TY - GEN
T1 - Towards parallel, 192 channel, 40MS/s/ch data acquisition for optical tomography
T2 - 16th IEEE SENSORS Conference, ICSENS 2017
AU - Fisher, E.
AU - Tsekenis, S-A.
AU - Yang, Y.
AU - Ouypornkochagorn, T.
AU - Chighine, A.
AU - Polydorides, N.
AU - Wright, P.
AU - McCann, H.
N1 - Acceptance date set to exclude from REF OA Policy
PY - 2017/12/21
Y1 - 2017/12/21
N2 - To investigate novel engine and fuel designs for greener aviation, instrumentation is required that can spatially and temporally resolve gas concentrations within aero-engine exhausts. This paper presents work towards a parallel, high-speed, distributed data acquisition (DAQ) system that employs in-situ demodulation of tunable diode laser absorption spectroscopy (TDLAS) signals. We briefly describe how this sits within a wider tomographic instrument, the electrical system of this scalable design and preliminary characterization. Being remote from the end-user (approx. 60m) and deployed within an industrial environment, we have used a hierarchical, embedded strategy. This uses photodiode pre-amplification, filtering, digitization, signal demodulation, Ethernet packaging and microprocessor control implemented both on a multi-node, distributed basis and with the DAQ physically mounted on the same mechanical 'ring' as the tomographic imaging array. Results show agreement with design but indicate that the first-generation interrupt-based direct-memory-access (DMA) between FPGA fabric memory and microprocessor memories is the predominant bottleneck.
AB - To investigate novel engine and fuel designs for greener aviation, instrumentation is required that can spatially and temporally resolve gas concentrations within aero-engine exhausts. This paper presents work towards a parallel, high-speed, distributed data acquisition (DAQ) system that employs in-situ demodulation of tunable diode laser absorption spectroscopy (TDLAS) signals. We briefly describe how this sits within a wider tomographic instrument, the electrical system of this scalable design and preliminary characterization. Being remote from the end-user (approx. 60m) and deployed within an industrial environment, we have used a hierarchical, embedded strategy. This uses photodiode pre-amplification, filtering, digitization, signal demodulation, Ethernet packaging and microprocessor control implemented both on a multi-node, distributed basis and with the DAQ physically mounted on the same mechanical 'ring' as the tomographic imaging array. Results show agreement with design but indicate that the first-generation interrupt-based direct-memory-access (DMA) between FPGA fabric memory and microprocessor memories is the predominant bottleneck.
KW - Chemical Species Tomography
KW - Data Acquisition
KW - TDLAS
KW - Tunable Diode Laser Absorption Spectroscopy
UR - http://www.scopus.com/inward/record.url?scp=85044288782&partnerID=8YFLogxK
U2 - 10.1109/ICSENS.2017.8234310
DO - 10.1109/ICSENS.2017.8234310
M3 - Conference contribution
AN - SCOPUS:85044288782
VL - 2017-December
T3 - IEEE Sensors
SP - 1302
EP - 1304
BT - IEEE SENSORS 2017 - Conference Proceedings
PB - Institute of Electrical and Electronics Engineers
Y2 - 30 October 2017 through 1 November 2017
ER -