The sensitivity of a detector pixel at a certain wavelength is defined as the detector electrical output per unit of radiation energy (photons) incident to that pixel. With a given A/D converter this can be expressed as the number of counts per mJ of incident radiation.
The relation between light energy entering the optical bench and the amount hitting a single detector pixel depends on the optical bench configuration. The efficiency curve of the grating used, the size of the input fiber or slit, the mirror performance and the use of a Detector Collection Lens are the main parameters. With a given set-up it is possible to do measurements over about 6-7 decades of irradiance levels. Some standard detector specifications can be found in Table 4 detector specifications. Optionally a DCL (Cylindrical Detector Collection) lens can be mounted directly on the detector array. The quartz lens (DCL-UV/VIS for AvaSpec-ULS2048/3648) will increase the system sensitivity by a factor of 3-5, depending on the fiber diameter used. The DCL-UV/VIS-200 can be used for the AvaSpec-ULS2048L/3648/2048XL to have a better vertical distribution of light focusing on the detector and is primarily for fiber diameters larger than 200 µm and round- to-linear assemblies.
The SensLine has the most sensitive detectors in Avantes’ instrument line, three backthinned detectors and two cooled CCD detectors.

In the tables below the UV/VIS detectors are depicted with their specifications, please find below some additional information on how those specifications are determined.

Pixel Well Depth (electrons)
This value is specified by the detector supplier and defines how many electrons can fit in a pixel well before it is saturated, this value determines the best reachable Signal to Noise (=√(Pixel well depth)).

Sensitivity in Photons/count @ 600 nm
The number of Photons of 600 nm that are needed to generate one count of signal on a 16-bit AD converter, the lower this number is, the better is the sensitivity of the detector. The calculation of the number of Photons/count is (Pixel Well depth in electrons)/16-bit AD/Quantum Efficiency @ 600 nm.

Sensitivity in counts/µW per ms integration time
Sensitivity here is for the detector types currently used in the UV/VIS AvaSpec spectrometers as output in counts per ms integration time for a 16-bit AD converter. To compare the different detector arrays we have them all built up with an optical bench with UA 300 lines/mm grating covering 200-1100 nm (AvaSpec-128 with grating VZ 350-1100 nm), DCL if applicable, and 50 µm slit. The measurement setup for 350-1100 nm has a 600 µm fiber connected to an AvaSpere-50-LS-HAL, equivalent to an optical power of 1.14 µW.For the UV/VIS measurement at 220-1100 nm we connected the 600 µm fiber to an AvaLight-DHS through a CC-VIS/NIR diffuser, equivalent to 2.7 µW power.

Peak wavelength and QE @ peak
The peak wavelength is provided by the detector supplier as well as the Quantum Efficiency, defined as the number of electrons generated by one photon.

Signal/Noise is measured for every detector at Avantes’ Quality Control Inspection and defined as the illuminated maximum Signal/Noise in Root Mean Square for the shortest integration time. The RMS is calculated over 100 scans.

Dark Noise
Dark noise is measured for every detector at Avantes’ Quality Control Inspection and defined as the non-illuminated noise in Root Mean Square for the shortest integration time. The RMS is calculated over 100 scans.

Dynamic Range
The dynamic range is defined as the (maximum signal level- baseline dark level)/dark noise RMS.

Photo Response Non-Uniformity
Photo Response Non-Uniformity is defined as the max difference between output of pixels when uniformly illuminated, divided by average signal of those pixels.
PRNU is measured for every detector at Avantes’ Quality Control Inspection.

The frequency is the clock frequency at which the data pixels are clocked out through the AD-converter.


Detector Specifications (based on a 16-bit AD converter)








CMOS linear array

CCD linear array

# Pixels, pitch

2048, 14µm

3648, 8 µm

Pixel width x height (µm)

14 x200

14 x 56

14 x 200

8 x 200

Pixel well depth (electrons)





Sensitivity Photons/count @600nm




in counts/µW per ms integration time

375,000 (AvaSpec-ULS2048CL)

310,000 (AvaSpec-ULS2048)

470,000 (AvaSpec-ULS2048L)

160,000 (AvaSpec-ULS3648)

Peak wavelength

700 nm

550 nm

450 nm

550 nm

QE (%) at peak 80%




200 :1

300 :1

350 :1

Dark noise (counts RMS)





Dynamic Range






± 5%

± 5%

Wavelength range (nm)



6 MHz

2 MHz

2 MHz

1 MHz






HAM1024x58 HAM1024x122

Back-thinned CCD array

 Cooled Back-Thinned CCD array 
 # Pixels, pitch  2048x14, 14µm  2048x64, 14µm  2048, 14µm  1024 x 58, 24 µm 1024 x 122, 24 µm
 Pixel width x height (µm)

 14 x 14

 14 x 500

 24 x 24 (total height 1.4 mm)  24 x 24 (total height 2.9 mm)
 Pixel well depth (electrons)


 Sensitivity Photons/count @600nm



in counts/µW per ms integration time

 200,000 (AvaSpec-ULS2048x16)

 600,000 (AvaSpec-ULS2048x64)

 460,000 (AvaSpec-ULS2048XL)





 Peak wavelength  600 nm   650 nm
 QE (%) at peak





450 :1

 1000 :1 
 Dark noise (counts RMS)


 Dynamic Range




 Wavelength range (nm)


 Frequency 1.33 MHz   1 MHz  250kHz  


* DUV coated

** Photo Response Non-Uniformity = max difference between output of pixels when uniformly illuminated, divided by average signal

Figure 5 Detector Spectral Sensitivity Curves

Sensitivity UV-VIS-NIR-2


In the next table the specification is given for the NIR spectrometers, followed by the spectral response curve for the different detector types are depicted.

For NIR detectors 2 different modes are available, the default setting is for high-sensitivity mode (HS), this means more signal at a shorter integration time. The other mode of operation is low-noise (LN), this means a better S/N performance.
Sensitivity, S/N, dark noise and Dynamic Range are given as HS and LN values.

NIR Detector Specifications

Detector NIR256-1.7 NIR256-1.7TEC NIR512-1.7TEC NIR256-2.5TEC-HSC-EVO NIR512-2.5TEC-HSC-EVO
Type Linear InGaAs array Linear InGaAs array with 2 stage TE cooling
# Pixels, pitch 256, 50 µm 512, 25 µm 256, 50 µm 512, 25µm
Pixel width x height (µm) 50 x 500 25 x 500 50 x 250 25 x 250
Sensitivity HS
in counts/µW per ms
1,300,000 (integral 1000-1750 nm) 2,770,000 (integral 1000-1750 nm) 2,770,000 (integral 1000-1750 nm)

990,000 (best estimate)

Noise (HS)
2000:1 1700:1 1800 :1 1900 :1
Dark noise HS (counts RMS) 14 13 16 15
Dynamic Range HS






Sensitivity LN
in counts/µW per ms
74,000 (integral 1000-1750 nm) 96,000 (integral 1000-1750 nm) 55,000 (integral 1000-2500 nm)
Signal/Noise (LN) 6000:1 3600:1 4000:1 3700:1
Dark noise LN (counts RMS) 8 12 13
Dynamic Range LN 8000 4500 5100
Peak wavelength 1550 nm 1500 nm 2300 nm
QE (%) @ peak 90% 70% 65%
PNRU** ± 5% 10% ±5%
Defective pixels (max) 0 12 26
Wavelength range (nm) 900-1750 1000-2500
Frequency 500 kHz 2.4 MHz 500 kHz

** Photo-Response Non-Uniformity


NIR Detector Sensitivity Curves

sensitivityCatIX NIR