MLCC Temperature Cycling: Four Different Cycle Counts Across 26 Walsin Datasheets, and 1,000 Cycles Means 1,000 Hours on One Sheet and 167 on Another

An MLCC temperature cycling test runs 5, 10, 100 or 1,000 cycles depending on which of our 26 Walsin datasheets governs the part. We read every cycling row. Two sheets state 1,000 cycles and mean six-fold different soak times, because one dwells 30 minutes at each extreme and the other dwells 5.

Venatech Limited, an authorised Walsin distributor in Hong Kong, holds 28 Walsin Technology (Walsin) datasheets — 451 pages, 26 multilayer ceramic capacitor documents and 2 chip resistor documents. We extracted the text of every page and read the temperature-cycling and thermal-shock rows. A cycle count on its own is not a specification; the dwell, the step structure and the temperature extremes are what decide how hard the part was actually worked. Every part number below is real.

How many cycles does an MLCC temperature cycling test actually run?

Four different numbers, and the spread is 200-fold once you multiply by the dwell. 25 of our 26 MLCC documents state a cycling or thermal-shock row; only 00_MLCC_Application_Guide.pdf, 12 pages, states none — it covers soldering temperature difference instead. We counted the cycle figure on all 25.

The same five sheets that state 1,000 cycles also state a separate Thermal Shock row: 300 cycles, MIL-STD-202 Method 107, 15 ± 3 minutes at each extreme and a maximum transfer time of 20 seconds. No commercial series sheet in our library states a thermal-shock item at all — the only other occurrences of the phrase in 451 pages are soldering-preheat warnings in the mounting sections, which are advice, not a test.

Two of the 100-cycle series, Mega Cap M and Mega Cap V, have a full approval sheet and zero rows in Walsin’s parametric export, so no part number exists for us to quote or cross on either. The 100-cycle condition is real and currently unbuyable from our data.

Why does 1,000 cycles mean two different things on two Walsin sheets?

Because the dwell differs by a factor of six and nobody puts the dwell in the headline. Page 15 of 11_Automotive_MT.pdf states Temperature Cycling, JESD22 Method JA-104, as 1,000 cycles with a two-step profile: minimum operating temperature +0/−3 for 5 ± 1 minutes, then maximum operating temperature +3/−0 for 5 ± 1 minutes. No room-temperature step. Page 6 of 17_Safety_X1Y2_Automotive_S2_H.pdf states 1,000 cycles with a four-step profile: minimum for 30 ± 3 minutes, room temperature 2 to 3 minutes, maximum for 30 ± 3 minutes, room temperature 2 to 3 minutes.

Multiply it out. At 5 ± 1 minutes a side, 1,000 cycles holds the part at an extreme for 10,000 minutes — about 167 hours. At 30 ± 3 minutes a side, the same 1,000 cycles holds it for 60,000 minutes — 1,000 hours, roughly 42 days. We have not seen this counted anywhere. Both figures are correct, both sheets say “1,000 cycles”, and a selector table that carries the cycle count and drops the dwell makes them look identical.

The S2-H sheet also cites both standards side by side — IEC 60384-21/22 clause 4.11 and JESD22 Method JA-104 — which is why its profile keeps the IEC four-step shape while carrying the automotive cycle count. The MT, ST, VT and RT sheets cite JESD22 Method JA-104 alone and use the two-step shape throughout.

Which real part numbers sit on which cycling condition?

Nine, from nine different approval sheets, every one present in our master part-number set. We measured each against the longest soak in the library — S2-H’s 1,000 cycles at 30 ± 3 minutes a side — and stated the differences field by field rather than calling anything a like-for-like swap.

One absence is worth naming. We hold a 12-page approval sheet for the S2-H safety automotive series and the parametric export carries 173 safety rows under a single series name, all with S242, S243, S252 or S255 prefixes and no automotive marker. We cannot attribute a part number to the S2-H variant from the export, so the sheet with the longest soak in our library is also the one we cannot put a code against. That is a gap in the data we hold, not a statement about what Walsin makes.

Which Walsin datasheets state the actual temperatures they cycle to?

Three of 28. We checked all 26 MLCC documents and both resistor documents. 24 of the 25 MLCC cycling rows write the extremes as “Min. operating temp. +0/−3” and “Max. operating temp. +3/−0” — a pointer to the part’s own temperature characteristic, not a figure. Exactly one MLCC sheet prints absolute values: page 5 of 09_Microwave_Automotive_RT.pdf states step 1 at −55 °C +0/−3 and step 2 at +125 °C +3/−0, both for 5 ± 1 minutes. Both chip resistor sheets print absolutes too, at −55 °C to +155 °C.

This matters more than it reads. The pointer is unambiguous for a part whose dielectric fixes the range, but an X5R part and an X7R part on the same sheet do not cycle to the same temperature, and a selector that prints “1,000 cycles” against a series rather than a part number is flattening a real difference. For how the dielectric code sets that range, see NP0 against X7R against X5R.

How much capacitance drift is allowed after cycling, and why does more cycling allow more?

The limit widens as the cycle count rises, which is the opposite of the intuition. On the 17 five-cycle sheets and the two 100-cycle Mega Cap sheets the Class II limit is within ±7.5 %, with Class I at ±2.5 % or 0.25 pF. On the five 1,000-cycle sheets it is within ±10.0 %, and on the 300-cycle thermal shock row it is ±10.0 % again. The feed-through sheet is wider still on one of its two part families, at ±12.5 % for FT15X against ±7.5 % for FT21B.

Read that pairing carefully before you use it in an argument. A part held through 1,000 cycles is permitted a wider post-test drift than a part held through 5 — so “passes 1,000 cycles” and “passes 5 cycles” are not points on one scale, they are two different acceptance criteria. The honest comparison is cycles, dwell and limit together, and that is three columns, not one. The 1,000-cycle sheets are also the AEC-Q200 sheets, and AEC-Q200 is stated by Walsin as compliant and qualified to, never certified — see what AEC-Q200 actually certifies.

What should you ask for before you accept a cycling figure for your part?

The governing page, the dwell and the extremes — 3 questions, in that order, and all 3 before you write anything into a qualification plan. A cycle count on its own is the one number in this whole row that cannot be acted on, because 1,000 cycles spans 167 to 1,000 hours of soak across the 5 sheets that state it.

1. Ask which sheet and page govern your exact part number. We hold 28 Walsin datasheets, 451 pages. Ask for the one covering your part and it will be in your inbox, not in a brochure — and if no page covers it, as it does not for Mega Cap M, Mega Cap V or the S2-H variant in the export, you get that answer the same day.
2. Ask for the qualification report if the part is automotive. The report for your exact part number is what closes the gap between a stated condition and a result. We make no qualification claim of our own. The safety series also carries certificate numbers you can check without asking us: UL file E182369, TUV R50381780 and R50195920.
3. Cross the whole list in one pass. Paste up to 500 part numbers or drop your Excel or CSV into the free cross-reference tool, filter by commercial or automotive grade, and read every stated difference. Then check what is physically in Hong Kong on ready stock — 1,125 part numbers as at 23 September 2026, a dated snapshot and not a live feed. The full capacitor line sits on the Walsin MLCC page.

The short version

25 of our 26 Walsin MLCC datasheets state a cycling row and the count takes four values: 5 cycles on 17 sheets, 10 on the feed-through sheet, 100 on the two Mega Cap sheets, and 1,000 on the five AEC-Q200 sheets. Those same five add a separate 300-cycle thermal shock at 15 ± 3 minutes a side. The two sheets that both say 1,000 cycles differ six-fold in soak: 167 hours at an extreme on MT, ST, VT and RT against 1,000 hours on S2-H, because one dwells 5 ± 1 minutes and the other 30 ± 3. Only one MLCC sheet prints the actual temperatures; 24 point at the part’s own operating range. And the permitted Class II drift widens with the cycle count, from ±7.5 % at 5 cycles to ±10.0 % at 1,000. Qualification stays with you; our job is to make sure the cycle count you quote comes with the dwell that makes it mean something.

Send us the thermally cycled lines off your bill of materials and we will return the governing page, the cycle count, the dwell, the step structure and the post-test limit for each one — including the lines where no page or no part number exists. And a question back: when a datasheet tells you 1,000 cycles, do you know whether that is a week in the chamber or six?

Frequently asked questions