Surface Insulation Resistance (SIR)

Surface Insulation Resistance (SIR)

Quality & Inspection Standards

PRODUCTS CENTER

Get Custom Quote

PRODUCTS CENTER

Supported formats: PDF, DWG, Gerber, Excel (Max 50MB)
Submit
Cleanliness Testing IPC‑TM‑650 Methods 2.6.3.3 / 2.6.3.5

Surface Insulation Resistance (SIR) Testing – IPC‑TM‑650 Methods, Reliability Limits & Process Qualification

Surface Insulation Resistance (SIR) is the only test that directly measures the functional reliability of a PCB under humidity and voltage bias – not just ionic load or conductivity. ROSE and IC tell you how much contamination is present. SIR tells you whether that contamination will cause failure in the field. We have seen boards pass ROSE at 1.2 µg/cm² but fail SIR after 168 hours at 85°C/85%RH. The root cause was a combination of marginal cleanliness and an aggressive flux that left hygroscopic residues. Per IPC‑6012E §3.9.1, SIR testing is required for qualification of soldering fluxes (J‑STD‑004) and can be invoked for PCB acceptance. This guide covers IPC‑TM‑650 2.6.3.3 (flux SIR) and 2.6.3.5 (bare board SIR), test patterns, failure criteria, and how to use SIR as a process validation tool.

Published: May 15, 2026  |  10 min read  |  #SIR #SurfaceInsulationResistance #FluxQualification

1. What Is SIR? – Functional Cleanliness vs Analytical Methods

Surface Insulation Resistance (SIR) is a direct electrical measurement of the resistance between two adjacent conductors (e.g., comb pattern fingers) under controlled temperature, humidity, and applied voltage. A low SIR value indicates that ionic contamination has created a conductive path, even if the total ionic load (by ROSE) appears acceptable.

Why SIR is essential:

  • Functional, not just analytical: SIR measures actual electrical behavior, not extracted ions. It captures the combined effect of all residues (ionic, hygroscopic, partially soluble).
  • Simulates field conditions: High humidity (85‑90% RH) and bias voltage accelerate electrochemical migration.
  • Process validation: SIR is required for qualifying fluxes per J‑STD‑004 and can be used to qualify cleaning processes.

In one automotive program, the supplier changed to a new no‑clean flux. ROSE results were stable at 1.1 µg/cm². However, SIR testing at 85°C/85%RH for 168 hours showed resistance dropping from 1×10¹⁰ Ω to 5×10⁷ Ω – below the 1×10⁸ Ω limit. The flux was requalified and rejected. ROSE alone would have missed the issue.

Inspection consideration: SIR testing evaluates whether ionic residues, moisture and process chemistry cause insulation resistance to degrade under environmental stress. The result should be reviewed together with the product construction, contamination source and qualification requirements. See our Inspection & Testing Capabilities for PCB and PCBA reliability verification support.

2. SIR Test Methods: 2.6.3.3 (Flux) vs 2.6.3.5 (Bare Board)

IPC‑TM‑650 defines two primary SIR methods. The choice depends on what you are testing.

Method What It Tests Sample Duration Typical Use
2.6.3.3 ("Flux SIR") Flux residues after reflow or wave solder Comb pattern test board (e.g., IPC‑B‑24) with applied flux 168 hours (7 days) Flux qualification per J‑STD‑004
2.6.3.5 ("Bare Board SIR") Cleanliness of bare PCBs (after fabrication) Bare board with comb pattern (e.g., IPC‑B‑25) 168‑500 hours PCB lot acceptance, process qualification

Table 1: SIR methods. For board acceptance, use 2.6.3.5. For flux qualification, use 2.6.3.3.

Method 2.6.3.5 is often specified by high‑reliability customers (aerospace, medical) for periodic validation. It is more severe than ROSE because it applies bias voltage (typically 50‑100V DC) and high humidity (85‑88% RH) for extended periods (168‑500 hours). A board that passes ROSE at 1.2 µg/cm² may still fail SIR if the ionic residues are hygroscopic and concentrated in fine gaps.

Assembly process connection: When SIR degradation is linked to solder paste, flux, reflow or cleaning chemistry, the corrective action belongs in the assembly process—not only in final inspection. See our SMT assembly capabilities for process and cleanliness control support.

3. Test Patterns – Comb Patterns, IPC‑B‑24, and IPC‑B‑25

SIR testing uses interdigitated comb patterns with defined line/space dimensions. The sensitivity to contamination scales with gap size – finer gaps are more susceptible to leakage.

  • IPC‑B‑24: Standard comb pattern used for flux qualification per J‑STD‑004. Typical line/space = 0.4mm (16mil) or 0.635mm (25mil). Used with method 2.6.3.3.
  • IPC‑B‑25: More aggressive comb pattern with finer lines/spaces (e.g., 0.2mm, 0.3mm). Used for bare board SIR (2.6.3.5) and high‑reliability applications.
  • Customer‑specific patterns: Some OEMs specify comb patterns that mimic their actual circuit geometries (e.g., BGA pitch, adjacent via spacing).

In one program, the customer required SIR testing on a comb pattern with 0.15mm lines and 0.15mm spaces (simulating BGA pitch). The board passed ROSE at 1.0 µg/cm² but failed SIR (resistance <1×10⁸ Ω) after 96 hours. The fine spacing trapped residues that would have been harmless at 0.4mm spacing. The cleaning process was tightened, and the board passed. Match the test pattern to your product's geometry.

4. Test Conditions – Temperature, Humidity, Bias & Duration

Typical SIR test conditions per IPC‑TM‑650 2.6.3.5:

  • Temperature: 85°C ± 2°C (high‑temperature humidity) or 40°C ± 2°C (ambient). 85°C is more severe.
  • Relative humidity: 85% ± 3% RH. Lower humidity (e.g., 50% RH) is used for some special cases but not typical for SIR.
  • Bias voltage: 50V DC (typical), 100V DC (severe), or 5V (for fine‑pitch low‑voltage circuits). The bias is applied continuously.
  • Measurement voltage: 50V (same as bias, or 5V for low‑voltage measurement).
  • Duration: Minimum 168 hours (7 days). Some specifications require 500 hours or more.
  • Measurement intervals: Typically at 24, 48, 96, 168 hours, then every 24 hours thereafter.

The board is placed in a humidity chamber with the comb pattern biased. A data logger measures resistance at defined intervals. During the test, the chamber must maintain stable conditions – a 5% RH drop or 2°C rise can dramatically affect results.

Acceptance criteria (common industry limits):

  • Initial (dry) resistance: Typically >1×10¹² Ω.
  • After 168 hours at 85°C/85%RH with bias: Resistance must remain ≥1×10⁸ Ω (100 MΩ) for most specifications. Aerospace often requires ≥1×10⁹ Ω (1000 MΩ).
  • No evidence of dendritic growth or CAF: Visual inspection after test.

In one lab, we saw a chamber that could not maintain 85% RH at 85°C – it fluctuated between 75% and 88%. The SIR results were inconsistent. After recalibrating the chamber, the results stabilized. Chamber control is as important as the board cleanliness.

5. Acceptance Criteria – SIR Limits (1×10⁸ Ω, 1×10⁹ Ω, etc.)

IPC‑6012E does not specify a numeric SIR limit for bare boards – it references IPC‑TM‑650 2.6.3.5 and J‑STD‑004 for fluxes. However, industry practice has established clear thresholds.

Application SIR Limit (after 168h) Bias Voltage Test Pattern Reference
J‑STD‑004 flux qualification ≥1×10⁸ Ω 50V IPC‑B‑24 IPC‑TM‑650 2.6.3.3
Telecom / Server (bare board) ≥1×10⁸ Ω 50V IPC‑B‑25 or finer Common OEM spec
Automotive (high‑reliability) ≥1×10⁹ Ω 50V Custom comb (0.2mm) AEC‑Q100 derived
Aerospace / Military ≥1×10⁹ Ω 100V IPC‑B‑25 or finer IPC‑6012ES
Medical implantable ≥1×10¹⁰ Ω 5V (low voltage) Fine (0.1mm) Customer spec

Table 2: Common SIR acceptance limits. Always confirm with customer – there is no single IPC limit for bare board SIR.

SIR failures are not binary. A gradual drop from 1×10¹⁰ to 1×10⁸ is concerning but may be accepted in some programs. A drop below 1×10⁶ Ω is a hard failure indicating electrochemical migration or dendritic growth. We recommend defining a "alert limit" (e.g., 5×10⁸) and a "reject limit" (e.g., 1×10⁸) in your specification.

6. Correlation with ROSE/IC – When SIR Fails but Cleanliness Passes

SIR failures can occur even when ROSE and IC indicate acceptable cleanliness. Common reasons:

  • Hygroscopic residues: Some flux activators (e.g., organic acids) absorb moisture at high humidity, forming a conductive film even at low ionic concentration. ROSE may measure 0.5 µg/cm², but SIR fails.
  • Localized contamination: ROSE measures average contamination over the whole board. A small, high‑concentration residue under a BGA may cause SIR failure but barely affect the average ROSE reading.
  • Incomplete extraction: The solvent used for ROSE (75/25 IPA/DI) may not extract all residues. SIR uses humidity and bias, which can activate residues that are not soluble in the ROSE solvent.
  • Electrochemical migration (dendrites): Even very low levels of silver or tin ions can grow dendrites under bias, causing a short. ROSE and IC may not detect these metal ions if they are not extracted.

In one study, we tested 20 boards with ROSE values between 0.8 and 1.2 µg/cm². Eight passed SIR, twelve failed. The failing boards had higher chloride/bromide ratios and were processed with a different flux that left hygroscopic residues. ROSE alone is not sufficient to guarantee SIR performance. For high‑reliability applications, run SIR on a sampling basis (e.g., first article, quarterly) to validate the process.

When SIR failure needs root-cause review: A failed SIR result may originate from ionic contamination, solder mask condition, cleaning chemistry, moisture exposure, spacing or electrochemical migration. If the failure mechanism is not clear, submit the test conditions, resistance data and available failure evidence for an engineering contamination and reliability review.

SIR Testing Checklist (IPC‑Anchored)

  1. Define SIR method and limits in PO. Specify 2.6.3.3 (flux) or 2.6.3.5 (bare board), test pattern, duration, bias voltage, and pass/fail resistance.
  2. Use the correct test pattern. IPC‑B‑24 for flux; IPC‑B‑25 or finer for bare board. Match line/space to your product's minimum spacing.
  3. Pre‑condition boards (bake at 105°C for 2 hours) to remove moisture. Do not test as‑received – moisture can cause false failures.
  4. Calibrate the humidity chamber before each test. Verify temperature and humidity at multiple points.
  5. Apply bias voltage continuously. Measure resistance at defined intervals (e.g., 24, 48, 96, 168 hours).
  6. Inspect for dendrites or CAF after test. Use 50‑100x magnification. Any visible growth is a failure.
  7. Document SIR curves. A plot of resistance vs time shows degradation trends. Compare with historical data.

7. FAQ

Q: What is the difference between SIR and insulation resistance (IR) testing?

A: IR testing is typically performed dry (room temperature, no humidity) and measures bulk insulation. SIR is performed under controlled humidity (e.g., 85% RH) and often with bias voltage, simulating field conditions. SIR is much more sensitive to ionic contamination.

Q: How long does SIR testing take?

A: Minimum 168 hours (7 days) per IPC methods. Some programs require 500 hours (21 days) or longer. Planning SIR testing requires scheduling chamber time well in advance.

Q: Can a board pass SIR but still have ionic contamination?

A: Yes – if the contamination is not hygroscopic or does not lower resistance under the test conditions. However, SIR passes does not guarantee long‑term reliability; some residues may degrade over longer exposure or under different humidity/temperature. SIR is a screen, not a guarantee.

Q: Is SIR required for every PCB lot?

A: No – SIR is expensive and time‑consuming. It is typically run on a sampling basis (first article, quarterly validation, or after process changes). For high‑volume production, ROSE is used for lot acceptance, and SIR is used for process qualification.

Related Manufacturing Capability

Inspection & Testing Capabilities

PCB and PCBA electrical reliability, cleanliness verification and process-quality support for prototype and production builds.

If SIR degradation, electrochemical migration, moisture exposure, flux residue or contamination-related failure is a concern, submit the test conditions and available data for an engineering review.

Request SIR Engineering Review →

References: IPC‑TM‑650 2.6.3.3, 2.6.3.5; IPC‑6012E §3.9.1; J‑STD‑004; IPC‑B‑24, IPC‑B‑25. IPC‑TM‑650 Official Method.

Get Custom Quote

PRODUCTS CENTER

Supported formats: PDF, DWG, Gerber, Excel (Max 50MB)
Submit