Nonlinearity correction of the integrated time-to-digital converter with direct coding

Author

Pelka, R. ; Kalisz, J. ; Szplet, R.

Author_Institution

Mil. Univ. of Technol., Warsaw, Poland

fYear

1996

fDate

17-21 June 1996

Firstpage

548

Lastpage

549

Abstract

A method and computer program for automated identification and correction of the nonlinearity error in time-to-digital converter (TDC) with 200-ps resolution, implemented on a single FPGA device is presented. The nonlinearity error is identified and corrected by the neural network trained with the fast Levenberg-Marquardt learning rule. Resulting nonlinearity was improved from 1.25 LSB (least significant bit) to about 0.12 LSB (24 ps).

Keywords

Newton method; analogue-digital conversion; automatic test software; computerised instrumentation; delay circuits; error correction codes; field programmable gate arrays; learning (artificial intelligence); neural nets; quantisation (signal); signal resolution; time measurement; automated identification; computer program; delay buffers; direct coding; error correction; fast Levenberg-Marquardt learning rule; fast Newton method; integrated time-to-digital converter; latch flip-flops; neural network; nonlinearity correction; parallel tapped delay lines; quantization; single FPGA device; CMOS logic circuits; Computer errors; Delay effects; Delay lines; Error correction; Field programmable gate arrays; Military computing; Propagation delay; Pulse measurements; Time measurement;

fLanguage

English

Publisher

ieee

Conference_Titel

Precision Electromagnetic Measurements Digest, 1996 Conference on

Conference_Location

Braunschweig, Germany

Print_ISBN

0-7803-3376-4

Type

conf

DOI

10.1109/CPEM.1996.547342

Filename

547342