ECLAIRs, a 2-D coded-mask imaging camera on-board the Sino-French SVOM space mission, will detect and locate Gamma-ray bursts (GRBs) in near real time in the 4-150 keV energy band. The design of ECLAIRs has been mainly driven by the objective of achieving a low-energy threshold of 4 keV, unprecedented for this type of instrument. The detection plane is an assembly of 6400 Schottky CdTe semiconductor detectors of size 4x4x1 mm3 organized on elementary hybrid matrices of 4x8 detectors. The detectors will be polarized from -300V to -500V and operated at -20°C to reduce both the leakage current and the polarization effect induced by the Schottky contact. The remarkable low-energy threshold homogeneity required for the detection plane has been achieved thanks to: i) an extensive characterization and selection of the detectors, ii) the development of a specific low-noise 32-channel ASIC, iii) the realization of an innovative hybrid module composed of a thick film ceramic (holding 32 CdTe detectors with their high voltage grid), associated to an HTCC ceramic (housing the ASIC chip within an hermetic enclosure). In this paper, we start describing a complete hybrid matrix, and then the manufacturing of a first set of 50 matrices (representing 1600 detectors, i.e. a quarter of ECLAIRs detector’s array). We show how this manufacturing allowed to validate the different technologies used for this hybridization, as well as the industrialization processes. During this phase, we systematically measured the leakage current on Detector Ceramics after an outgassing, and the Equivalent Noise Charge (ENC) for each of the 32 channels on ASIC Ceramics, in order to optimize the coupling of the two ceramics. Finally, we performed on each hybrid module, spectral measurements at -20°C in our vacuum chamber, using several calibrated radioactive sources (241Am and 55Fe), to check the performance homogeneity of the 50 modules. The results demonstrated that the 32-detector hybrid matrices presented homogeneous spectral properties and that a lowenergy threshold of 4 keV for each detector could be reached. In conclusion, our hybrid module has obtained the performance required at the SVOM mission level and successfully withstood the space environment tests (TRL 6/7). This development phase has given us the opportunity to build a detector’s array prototype (Engineering Model) equipped with 50 hybrid modules. Thanks to this prototype we are in the process of validating a complete detection chain (from the detectors to the backend electronics) and checking the performance. In addition it enables us to consolidate the instrument’s mechanical and thermal design, and to write preliminary versions of the quality procedures required for integration, functional tests and calibration steps. At the end of this prototype development and testing, we will be ready to start the detailed design of the detection plane Flight Model.