A 0603 X7R 100 nF 50 V specification does not cross to one Walsin Technology (Walsin)
part. It crosses to 13, across six series. They split into two dimensional tolerance bands
— eight at +0.15/−0.10 mm and five at ±0.30 mm — which at worst case is a 0.95 mm part
against a 1.10 mm one. The rest is grade and termination.
This is what an MLCC cross reference looks like underneath the headline. Venatech Limited,
an authorised Walsin distributor in Hong Kong, holds an extract of 24,026 Walsin MLCC part
numbers. Filtering it on 0603, X7R, 100 nF and 50 Vdc returns 13 rows. Every part number
below came out of that query.
How many Walsin parts share a 0603 X7R 100 nF 50 V specification?
Thirteen, spread over six series. A cross tool returns one of them, correctly, as the
closest match — and it is the closest match. The other twelve are not wrong answers. They
are the same headline specification built to different grades, terminations and dimensional
tolerances.
| Walsin part | Series | Capacitance tolerance | L × W × T (mm) | Match / mismatch vs a 0603 X7R 100 nF ±10% 50 V line |
|---|---|---|---|---|
0603B104K500 |
General Purpose | ±10% | 1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Match on all five headline fields |
0603B104J500 |
General Purpose | ±5% | 1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Mismatch — tolerance is tighter |
0603B104M500 |
General Purpose | ±20% | 1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Mismatch — tolerance is wider |
MT18B104K500 |
Automotive MT, AEC-Q200 | ±10% | 1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Match on five fields, mismatch on grade — AEC-Q200 |
MT18B104J500 |
Automotive MT, AEC-Q200 | ±5% | 1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Mismatch — tighter tolerance, AEC-Q200 grade |
MT18B104M500 |
Automotive MT, AEC-Q200 | ±20% | 1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Mismatch — wider tolerance, AEC-Q200 grade |
MG18B104K500 |
Automotive MG, no AEC-Q200 | written -0.1 |
1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Match on dimensions, tolerance written in a non-percentage form |
MG18B104J500 |
Automotive MG, no AEC-Q200 | written -0.05 |
1.60+0.15/−0.10 × 0.80+0.15/−0.10 × 0.80+0.15/−0.10 | Mismatch — tighter tolerance, written in a non-percentage form |
ST18B104K500 |
Automotive soft term ST, AEC-Q200 | ±10% | 1.60±0.30 × 0.80±0.30 × 0.80±0.30 | Match on five fields, mismatch on dimensional tolerance and termination |
ST18B104M500 |
Automotive soft term ST, AEC-Q200 | ±20% | 1.60±0.30 × 0.80±0.30 × 0.80±0.30 | Mismatch — wider tolerance, wider dimensions, soft termination |
SG18B104K500 |
Soft Termination SG, Cu-poly | ±10% | 1.60±0.30 × 0.80±0.30 × 0.80±0.30 | Match on five fields, mismatch on dimensional tolerance and termination |
SH18B104K500 |
Soft Termination SH, Ag-poly | ±10% | 1.60±0.30 × 0.80±0.30 × 0.80±0.30 | Match on five fields, mismatch on dimensional tolerance and termination |
SH18B104J500 |
Soft Termination SH, Ag-poly | ±5% | 1.60±0.30 × 0.80±0.30 × 0.80±0.30 | Mismatch — tighter tolerance, wider dimensions, soft termination |
Unique to Venatech’s data. One headline MLCC specification — 0603, X7R, 100 nF,
50 Vdc — returns 13 Walsin part numbers across six series in our extract of Walsin’s own
parametric data: General Purpose, Automotive MT, Automotive MG, Automotive soft
termination ST, Soft Termination SG, Soft Termination SH. The count and the series split
are not published anywhere. Re-derived from the 24,026-row extract before publication.
Datasheets, one per series:
General Purpose ·
Automotive MT ·
MG series ·
Soft term ST ·
Soft term SG ·
Soft term SH
What separates the 13 from each other?
Three axes: capacitance tolerance, qualification grade and termination construction. The
headline specification fixes case size, dielectric, capacitance and voltage. It says nothing
about the other three, and those three are what a design engineer actually chooses between.
Capacitance tolerance splits the 13 into ±5%, ±10% and ±20% groups, plus the two MG parts
whose tolerance is written in a different form. Grade splits them into commercial,
automotive with AEC-Q200 qualification, and automotive without it. Termination splits them
into standard metal terminations and conductive-polymer soft terminations.
No cross-reference tool can make those calls. They depend on where the part sits on the
board, whether the panel gets depanelled after reflow, and what the qualification plan
demands. That is not a deficiency in anybody’s software — it is a design decision that needs
a designer.
Which of the 13 differ dimensionally?
Five of them. The three soft-termination series — ST, SG and SH — publish their 0603 body
dimensions as 1.60 ± 0.30 × 0.80 ± 0.30 × 0.80 ± 0.30 mm. The other eight publish
1.60 +0.15/−0.10 × 0.80 +0.15/−0.10 × 0.80 +0.15/−0.10 mm. Same nominal size, different
band.
| Dimension | Group A: General Purpose, MT, MG — 8 parts | Group B: ST, SG, SH — 5 parts | Match / mismatch |
|---|---|---|---|
| Length, nominal | 1.60 mm | 1.60 mm | Match |
| Length, range | 1.50 – 1.75 mm | 1.30 – 1.90 mm | Mismatch — band is 0.35 mm wider |
| Width, nominal | 0.80 mm | 0.80 mm | Match |
| Width, range | 0.70 – 0.95 mm | 0.50 – 1.10 mm | Mismatch — band is 0.35 mm wider |
| Thickness, nominal | 0.80 mm | 0.80 mm | Match |
| Thickness, worst case | 0.95 mm | 1.10 mm | Mismatch — 0.15 mm taller at worst case |
| Thickness, minimum | 0.70 mm | 0.50 mm | Mismatch — 0.20 mm thinner at minimum |
Both bands are published by Walsin and both are correct for their series. The mismatch is
between the two bands, not between a band and reality.
Why does 0.15 mm of worst-case thickness matter?
Because 0.15 mm is a nozzle height, a stencil aperture and sometimes an enclosure clearance.
A 0603 part specified at 0.80 mm nominal can arrive at 0.95 mm or at 1.10 mm depending on
which of the 13 you bought, and nothing in the BOM line tells you which.
Three places it bites. Placement: a pick-and-place recipe set for one thickness band may not
pick reliably from the other. Height-restricted assemblies: a part under a shield can or
inside a slim housing has a budget measured in tenths of a millimetre. Process rules: Walsin
permits wave soldering for 0603, 0805 and 1206 only below 1 mm, so a part whose worst case
is 1.10 mm needs its actual thickness checked against that rule rather than assumed from the
case code.
This is precisely the check Walsin’s own cross-reference tool tells its users to make. Step
four of its instructions reads: double check the thickness and tolerance of the suggested
part to ensure accuracy. A manufacturer that sends you to the drawing is doing you a favour,
and the 13-part table above is the reason why.
The minimum matters too, and gets discussed less. Group B can be 0.50 mm thick, 0.30 mm
under nominal. Where a thin part sits low on the pads, solder volume and fillet shape change
with it.
Why are two of the 13 hard to find with a filter?
Because the MG series writes its capacitance tolerance as -0.05 and -0.1 rather than
±5% and ±10%. A numeric or string filter on tolerance drops both MG parts silently. They are
real, orderable Walsin parts that a tolerance filter will not return.
MG18B104J500 carries -0.05 and MG18B104K500 carries -0.1 in Walsin’s parametric
data. Read them alongside the MT parts and the pattern is obvious: J is the ±5% code and
K is the ±10% code in the part number itself, exactly as in every other series. The part
number is consistent; the tolerance column is not.
The MG series is the automotive line without AEC-Q200 certification, so anyone building a
cost-down automotive variant is in exactly the position of needing these two parts and
exactly the position of filtering them out. Read the part number, not just the column — see
How to read a Walsin MLCC part number.
Which of the 13 are qualified to AEC-Q200?
Five, by the series names in Walsin’s own data. The three Automotive MT parts and the two
Automotive soft-termination ST parts sit in series that state AEC-Q200 qualification. The
two MG parts sit in the automotive series that states it is without AEC-Q200 certification.
The remaining six state neither.
| Group | Parts | Series statement on qualification |
|---|---|---|
| AEC-Q200 automotive | MT18B104J500, MT18B104K500, MT18B104M500 |
Automotive Caps Qualified to AEC-Q200 (MT series) |
| AEC-Q200 automotive, soft termination | ST18B104K500, ST18B104M500 |
Automotive Soft term Caps Qualified to AEC-Q200 (ST series) |
| Automotive, not AEC-Q200 | MG18B104J500, MG18B104K500 |
Automotive Caps Without AEC-Q200 Certification (MG series) |
| Commercial general purpose | 0603B104J500, 0603B104K500, 0603B104M500 |
General Purpose Caps |
| Soft termination, commercial | SG18B104K500, SH18B104J500, SH18B104K500 |
Soft Termination Capacitors (SG Cu-poly, SH Ag-poly) |
If the board is automotive, that table removes eight of the 13 in one step. If it is not,
it removes five. Either way it is a bigger cut than the tolerance column makes.
Which of the 13 does Venatech hold in Hong Kong?
Three of them appear in our stock snapshot dated 23 September 2026: 0603B104K500,
0603B104M500 and SH18B104K500. That snapshot covers 1,125 part numbers. It is a
snapshot of part numbers, not a standing quantity, and it moves.
All three appear with a CT packaging suffix on the order code — 0603B104K500CT,
0603B104M500CT, SH18B104K500CT — which the catalogue code does not carry. 865 of the
1,125 lines in that snapshot are the same way round. A catalogue code pasted into a purchase
order is not orderable as it stands.
How do I narrow 13 candidates down to one?
In four questions, in this order. Is the board automotive and does it need AEC-Q200? Does it
flex, get depanelled or take board-level mechanical stress? What capacitance tolerance does
the circuit actually need? And what is the height budget at that location?
Answer those and the 13 usually collapse to one or two. Automotive plus AEC-Q200 plus no
flex concern points at the MT series. Flex concern plus AEC-Q200 points at ST. Flex concern
without an automotive requirement points at SG or SH. Neither concern points at General
Purpose, which is where a cross tool will usually land you first, and quite reasonably.
None of that makes the crossed part approved. A specification match against Walsin’s
published catalogue is a candidate for your evaluation. Check both datasheets, confirm the
dimensional tolerance band against your own process window, and qualify the part in your
application before production use.
FAQ
How many Walsin parts match a 0603 X7R 100 nF 50 V specification?
Thirteen, across six series, in Venatech’s extract of 24,026 Walsin MLCC part numbers. They differ by capacitance tolerance, qualification grade and termination construction. Three are General Purpose, three Automotive MT, two Automotive MG, two Automotive soft termination ST, one SG and two SH.
Which Walsin MLCC series have a ±0.30 mm dimensional tolerance at 0603?
The three soft-termination series: ST, SG and SH. They publish 1.60 ± 0.30 × 0.80 ± 0.30 × 0.80 ± 0.30 mm. The General Purpose, MT and MG series publish 1.60 +0.15/−0.10 × 0.80 +0.15/−0.10 × 0.80 +0.15/−0.10 mm at the same nominal size.
What is the worst-case thickness of a 0603 Walsin MLCC?
It depends on the series. Eight of the 13 parts at 0603 X7R 100 nF 50 Vdc top out at 0.95 mm and five top out at 1.10 mm. Both are nominally 0.80 mm. Check the figure against your placement recipe and any height restriction at that board location.
Why does a tolerance filter miss some Walsin MG parts?
Because the MG series writes capacitance tolerance as -0.05 and -0.1 in Walsin’s parametric data rather than as ±5% and ±10%. A numeric or string filter on tolerance drops those rows silently. The part numbers still carry the standard J and K tolerance codes, so read the part number as well as the column.
Which of the 13 are AEC-Q200 qualified?
Five, by the series names in Walsin’s own data: the three MT parts and the two ST parts. The two MG parts sit in the automotive series that states it is without AEC-Q200 certification. The three General Purpose and three SG/SH soft-termination parts state neither way; confirm grade against the datasheet.
Does a soft termination part fit the same footprint as a general purpose one?
The nominal body size is the same at 0603, but the dimensional tolerance band is wider — ±0.30 mm against +0.15/−0.10 mm — so the extremes differ in length, width and thickness. Check the land pattern against the worst case, not the nominal, before assuming the footprint carries over.
Is a 13-result cross a problem with the cross tool?
No. A cross tool is built to return the closest single match, and it does. The remaining choice is qualification grade, termination and tolerance, which are design decisions rather than matching problems. Walsin’s tool names the part; a distributor and a designer settle the rest between them.
Are any of the 13 held in Hong Kong?
Three appear in Venatech’s stock snapshot dated 23 September 2026: 0603B104K500, 0603B104M500 and SH18B104K500. All three carry a CT packaging suffix on the order code. That is a part-number snapshot rather than a standing quantity, so confirm current availability before planning a build around it.
Next step
Send the specification, not just the part number, and get the whole candidate list back
rather than one row of it.
Cross a part with VenaMatch · Talk to a person at Venatech ·
Browse the Walsin MLCC range · Read
Soft termination MLCC: SH, SG, ST and VT and
AEC-Q200 MLCC: MT vs MG series.