Goldeye SWIR InGaAs sensor technology cameras

Select the spectral band by matching the target material’s absorption characteristics with the illumination and optical transmission. The G/CL-030 and G/CL-130 VSWIR models cover 400 to 1,700 nm, making them suitable when visible and SWIR information must be captured with one camera. Standard SWIR models cover 900 to 1,700 nm, while XSWIR 1.9 models extend from 1,100 to 1,900 nm and XSWIR 2.2 models cover 1,200 to 2,200 nm. The lens, window, filter and illumination must all transmit the selected wavelengths. A longer spectral range is not automatically preferable because XSWIR models require TEC2 cooling and a larger thermal design.

Choose Goldeye G when Gigabit Ethernet integration, network cabling or Power over Ethernet is useful, and choose Goldeye CL when the host system already uses a Camera Link frame grabber. Goldeye G uses IEEE 802.3 1000BASE-T and can operate from PoE or a 10.8 to 30.0 V DC input. Goldeye CL uses the Camera Link Base configuration and requires a separate DC supply and compatible frame grabber. Camera Link does not support hot plugging, so its data cable must be connected before power is applied and power should be removed before disconnection. The image-sensor options are paired across both interface families, allowing the host architecture to guide the interface decision.

The appropriate model depends on whether spatial detail or acquisition rate is more important to the inspection. The 320 × 256 G/CL-008 models reach 344 fps, while the 656 × 520 G/CL-030 reaches 234 fps and the 1280 × 1024 G/CL-130 reaches 94 fps. At approximately 0.3 megapixels, the G/CL-032 reaches 100 fps, the G/CL-033 reaches 301 fps and the G/CL-034 reaches 303 fps. Pixel sizes also vary from 5 to 30 µm, affecting optical sampling, sensor area and photon collection. Regions of interest can raise acquisition speed, although final performance also depends on pixel format, interface bandwidth and host processing.

TEC2 is generally appropriate for extended-SWIR sensing, long exposure times or installations where sensor temperature must remain well below ambient. The XSWIR and G/CL-034 TEC2 models use calibrated sensor setpoints down to −30 °C, with a minimum temperature difference of up to 60 K for the relevant configurations. TEC1 models use single-stage stabilisation, commonly with calibrated operating points around +20 or +25 °C, while the G/CL-008 Cool version is calibrated at +5 °C. The G/CL-033 TECless variant removes thermoelectric cooling and retains a 301 fps full-resolution rate, but its dark current is more dependent on sensor temperature. Fan vibration, heat removal, power consumption and exposure duration should all be assessed before selecting the cooling method.

Start by confirming that the lens covers the selected sensor format and transmits the complete wavelength band used by the application. C-Mount is the standard configuration, while F-Mount and M42-Mount options are available for selected models; some VSWIR and XSWIR combinations require confirmation before ordering. The range includes Type 1/4, Type 1/2, Type 1 and Type 4/3 sensors, so the required image circle varies considerably. Optional IRBP 1450, BP 1450 and LP 920 filters can isolate water-absorption bands or remove visible wavelengths, subject to model compatibility. Removed-cover-glass variants are available for selected models but require stricter contamination control, optical cleaning procedures and careful handling.

The camera must be installed so that its housing remains within −20 to +55 °C under the intended load and ambient conditions. Compact industrial models use a 78 × 55 × 55 mm body, while scientific, Cool and XSWIR configurations use a 90 × 80 × 80 mm body. A thermally conductive mounting base or heat sink should be used, particularly for high-duty-cycle acquisition, PoE operation or enclosed installations. With a suitable lens fitted, standard models are rated IP40, while Cool models are IP20 and may need additional enclosure protection. Heavy or long lenses should have independent support, even though the cameras have been tested against IEC 60068 shock and vibration procedures.

The onboard processing functions can reduce the amount of corrective work required in the host application. Non-uniformity correction compensates for pixel-to-pixel response differences, while background correction and defect-pixel correction address offset behaviour and unusable pixels. Binning and multiple regions of interest can reduce the amount of transmitted data or concentrate processing on relevant image areas. Look-up tables, auto contrast and auto exposure can support visualisation and automatic adjustment, while the 256 MB buffer provides temporary storage during short interruptions. Correction settings should still be validated at the intended exposure time, sensor temperature, gain and illumination level before the system is released for production.

The cameras provide hardware and software controls for integration with lighting, motion equipment, sensors and process controllers. Available connections include one LVTTL input, one LVTTL output, one opto-isolated input, two opto-isolated outputs and an adjustable RS232 channel operating at approximately 115 kbaud. Trigger control, acquisition-frame-rate control, event reporting and image chunk data can be used to associate each image with the process state. Goldeye G models can use PoE or external DC power, while Goldeye CL models require external DC power and the prescribed Camera Link connection sequence. Shielded cables, strain relief and measured trigger latency should be included in the final installation plan.