You are on page 1of 6

Impulse Noise

Measurement
Test Setup

Ramin Shirani
1/27/2015 Larry Cohen
Impulse Noise Problem Overview
•Problem: Impulse noise events in the enterprise environment may degrade the
operational BER of otherwise compliant NGBase-T links to unacceptable levels
–Impulse noise effects in the enterprise environment on systems with bandwidth
beyond 1000Base-T are not fully understood

•What we already know


–10GBase-T is mainly deployed in the data center environment so the existing
knowledge base is not fully applicable to the enterprise environment
–1000Base-T equipment is known to operate properly in the enterprise environment

•First step: Create a test setup to capture, record, and analyze the characteristics of
impulse noise waveforms in the enterprise environment and measure the important
characteristics that could affect NGBase-T link quality
–Frequency of occurrence (inter-arrival time)
–Signal bandwidth and the effect of AFE lowpass filtering
–Common-mode and differential-mode waveforms (peak-to-peak voltage) induced
in cabling; main interest is worst-case differential waveforms
–Determine relative susceptibility of different cable and connector categories

1/27/2015 2
Impulse Noise Measurement Considerations
•Instrumentation must be placed at multiple locations to obtain a valid (and
useful) noise profile of the enterprise environment

•Example location #1: Wired port in an office “bullpen” cubicle


–Lots of human activity in an area with no ESD mitigation
–Potential for lots of ESD induced impulse noise events near LAN ports
–Additional potential for on/off switch arc transient induced impulse noise events
•Example location #2: Wireless access point placement
–Less nearby human activity; potentially less ESD events
–Potential for on/off switch arc transient induced impulse noise events
•Relevant for all locations
–How bad is the impulse noise produced from office HVAC equipment?
–Are there additional unexpected noise sources picked up due to cable routing
(floor and/or ceiling) and other office equipment?

1/27/2015 3
Impulse Noise Event Capture Test Setup #1
Control Computer
(PC with GPIB)

GPIB

This is the
100  Differential-mode receive "antenna" Digital Storage
and 50  Common- Oscilloscope
mode for all four pairs All unused pairs
terminated with 50  0-degree splitter 50 
L (1 to 10 meters)
for common-mode
Optional Keystone Shielded
or inline connector +20 dB Gain
CM-DM RJ45-to-SMA
Breakout/ 50 
Termination RJ45 RJ45 Power 50  50 
Block Termination LNA
Splitter
Board 50 
Cable channel under test Low-pass filter
(Cat 5e/6/6a UTP) (f3dB = 200 MHz to 400 MHz)
180-degree splitter
for differential-mode

• Scope channel should have at least 500 MHz bandwidth, 2 Gsps sampling rate, and
10000 samples memory capture depth (5 usec time span)
• LNA is mainly required to capture low-level differential-mode waveforms
• Lowpass filter simulates effect of AFE and adds additional anti-alias filtering
• Impulse capture threshold controlled by scope trigger level
• Captured impulse waveforms recorded on control computer for later analysis
• Proper site placement of noise survey instrumentation is critical to gathering valid data

4
1/27/2015
Impulse Noise Event Capture Test Setup #2
This is the
100  Differential-mode receive "antenna" Control Computer
and 50  Common- (PC with GPIB)
mode for all four pairs All unused pairs
terminated with 50  GPIB
L (1 to 10 meters)

Optional Keystone Shielded


or inline connector RJ45-to-SMA CH1
CM-DM
Breakout/ 50 
Termination RJ45 RJ45 50 Digital Storage
Block Termination
CH2 Oscilloscope
Board 50 
Cable channel under test 50 
(Cat 5e/6/6a UTP)
Matched coaxial
cable pair

• Captures both common-mode and differential mode impulses simultaneously, but requires
better scope with more bandwidth, memory depth, and sensitivity
• Scope channels should have at least 1 GHz MHz bandwidth, 5 Gsps sampling rate, and
25000 samples memory capture depth (5 usec time span)
• Use post-processing of captured waveforms to generate common-mode and differential
mode impulse noise events and simulate any AFE lowpass filtering
• Scope vertical sensitivity may limit resolution of differential-mode waveforms
• Impulse capture threshold controlled by scope trigger level

1/27/2015 Impulse noise in the enterprise 5


environment
Design of RJ45 CM-DM Termination Block
RJ45 Termination Block
(RDM = 100RCM = 50

Proposed symmetrical PCB


layout of resistors to preserve
Edge of connector Pair 1 R1 R1 termination balance.
enclosure should be
at the edge of the R2
board. R1 R1

1 Via to ground
2
R1 R2 R1
3
Shielded R2
6 Pair 2 Pair 4
RJ45 4 R2
Connector Resistor values:
5 R1 R2 R1 R1 = 49.9+/- 0.1%
7 R2 = 24.9 +/- 1%
8

R1 R1
All signal traces are 100 Ohms
(+/- 10%) weakly coupled
Use "tab up" Pair 3 matched differential pairs, and
RJ45 jack all are the same length.

1/27/2015 6