Transcript Final Presentation

1394b Gigabit Opto-electronic
Data Communication Design
and Test
Transmitter Group(G8)
Final Presentation
Zubair P. Siddiqui
Mubin Ansari
ECE 4006C
April 23, 2002
Background Information
IEEE 1394
High Performance Serial Bus
Scalable, Flexible, Low Cost Standard
Integrates Consumer Electronics and PCs
Simple common plug-in serial connector
Hot-Plug and Plug and Play © compatible
Peer-to-peer interface
Apple’s FireWire and Sony’s i.Link use 1394.
1394-1995
IEEE
1394a
1394b
Original specification called 1394-1995
Further developments lead to 1394a
 Greater efficiency
 Arbitration acceleration
 Fast reset
 Suspend/Resume feature
1394b – Major Improvements
Increased Speed – 800Mbps and 1.6 Gbps
Architectural infrastructure supports 3.2 Gbps
Greater variety of transmission mediums
Increased transmission distances (up to 100m)
Bi-lingual’ ports for backward compatibility
• ‘Pure beta’ for 1394b on both ends
• Reduced cost – simple integration of PHY
MAX3287 Evaluation Kit
Specifications
MAX3287 Evaluation Kit
Specifications (cont’d)
Surface-mount Demonstration board
The MAX3287: A high-speed laser driver for
fiber optic LAN transmitters
Allows optical and electrical evaluation of
MAX3287




1.25Gbps
Common-Cathode configuration
3.5 to 5.5 supply voltage
30mA laser modulation current
MAX3287 Evaluation Kit
Specifications (cont’d)
Components
Bias generator with APC
Laser modulator
Safety circuit
Power-on reset (POR) circuit
Modification on MAX3287
EV Board
Purpose: AC coupled
output for VCSEL
Resistor parallel to the
SMA connector
(R20=49.9) was
removed
R24=25 was removed
and replaced by R20
Blue circles on the figure
show the modifications
Designed Transmitter Board
Designed Transmitter Board
Board Design
Objective: Minimize cost and increase simplicity
Essentially, same as the MAX3287 EV kit without
unnecessary parts
Output was AC coupled
Capacitors helped to reduce the noise level
RTC left open and potentiometer used for Rmod
Rmod controls the modulation current (set to 10k)
Design of Transmitter board shown on next slide
Designed Transmitter Board
Board Design (cont’d)
Designed Transmitter Board
Board Layout
SuperPCB used to make the design layout
Manufacturer specified dimensions used for
all surface mounts including MAX3287 chip
Red traces go on top and blue go on the
bottom of the board
Green color represents holes and connection
pads
Dotted green line marks the outline of the
board
Next slide shows the TX board layout
Designed Transmitter Board
Board Layout (cont’d)
Designed Transmitter Board
Final Transmitter Board
Poorly Fabricated
Short circuits at two different places
Exact-o-knife used to scrape away short
circuits
Lands not stable and came off while soldering
Jumper wires used to overcome this problem
Difficult to solder the surface mounts,
especially the Maxim chip
Assembled TX board on next slide
Designed Transmitter Board
Final Transmitter Board (cont’d)
Testing and Results
Eye Pattern Analysis Theory
A display in which a
pseudorandom digital
data signal is
repetitively sampled
Vertical input and the
data rate is utilized to
trigger the horizontal
sweep
Wide-open eye=minimal
signal distortion
Close eye pattern=high
noise and greater bit
error losses.
Three significant
characteristics



Width: time interval over
which the signal received
can be sampled with no
errors from inter-symbol
interference.
Height: noise margin of a
system at a particular
time interval
Sensitivity: the rate of
closure of the eye, as the
sampling time is varied
Test Setup for Maxim board
Tested to verify its
functionality
Two test conducted for
Maxim board


DC coupled output of
Maxim board
AC coupled output of
Maxim board
Equipment used



5V power supply
SMA cables and
connectors
Oscilloscope and
function generator
Test Results for Maxim board
DC coupled
Near perfect open
eye
Fits the compliance
mask
Signifies very low
SNR
Slight dip at the
bottom (unexplained
phenomenon)
Test Results for Maxim board
AC coupled
Again a near perfect
open eye
Low SNR
Fits the compliance
mask
The dip on bottom
disappears
(mysteriously)
Test Setup for TX-OE Setup
Similar to the setup
used for testing
Maxim board
Separate power
supplies for each
board
Maxim board with
AC coupled output
used
Test Results for TX-OE Setup
Eye pattern signifies
lots of jitter and
noise
Noise and Jitter
from Bias circuitry
for VCSEL and OE
module
Test Setup for Final
Transmitter Board
Similar to test setup
for the Maxim
boards
D21.5 and PRBS7
used for testing
5V power supply
used
Test Results for designed
Transmitter board
Jitter and noise
observed when testing
with PRBS7
Passes the compliance
mask test when using
D21.5 but dip observed
Possible problems:


Long traces
Bad Solder Joints
Height of eye can be
adjusting by varying the
pot at Rmod
Test Results from D21.5
Test results from PRBS7
Sources of Errors
Possible sources of error include
Cold solder joints
 Use of jumper wires
 Use of unregulated power supply
 Sharp or right-angle bends in the traces
 Short circuits on the board
 Improper Impedances

Suggestion for improvements
Better quality boards should be used
Better understanding of transmission
line concepts required
Shorter traces in the layout
Avoid using jumper wires
Suggestion for improvements
Solder power connector directly on the
board
Use one side of the board only (for
layout)
Experiment with the MAX3288 chip
Use photodiode for feedback
Conclusion
MAX3287 EV kit studied carefully and
tested successfully
Unavailability of 1394b Gigabit Ethernet
card made our work no different than
other transmitter groups
Never reached the point where the
advantages of 1394b over its
predecessors could be realized.
Conclusion
New learning opportunities explored
SuperPCB
 Soldering and de-soldering process
 Practical training

Over all, the project was a good
learning experience and the objective of
the class was achieved.