Jan 21, 2014 ... Used ISO 7637-3 specification for transient noise tests on signal lines for road
vehicles. 2. Used 2m cables with TESEQ capacitive coupling ...
RTPGE TRANSIENT NOISE MEASUREMENTS AND MODELING BASED ON ISO 7637-3 January 21, 2014
Ahmad Chini & Mehmet Tazebay Broadcom Corporation
Version 1.0
IEEE 802.3bp RTPGE Task Force– January 21, 2014
Page 1
AGENDA • Objectives & Methodology • Set-up & Calibration Information
• Measurement and Analysis Results • A Transient Noise Model for system analysis • Summary
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 22 Page
OBJECTIVES 1. Estimate transient (pulse) noise on the signal lines at the RTPGE receiver for single pair UTP cables. 2. Investigate the effect of test head grounding (floating) on the transient noise magnitude and duration. 3. Provide simple noise models for initial analysis of modulation and coding solutions for RTPGE.
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 33 Page
METHODOLOGY 1. Used ISO 7637-3 specification for transient noise tests on signal lines for road vehicles. 2. Used 2m cables with TESEQ capacitive coupling clamp (1m long) as specified in ISO 7637-3. 3. Used Agilent VNA to obtain coupling transfer function and mathematical calculation to estimate noise from simulated test pulses. 4. Used BMW suggested test levels to obtain differential noise levels at the receiver.
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 44 Page
Transient test setup per ISO 7637-3
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 55 Page
Negative Test Pulses per ISO 7637-3
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 66 Page
Positive Test Pulses per ISO 7637-3
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 77 Page
BMW Suggested Test Levels
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 88 Page
Transient Noise Test Setup •
Transient coupler (by TESEQ) designed per ISO 7637-3 was used to couple transient noise.
•
Agilent VNA measured the coupling transfer function.
•
One side of the cable connected to port 1 and 2 of VNA and other side terminated with 50Ω loads (100Ω differential).
•
Test heads had optional grounding stands in order to study the effect of grounding or floating.
VNA Transient Coupler
1-Pair, 2m UTP
10-Cm Foam
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 99 Page
Transient Coupler Calibration 0 -5 -10
S21, dB
-15 -20 -25 -30 -35 -40 -45 -50 0
•
Version 1.0
100
200
300
400 500 600 Frequency, MHz
700
800
900
1000
Almost no insertion loss in the lower frequency band where test pulses have their most energy. Port 1 is noise injection port and port 2 is the measurement port for calibration. 802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 10 10 Page
Simulated Transient Test Pulse a 200
tr = 5nS td = 100nS Vs = -150V
150 100
Volts
50 0 -50 -100 -150 -200 0
Version 1.0
50
100
150
200 nS
250
300
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
350
400
Page 11 11 Page
Simulated Transient Test Pulse and Calibration 200
tr = 5nS td = 100nS Vs = -150V
150 100
Teseq coupler input Teseq coupler output
Volts
50 0 -50 -100 -150 -200 0
•
Version 1.0
50
100
150
200 nS
250
300
350
400
No need for calibration, Input/output pulses are about the same magnitude. Input port is where noise is injected and output port is where noise level is measured for calibration. 802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 12 12 Page
Receiver HPF Effect 200 No HPF HPF 20MHz HPF 50MHz
150 100
Volts
50 0 -50 -100 -150 -200 0
50
100
150
200 nS
250
300
350
400
• Receiver HPF reduces the intensity and duration of the transient test pulses and needs to be considered in analysis of transient noises. Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 13 13 Page
Example Measured Transient Transfer Function -30 -35 -40
Sds21, dB
-45 -50 -55 -60 -65 -70 -75 -80 1 10
2
10 Frequency, MHz
10
3
• Port 1 is noise injection port and port 2 is the differential port at DUT Test Head. Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 14 14 Page
Simulated DM Transient Noise Example 100
After RX HPF of 20MHz
80 60 40
mVolts
20 0 -20 -40 -60 -80 -100 0
50
100
150
200 nS
250
300
350
400
The received noise 𝑛 𝑡 is Fourier inverse of 𝑁 𝑓 where 𝑁 𝑓 = 2 𝑃 𝑓 𝑆𝑑𝑠21(𝑓) 𝐻(𝑓) and 𝑃(𝑓) is frequency response of the injected pulse noise and 𝑆𝑑𝑠21(𝑓) is the transient noise transfer function and 𝐻(𝑓) is receiver highpass filter Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 15 15 Page
Test Heads Grounding/Isolating Effect 100 Test Heads on the GND plane Test Heads isolated from the GND plane
80 60
After RX HPF of 20MHz
40
mVolts
20 0 -20 -40 -60 -80 -100 0
50
100
150
200 nS
250
300
350
400
• Grounding/ Isolating the test heads does not affect the noise magnitude or duration significantly.
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 16 16 Page
Simulated DM Transient Noises, Various Cables 100
Before HPF
80 60 40
mVolts
20 0 -20 -40 -60 -80 -100 0
50
100
150
200 nS
250
300
350
400
• Cables are from 3 different manufacturers, results shown are collected at both ends of coupler. Test heads are grounded. • At the receiver and before HPF, the noise intensity is below 100mV and noise duration is less than 50nS. Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 17 17 Page
Simulated DM Transient Noises, Various Cables 100
After RX HPF of 20MHz
80 60 40
mVolts
20 0 -20 -40 -60 -80 -100 0
50
100
150
200 nS
250
300
350
400
• Cables are from 3 different manufacturers, results shown are collected at both ends of coupler. Test heads are grounded. • Note that intensity of transient noise pulses are reduced using the HPF at the receiver. Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 18 18 Page
Transient Noise Model at Receiver 100
Before RX HPF
mVolts
50
0
-50
-100 0
50
100
150
200 nS
250
300
350
400
• One cycle of a 20MHz sinusoidal signal with magnitude of 100mV is suggested for system analysis of transient noises.
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 19 19 Page
Summary •
Suggested a method for estimation of DM transient noises based on ISO 7637-3 – Used a transient noise coupler and VNA to measure transfer function. – Used simulated pulse noises and mathematical calculation to estimate noise at the receiver. – Provided results for multiple cables from various manufacturers.
•
Grounding Effects for the Test Heads – Comparing noise results for grounded and floating test heads, show no significant change in magnitude or duration of noise.
•
Transient Noise Model – One cycle of a 20MHz sinusoidal signal with magnitude of 100mV is suggested for system analysis of transient noises. Effect of receiver HPF on the noise pulses should be considered in analysis.
Version 1.0
802.3bp RTPGE Taskreport Force––January 21, 2014 IEEEIEEE P802.3 Maintenance July 2008 Plenary
Page 20 20 Page
