Evaluation of a simple DC-balanced encoding method for LVDS data transmission over CAT-5 cable
Abstract (Summary)
In this study, a simple dc-balanced encoding scheme was developed to reduce
the bit error rate for high-speed data transmission over CAT-5 cable using Low-voltage
Differential Signaling (LVDS). The dc balance encoder was implemented to make binary
sequences with a spectral null at zero. A data transmission controller using the dc
balanced scheme was implemented in an Altera Field Programmable Gate Array
(FPGA), including data generators that send data to a DS92LV1023 10-bit bus LVDS
serializer, data checkers for checking the data coming from a DS92LV1224 10-bit bus
LVDS deserializer, and dc balance and non-dc balance encoders. Data was transmitted
over various lengths of CAT-5 cable with and without dc balance to determine the effect
of the dc balance scheme on transmission errors.
To generalize the type of data used in the transmission tests, three different
kinds were selected for the error testing: pseudo-random numbers generated by a 32-bit
Linear Feedback Shift Register (LFSR) binary polynomial generator; consecutive
numbers generated by a counter; and data looked-up from a Read Only Memory (ROM)
implemented in an FPGA embedded memory block.
The data transmission controller was constructed, configured, and tested both
with and without dc balance. It provides a data transmission rate of 538.8 Mbps, and is
able to send the number of errors encountered during the transmission process to a PC
via the PCI bus. Testing results verify that the dc balance scheme adopted in this thesis
significantly improves the accuracy of the serial data transmission. Both dc-balanced
and non-dc-balanced encoding proved error-free out to cable lengths of about 19.8
meters. DC-balanced encoding also extended the error-free cable length by about 1.5
meters and reduced errors by about 60% for longer cables.
iv
Bibliographical Information:
Advisor:
School:The University of Tennessee at Chattanooga
School Location:USA - Tennessee
Source Type:Master's Thesis
Keywords:
ISBN:
Date of Publication: