What Is an Infrared Core Module?

An infrared core module (IR core module) is the primary imaging unit inside every thermal camera — and the single component that most directly governs system sensitivity, resolution, and dynamic range. It is not a bare detector chip: it is a fully integrated assembly combining a detector array, readout integrated circuit (ROIC), signal processing electronics, and a digital output interface in one sealed package. Selecting the right infrared core module matters more than any other hardware decision in a thermal imaging design.


How Does an Infrared Core Module Work?

From photon to pixel, an infrared core module processes radiation through four functional layers:

  1. Detector Array — Converts incident infrared radiation into an electrical signal. Uncooled designs use vanadium oxide (VOx) microbolometers, which respond to the 8–14 μm long-wave infrared (LWIR) band. Cooled designs use mercury cadmium telluride (HgCdTe/MCT) or indium antimonide (InSb) to cover mid-wave infrared (MWIR, 3–5 μm) and short-wave infrared (SWIR, 0.9–1.7 μm) through quantum-effect detection rather than thermal absorption.

  2. Readout Integrated Circuit (ROIC) — Mounted directly behind the detector array, the ROIC integrates, amplifies, and multiplexes the analog signal from every pixel before feeding it to an analog-to-digital converter (ADC). ROIC architecture strongly influences full-well capacity, dynamic range, and fixed-pattern noise floor.

  3. Packaging and Thermal Control — Uncooled cores use hermetically sealed enclosures (typically dry-nitrogen filled) to protect the microbolometer from humidity and contamination. Cooled cores require a Stirling-cycle cryocooler that holds the detector at approximately 77 K (−196 °C) to suppress thermal noise below the scene signal level.

  4. Digital Signal Processor (DSP) Board — Executes non-uniformity correction (NUC), bad-pixel replacement, and temporal filtering, then outputs a standardized video stream over LVDS, MIPI CSI-2, or Camera Link.

These four layers explain why the core module — not the lens or enclosure — is the performance-defining subsystem of any thermal imaging design.


Uncooled vs. Cooled Infrared Core Modules: Key Differences

The choice between an uncooled and a cooled core is the most consequential specification decision in thermal camera development. The table below captures the principal trade-offs:

Parameter Uncooled (LWIR) Cooled (MWIR / SWIR)
NETD 30–50 mK < 10 mK
Module Power < 3 W 8–35 W (including cryocooler)
Start-up Time Instant 3–8 min
Cryocooler MTBF N/A 8,000–20,000 h
Typical Pixel Pitch 12 μm, 17 μm 10–15 μm
Relative Cost Baseline 3–15× baseline

Selection logic: Where size, weight, and power (SWaP) budgets are tight — as in airborne and UAV platforms or power-line inspection systems — uncooled LWIR is the mainstream choice. When an application demands detection of temperature differences below 20 mK, MWIR spectral selectivity, or reliable tracking of fast-moving targets, only a cooled MWIR core can meet the requirement, at the cost of higher procurement price, longer warm-up time, and a finite cryocooler service life.


Key Parameters to Evaluate in an Infrared Core Module

Pixel Pitch

Pixel pitch is the center-to-center spacing between adjacent pixels on the detector array. The current uncooled LWIR mainstream is 12 μm — as used in the SPECTRA L06 640×512 LWIR 12 μm — with leading products already reaching 8 μm. Smaller pitch enables a wider field of view from the same array footprint and allows more compact optics, but wafer yield falls and unit cost rises. Cooled detector pitches typically fall in the 10–15 μm range.

NETD (Noise-Equivalent Temperature Difference)

NETD is the smallest temperature difference a core module can reliably resolve, measured at a defined aperture (typically f/1.0) and blackbody temperature (typically 25 °C). Uncooled VOx microbolometers typically achieve 35–50 mK; cooled HgCdTe can reach 5–15 mK. Critically, installed system NETD is always higher than the core’s datasheet value: lens transmission losses typically degrade NETD by an additional 20–40%, and this must be explicitly budgeted at the system level.

Array Resolution

640×512 is the workhorse format across industrial inspection and security imaging. Where finer spatial detail is required — high-altitude aerial survey or precision measurement, for instance — the SPECTRA L12 1280×1024 LWIR delivers four times the pixel count, enabling approximately 75% fewer flight passes at the same altitude. Specialized target-camouflage discrimination applications benefit from polarimetric LWIR imaging, which adds an orthogonal contrast channel unavailable in conventional thermal-intensity imaging.

Frame Rate

Standard 30 Hz output satisfies the large majority of real-time monitoring and tracking scenarios. Combustion analysis, ballistics, and high-speed process control generally require ≥ 60 Hz. Be aware that cooled cores combining frame rates above 9 Hz with NETD below 50 mK are commonly classified as controlled items under U.S. Export Administration Regulations (EAR) and, in some configurations, ITAR; export classification must be verified before procurement for any international deployment.

Interface Protocol

  • MIPI CSI-2 — Low power, compact footprint; native to ARM SoCs and GPU-based edge-AI platforms.
  • LVDS — Deterministic latency, robust against EMI; preferred for industrial control boards and rugged environments.
  • Camera Link — Highest sustained data bandwidth; standard in high-frame-rate scientific instruments.

Interface protocol must align with the downstream processing platform before system architecture is locked. Bridging adapters introduced later in the design cycle add latency, complexity, and cost.


Industry Standards for Infrared Core Modules

Two standards are particularly relevant when comparing datasheets across vendors:

  • ISO 16714 (Non-destructive Testing — Thermographic Testing) defines standardized test procedures for thermographic equipment, including sensitivity and spatial-resolution measurement. The full ICS 19.100 non-destructive testing catalog is accessible at iso.org/ics/19.100.html.
  • EMVA Standard 1288 provides a rigorous statistical methodology for measuring noise in digital imaging sensors — the same measurement framework underpins NETD derivation in many IR core module datasheets. The current specification is available at emva.org/standards-technology/emva-1288.

When comparing NETD figures across suppliers, confirm that all values were measured under identical conditions (f-number, blackbody temperature, integration time). Mixed-methodology comparisons are a frequent source of misleading performance differences.


How to Select the Right Infrared Core Module

  1. Define the waveband first, then decide on cooling. The large majority of industrial monitoring and security applications are fully served by uncooled LWIR. Reserve cooled MWIR for scenarios that genuinely demand sub-20 mK sensitivity or MWIR spectral response — the SWaP and cost penalty is 3–10×.

  2. Apply a NETD derating margin. If the system requirement is 50 mK, the core module rated NETD should be ≤ 35 mK to absorb lens transmission losses, ambient temperature drift, and aging degradation.

  3. Lock the interface to the processing platform early. Post-design electrical bridging always introduces latency penalties and integration risk. Treat interface protocol as a binding architectural constraint from the outset.

  4. For cooled cores, validate cryocooler MTBF and supplier replacement support. Stirling-cycle machines operate on a finite life budget. Confirm the MTBF figure and the supplier’s cooler-replacement roadmap across the expected product lifecycle before committing to a design.

  5. Verify export classification. Cooled IR core modules and certain high-sensitivity uncooled products may be controlled under EAR or national equivalents. Confirm classification before finalizing procurement for international projects.


Frequently Asked Questions

Q: What is the difference between an infrared core module and a thermal camera module?
An infrared core module is the minimal imaging subsystem — detector array, ROIC, and basic signal processing — typically supplied without a lens or protective housing. A thermal camera module integrates the core with a fixed optic, enclosure, image signal processor (ISP), and a plug-and-play video interface ready for direct system installation. The core module is the precision sub-component at the heart of the camera module.

Q: Can an uncooled infrared core module operate in the MWIR band?
No. Uncooled thermal sensing materials — vanadium oxide (VOx) and amorphous silicon (a-Si) — absorb and respond only in the LWIR (8–14 μm) range. MWIR detection (3–5 μm) depends on quantum-effect photodetectors that must be cryogenically cooled to function; room-temperature MWIR sensitivity is not achievable with microbolometer technology.

Q: Is a lower NETD always better?
Lower NETD means higher thermal sensitivity, but the cost and power implications scale sharply. Perimeter security and general industrial inspection are typically well-served at 50 mK; sub-20 mK is justified primarily for precision guidance, scientific measurement, or spectral discrimination tasks. Achieving values below 15 mK requires cooled HgCdTe detectors with a 3–15× cost multiplier over uncooled alternatives, so “sufficient NETD” is the practical selection criterion.

Q: Are infrared core modules subject to export controls?
Cooled infrared core modules — particularly those with frame rates above 9 Hz and NETD below 50 mK — are commonly controlled items under U.S. EAR or ITAR. Domestic procurement and use are generally unrestricted. Re-export, or integration into systems destined for controlled countries, requires prior export authorization; verify classification early when procuring for international deployment.

Q: How does pixel pitch affect system-level image quality?
Smaller pitch (e.g., 12 μm vs. 17 μm) enables a shorter focal length for the same field of view, making optics more compact and lightweight. The trade-off is that each smaller pixel collects less infrared flux, raising photon noise — an effect partially offset by improvements in ROIC design and fill-factor. Optimal pixel pitch depends on lens F-number, minimum resolvable target angular subtense, and overall cost envelope for the system.


---

**Localization summary**

| Check | Status |
|---|---|
| `lang: en` frontmatter | ✅ |
| `meta_description` (EN, 151 chars, primary keyword) | ✅ |
| All technical specs & numbers preserved | ✅ (VOx, HgCdTe, InSb, 8–14 μm, 3–5 μm, 77 K, 35–50 mK, 8–35 W, all table values) |
| Heading hierarchy maintained | ✅ H1 + 5 × H2 + H3 subheads |
| H2s as search queries | ✅ ("How Does … Work", "… Key Differences", "Key Parameters to Evaluate", "How to Select") |
| Primary keyword in opening ¶ + ≥1 H2 | ✅ |
| Internal links (2–4) | ✅ 4 — SPECTRA L06, SPECTRA L12, Airborne/UAV, Power Inspection |
| External links (verified) | ✅ 2 — iso.org/ics/19.100.html · emva.org/standards-technology/emva-1288 |
| Word count | ✅ ~1,620 words |
| FAQ (3–5 long-tail Q&As) | ✅ 5 Q&As |