, 40 state 1.00 mm and 89 state 1.10 mm. The nominal is identical, so read the full thickness string for the exact part number.
How many Walsin MLCC rows publish a readable thickness?
23,552 of 24,026 state a nominal plus a tolerance we could read. Another 127 state only a ceiling of 1.78 mm, all in the Microwave RF series at 1111, and 347 state no thickness at all. Check the datasheet outline for those rows.
Which Walsin MLCC series have the widest height tolerance?
The soft termination series. At 0603 the 89 rows with plus or minus 0.30 mm are soft termination, at 0805 the 40 rows of plus or minus 0.30 mm are too, and at 1206 the 58 rows of plus or minus 0.50 mm are. General Purpose reaches plus or minus 0.20 mm at 0603.
Does the thickness symbol letter tell me the maximum height?
Not alone. The letter names a thickness class, and eight of 64 case-size and letter pairs name two thicknesses. Within one nominal, rows can still differ in upper tolerance. Use the thickness string for the exact part number and not the letter.
Is a filter on nominal height safe for a hard height limit?
No. At 0603, a nominal filter passes all 4,672 parseable rows, but only 3,929 are at or below 0.87 mm maximum. At 1210, 288 of 291 rows at 2.50 mm nominal exceed 2.70 mm. Add the upper tolerance first.
Can a part with a 0.95 mm maximum replace one with 0.87 mm?
We do not make that call for you. The two differ by 0.08 mm at maximum, and match on case size and nominal. Whether the extra height fits depends on your assembly. Confirm against both datasheets and qualify in your own circuit.