Design and implementation on FPGA of an embedded data-logger with a scalable architecture for a large number of analog inputs

Corrosion related issues cost billions every year in the oil and gas industry. One consequence of corrosion is the wall thickness loss in pipelines, this issue could cause leaks and stop the production which leads to an economic slowdown. In-line inspection tools measure pipe wall thickness using Magnetic Flux Leakage (MFL) technique. In order to identify critical points over the pipeline, an embedded system capable of register hundreds of hall effect sensor signals is needed. It also has to comply specific mechanical and electrical requirements to be suited for inline inspection. This work presents the architectural design and FPGA (Field-Programmable Gate Array) implementation of an embedded data logger used to acquire 160 analog signals for an MFL inspection tool currently under development. The architecture was implemented in a XilinxArtix FPGA and was designed to be scalable to allow a large number of analog inputs. The final implementation is capable of acquiring 10k samples per second. It was successfully tested to continuously acquire and storage in flash memory the data from 132 hall effect sensors at 936kB/s.