- Thermal cycling is a method-family question first, not a generic reliability badge.
- Public IPC methods do contain usable parameter examples, but they are method-scoped or application-scoped, not universal defaults.
- The release package should freeze the temperature profile, coupon or board representation, and failure-detection method before chamber time starts.
- A pass only proves the board survived the chosen screen; it does not prove field life.
Quick Answer
Thermal cycling test for PCB reliability should be reviewed as an interconnect-fatigue and material-mismatch test. If you need a concrete parameter table, use IPC method-scoped examples and label them clearly as such, rather than treating them as universal PCB rules. If the real risk is moisture or static heat, a different test is usually the better fit.
A PCB thermal cycling test moves bare boards, coupons or assembled boards repeatedly between a cold and a hot temperature in a chamber, while their resistance is monitored or they are inspected afterwards. It is the standard way to find plated through-hole cracks, microvia separation, solder joint fatigue and delamination before they appear in the field.
For the broader release-readiness workflow that connects DFM, test strategy, and reliability evidence before pilot or volume release, see the PCB Design for Manufacturing Guide.
Table of Contents
- What does the test actually measure?
- How is a PCB thermal cycling test run?
- What parameters can you actually publish?
- When does thermal cycling make sense?
- What should be frozen before chamber work starts?
- How should failure modes be read?
- Thermal cycling vs thermal shock vs humidity
- How should results be read?
- FAQ
- Public references
What does the test actually measure?
Thermal cycling stresses the board by repeatedly moving it through hot and cold extremes. The point is not just to count cycles. It is to expose how interconnects and materials behave when expansion and contraction repeat.
| Review axis | What to freeze | Why it matters | What it can reveal |
|---|---|---|---|
| Failure mechanism | Fatigue, crack growth, or delamination | The test only helps if the mechanism matches the risk | Interconnect or material weakness |
| Structure under test | PTH, solder joint, microvia, or layered stack | Different structures fail differently | The real weak link |
| Measurement method | Continuity, resistance change, or microsection | The report must detect the failure path, not just the end state | Opens, intermittents, hidden cracks |
| Scope boundary | Screening, comparison, or qualification context | The same cycle run can answer different questions | Whether the build is ready for the next gate |
How is a PCB thermal cycling test run?
The sequence is similar whether the sample is a coupon, a bare board or an assembly:
- Define the objective and the structure under test. For example: barrel cracking in 0.25 mm PTHs on a 2.4 mm board, or BGA solder joint fatigue.
- Precondition the samples. Usually a bake followed by reflow simulations, so the samples have seen the same thermal history as production boards (see the IPC-TM-650 2.6.7.2C example below).
- Set the chamber profile. Cold and hot extremes, ramp rate, dwell at each extreme and cycle count. Single-chamber cycling uses controlled ramps; two-chamber shock testing moves the samples between zones within seconds.
- Monitor during the test. Daisy-chain resistance logged continuously or at intervals, with an event detector if intermittent opens matter.
- Inspect after the test. Microsections through the failed and unfailed structures confirm where and how the crack formed. The PCB microsection analysis guide covers coupon placement and reading.
- Report against the frozen criteria. Resistance-change limit, cycles to first failure and failure location.
For solder joint reliability of assemblies, IPC-9701 defines its own test conditions. Its preferred reference condition (TC1) is 0 °C to +100 °C, and harsher conditions run down to −40 °C or −55 °C and up to +125 °C. As with the IPC-TM-650 values below, treat these as method examples and take the project values from the program's own requirement.
PCB temperature testing beyond cycling
"PCB temperature test" can mean several different tests. Thermal cycling is only one of them:
| Test | Typical purpose |
|---|---|
| Thermal cycling (single chamber, controlled ramp) | Interconnect and solder joint fatigue |
| Thermal shock (two-zone, fast transfer) | Faster screening of the same failure modes, plus abrupt-stress cracks |
| Thermal stress / solder float | Bare-board survival of soldering temperatures, checked by microsection |
| Interconnect stress test (current-induced heating, IPC-TM-650 2.6.26) | Fast via reliability comparison on coupons |
| High-temperature storage / operating life | Static heat aging of materials and components |
| Thermal imaging under power | Hot-spot check on a working assembly, not a reliability test |
What parameters can you actually publish?
Public IPC methods give you real parameter examples, but they come with context.
| Method-scoped example | Public value | How to read it |
|---|---|---|
| IPC-TM-650 2.6.26A default condition | 6 samples, 150 °C, 10% resistance-change threshold, 250 cycles | A method example for interconnect evaluation, not a universal PCB rule |
| IPC-TM-650 2.6.26A microvia condition | 6 samples, 190 °C, 10% resistance-change threshold, 250 cycles | A microvia example, not a default for every via structure |
| IPC-TM-650 2.6.26A survivability examples | 230 / 245 / 260 °C, 10 cycles, 10% threshold | A survivability lane, not a blanket reliability claim |
| IPC-TM-650 2.6.7.2C qualification/conformance example | 6 h bake at 105 to 125 °C, 6 reflow simulations, 100 cycles, 15 min dwell, 5% resistance-change limit | A board-level method example for qualification or quality conformance, not a sector-independent acceptance table |
When you cite these values in a project review or supplier discussion, label them as method examples or qualification conditions so nobody mistakes them for universal PCB defaults.
When does thermal cycling make sense?
Thermal cycling makes sense when the product will actually see repeated temperature swings in service.
- automotive electronics
- industrial hardware with repeated power cycling
- thick or high layer-count boards, where laminate Z-axis expansion drives barrel stress (a High-Tg PCB with lower Z-CTE is the usual first mitigation)
- HDI or microvia structures
- mixed-material boards with copper, ceramic, or laminate mismatch
- rigid-flex or other assemblies with multiple stress boundaries
If the main risk is moisture ingress, corrosion, or static heat aging, humidity testing or high-temperature storage is usually the better first choice.
Practical reading rule
A board can pass electrical bring-up and still fail thermal cycling because the weakness sits in the interconnect, not the logic. A plated through hole can look fine until repeated expansion and contraction opens a crack in the barrel. A solder joint can also work at room temperature and still fatigue at the pad edge after enough hot-cold repeats.
That is why the result should always be read with the failure mode in mind.
What should be frozen before chamber work starts?
Freeze these items before the test begins:
- test objective
- temperature profile
- ramp rate
- dwell time
- cycle count
- coupon or board representation
- detection method for opens and intermittents
| Frozen item | Why it matters |
|---|---|
| Test objective | Defines whether the run is screening, comparison, or qualification support |
| Temperature profile | Sets the stress severity and the failure mode being exercised |
| Ramp rate | Helps distinguish cycling from shock-like behavior |
| Dwell time | Affects thermal soak and material response |
| Cycle count | Changes how much accumulated fatigue is visible |
| Detection method | Determines whether intermittent failures are actually seen |
How should failure modes be read?
The strongest value of thermal cycling is the failure story.
| Failure mode | What it usually points to | What to review first |
|---|---|---|
| PTH barrel crack | Z-axis mismatch and plating stress | Hole wall integrity and laminate behavior |
| Solder-joint fatigue | Component-to-board mismatch | Pad geometry, package mass, and support |
| Microvia separation | Local interconnect stress | Build-up structure and interface quality |
| Delamination | Material weakness or weak bonding | Resin system, moisture history, and stack behavior |
That makes the test more useful as an engineering feedback loop than as a slogan.
Thermal cycling vs thermal shock vs humidity
The distinction matters.
| Test type | What it is for | What it is not for |
|---|---|---|
| Thermal cycling | Repeated expansion and contraction | Moisture-only risk |
| Thermal shock | Faster temperature change and abrupt stress | Slow fatigue screening |
| Humidity testing | Moisture ingress and corrosion risk | Mechanical fatigue proof |
| High-temperature storage | Static heat aging | Temperature swing fatigue |
IPC TM-650 public methods support this vocabulary at method level, but the exact severities and limits still come from the program, drawing, or licensed standard text. For bare boards, the performance class on the drawing (for example IPC 6012 Class 3) decides which thermal stress and coupon evaluations are required at all.
How should results be read?
Look for:
- where the first intermittent appeared
- whether the failure was an open, a resistance rise, or a visible crack
- whether the same structure failed repeatedly
- whether microsectioning confirmed the suspected path
- whether the result points to a board-level mismatch or a program-level boundary issue
That matters because a pass only proves the board survived the plan that was actually run.
FAQ
Is thermal cycling the same as thermal shock?
No. Thermal cycling is about repeated fatigue. Thermal shock is about much faster temperature change.
What is the difference between a PCB thermal cycling test and a thermal stress test?
A thermal stress test (for example a solder float) exposes the bare board to soldering temperature once or a few times and then checks it by microsection. Thermal cycling repeats a hot-cold profile many times to accumulate fatigue. The first checks whether the board survives assembly; the second checks whether it survives use.
Does a pass prove long-term field life?
No. It only proves the board survived the defined screen.
Why are coupons used?
They let the test focus on a known failure path before a full product build is exposed.
Should humidity be tested separately?
Yes, if moisture or corrosion is the real concern.
Public references
- IPC-TM-650 2.6.26A D.C. Current Induced Thermal Cycling Test for Interconnect Evaluation
- IPC-TM-650 2.6.7.2C Thermal Shock, Thermal Cycle and Continuity
- IPC test methods
Next steps
If the board is still under design, decide whether the main risk is fatigue, moisture, or static heat before you choose a test plan.
If you need help deciding whether thermal cycling is the right evidence layer, send the stackup, intended use profile, suspected failure mode, and validation questions to [email protected] or upload the stackup and test plan for a reliability review. Assembly-level reliability screens are covered on our PCBA reliability testing page. APTPCB's engineering team can help determine whether the real gap sits in test selection, coupon or sample definition, dwell profile, or failure-reading method.
Conclusion
Thermal cycling is useful when you want to expose fatigue from repeated temperature change. It works best when the objective, profile, and failure-reading method are defined before the chamber starts.