You are on page 1of 137

The “Ground” Myth

Bruce Archambeault, Ph.D.


IBM Distinguished Engineer, IEEE Fellow
barch@us.ibm.com

18 November 2008

IEEE
Introduction
• Electromagnetics can be scary
– Universities LOVE messy math
• EM is not hard, unless you want to do the
messy math
• Goal:
– Intuitive understanding
– Understand the basic fundamentals
– Understand how to read the math
November 2008 Bruce Archambeault, PhD 2
November 2008 Bruce Archambeault, PhD 3
Overview
• What does the derivative mean?
• What does integration mean?
• Weird vector notation
• In the beginning – Faraday and Maxwell
• Inductance
• “Ground”
• Primary cause of EMI problems on
PCBs
November 2008 Bruce Archambeault, PhD 4
Derivative
• How fast is something
changing?

d
[something ] Changing with
respect to time
dt
d
[something ] Changing with respect
to position (x)
dx
November 2008 Bruce Archambeault, PhD 5
Partial Derivative
• How fast is something changing for
one variable?
Changing with respect

[something (t , x)] to time (as ‘x’ is
∂t constant)


[something (t , x)] Changing with respect
∂x to position (x) (as time
is constant)
November 2008 Bruce Archambeault, PhD 6
Integration
• Simply the sum of parts (when the parts
are very small)
– Line Integral --- sum of small line segments
– Surface Integral -- sum of small surface
patches
– Volume Integral -- sum of small volume blocks

November 2008 Bruce Archambeault, PhD 7


Line Integral
(find the length of the path)

‘piece’ of E field

dl

stop →

V = − ∫ ( E • dl )
start
November 2008 Bruce Archambeault, PhD 8
Line Integral -- Closed
Circumference = ∫ path around box
x =l y=w x =0 y =0

= ∫ dx + ∫ dy + ∫ dx + ∫ dy
x =0 y =0 x =l y=w

x
November 2008 Bruce Archambeault, PhD 9
Line Integral -- Closed
• Closed line integrals
find the path length
• And/or the amount of
some quantity along
that closed path length

November 2008 Bruce Archambeault, PhD 10


Surface Integral
(find the area of the surface)

Area = ∫ da
da = dx ∗ dy
Area = ∫∫ dx ∗ dy
As dx and dy become
smaller and smaller, the
area is better calculated
November 2008 Bruce Archambeault, PhD 11
Volume Integral
(find the volume of an object)

Volume = ∫ dv
dv = dx ∗ dy ∗ dz
Volume = ∫∫∫[dx ∗ dy ∗ dz ]

November 2008 Bruce Archambeault, PhD 12


Electromagnetics
In the Beginning
• Electric and Magnetic effects not
connected
• Electric and magnetic effects were due to
‘action from a distance’
• Faraday was the 1st to propose a
relationship between electric lines of force
and time-changing magnetic fields
– Faraday was very good at experiments and
‘figuring out’ how things work
November 2008 Bruce Archambeault, PhD 13
Maxwell
• Maxwell was
impressed with
Faraday’s ideas
• Discovered the
mathematical link
between the “electro”
and the “magnetic”
• Scotland’s greatest
contribution to the
world (next to Scotch)
November 2008 Bruce Archambeault, PhD 14
“Maxwell’s Equations”
• Maxwell’s original work included 20
equations!
• Heaviside reduced them to the existing
four equations
– Heaviside refused to call the equations his
own
• Hertz is credited with proving they are
correct

November 2008 Bruce Archambeault, PhD 15


Maxwell’s Equations
are NOT Hard!

∂D
∇×H = J +
∂t
∂B
∇×E = −
∂t
November 2008 Bruce Archambeault, PhD 16
Maxwell’s Equations are not Hard!

• Change in H-field across space ¡ Change


in E-field (at that point) with time
• Change in E-field across space ¡ Change
in H-field (at that point) with time
• (Roughly speaking, and ignoring
constants)

November 2008 Bruce Archambeault, PhD 17


Current Flow
• Most important concept of EMC
• Current flow through metal changes as
frequency increases
• DC current
– Uses entire conductor
– Only resistance inhibits current
• High Frequency
– Only small part of conductor (near surface) is used
– Resistance is small part of current inhibitor
– Inductance is major part of current inhibitor
November 2008 Bruce Archambeault, PhD 18
Skin Depth
• High frequency current flows only near the
metal surface at high frequencies
Frequency Skin Depth Skin Depth
60 Hz 260 mils 8.5 mm
1 KHz 82 mils 2.09 mm
10 KHz 26 mils 0.66 mm
1
δ=
100 KHz 8.2 mils 0.21 mm
1 MHz 2.6 mils 0.066 mm
π fμσ 10 MHz 0.82 mils 0.021 mm
100 MHz 0.26 mils 0.0066 mm
1 GHz 0.0823 mils 0.0021 mm

November 2008 Bruce Archambeault, PhD 19


Inductance
• Current flow through metal =>
inductance!
• Fundamental element in EVERYTHING
• Loop area first order concern
• Inductive impedance increases with
frequency and is MAJOR concern at
high frequencies
X L = 2πfL
November 2008 Bruce Archambeault, PhD 20
Current Loop => Inductance

Courtesy of Elya Joffe

November 2008 Bruce Archambeault, PhD 21


Inductance Definition

• Faraday’s Law ∂B
∫ E ⋅ dl = − ∫∫ ∂t ⋅ dS
• For a simple rectangular loop
Area = A
∂B
V = −A
V
B
∂t
The minus sign means that the induced
voltage will work against the current that
originally created the magnetic field!

November 2008 Bruce Archambeault, PhD 22


Self Inductance
• Isolated circular loop ⎛ 8a ⎞
L ≈ μ 0 a⎜⎜ ln − 2 ⎟⎟
⎝ r0 ⎠
• Isolated rectangular loop

2μ0 a ⎜ ⎛ p + 1 + p 2
1 1 ⎞
2 ⎟
L= ln + −1 + 2 − 1+ p
π ⎜ 1+ 2 p p ⎟
⎝ ⎠
Note that inductance is directly influenced length of side
p=
by loop AREA and only less influenced by wire radius
conductor size!
November 2008 Bruce Archambeault, PhD 23
Partial Inductance
• Simply a way to break the overall loop
into pieces in order to find total
inductance

L2

L1 L3

L4 L total=Lp11+ Lp22 + Lp33 + Lp44 - 2Lp13 - 2Lp24

November 2008 Bruce Archambeault, PhD 24


Important Points About
Inductance

• Inductance is everywhere
• Loop area most important
• Inductance is everywhere

November 2008 Bruce Archambeault, PhD 25


Decoupling Capacitor Mounting
• Keep as to planes as close to capacitor
pads as possible

Via Separation

Inductance Depends
on Loop AREA
Height above Planes

November 2008 Bruce Archambeault, PhD 26


Via Configuration Can Change
Inductance
SMT Capacitor

Via
The “Good” Best
Capacitor Pads

The “Bad”
Better
The “Ugly”

Really “Ugly”

November 2008 Bruce Archambeault, PhD 27


Comparison of Decoupling Capacitor Impedance
100 mil Between Vias & 10 mil to Planes
1000

1000pF
100
0.01uF
0.1uF
1.0uF
Impedance (ohms)

10

0.1

0.01
1.0E+06 1.0E+07 1.0E+08 1.0E+09 1.0E+10
Frequency (Hz)

November 2008 Bruce Archambeault, PhD 28


Comparison of Decoupling Capacitor
Via Separation Distance Effects

Via Separation

10 mils

0603 Typical
Minimum
Dimensions

November 2008 Bruce Archambeault, PhD 29


Connection Inductance for Typical
Capacitor Configurations
Distance into board 0805 typical/minimum 0603 0402 typical/minimum
to planes (mils) (148 mils between via typical/minimum (106 mils between via
barrels) (128 mils between via barrels)
barrels)

10 1.2 nH 1.1 nH 0.9 nH


20 1.8 nH 1.6 nH 1.3 nH
30 2.2 nH 1.9 nH 1.6 nH
40 2.5 nH 2.2 nH 1.9 nH
50 2.8 nH 2.5 nH 2.1 nH
60 3.1 nH 2.7 nH 2.3 nH
70 3.4 nH 3.0 nH 2.6 nH
80 3.6 nH 3.2 nH 2.8 nH
90 3.9 nH 3.5 nH 3.0 nH
100 4.2 nH 3.7 nH 3.2 nH
November 2008 Bruce Archambeault, PhD 30
‘Ground’
• Ground is a place where
potatoes and carrots thrive!
• ‘Earth’ or ‘reference’ is more descriptive
• Original use of “GROUND”
• Inductance is everywhere

X L = 2πfL
November 2008 Bruce Archambeault, PhD 31
What we Really Mean when we
say ‘Ground’

• Signal Reference
• Power Reference
• Safety Earth
• Chassis Shield Reference

November 2008 Bruce Archambeault, PhD 32


‘Ground’ is NOT a Current Sink!

• Current leaves a driver on a trace and


must return (somehow) to its source
• This seems basic, but it is often forgotten,
and is most often the cause of EMC
problems

November 2008 Bruce Archambeault, PhD 33


‘Grounding’ Needs Low Impedance
at Highest Frequency
• Steel Reference Plate
– 4 milliohms/sq @ 100KHz
– 40 milliohms/sq @ 10 MHz
– 400 milliohms/sq @ 1 GHz
• A typical via is about 2
nH
– @ 100 MHz Z = 1.3 ohms
– @ 500 MHz Z = 6.5 ohms
– @ 1000 MHz Z = 13 ohms
– @ 2000 MHz Z = 26 ohms
November 2008 Bruce Archambeault, PhD 34
Where did the Term “GROUND”
Originate?

• Original Teletype connections


• Lightning Protection

November 2008 Bruce Archambeault, PhD 35


Ground/Earth

Teletype
Receiver

Teletype
Transmitter

November 2008 Bruce Archambeault, PhD 36


Ground/Earth

Teletype
Receiver

Teletype
Transmitter

November 2008 Bruce Archambeault, PhD 37


FIG 7 Lightning striking house

Lightning

November 2008 Bruce Archambeault, PhD 38


Lightning effect without rod

November 2008 Bruce Archambeault, PhD 39


Lightning effect with rod

Lightning
Lightning rod

November 2008 Bruce Archambeault, PhD 40


What we Really Mean when we
say ‘Ground’

• Signal Reference
• Power Reference
• Safety Earth
• Chassis Shield Reference

A
D

Circuit Chassis Digital Analog


“Ground” “Ground” “Ground” “Ground”
November 2008 Bruce Archambeault, PhD 41
November 2008 Bruce Archambeault, PhD 42
Schematic with return current
shown

Signal trace currents


IC1 IC2 IC3

Return currents on ground

November 2008 Bruce Archambeault, PhD 43


Actual Current Return is 3-Dimensional
Signal Trace

IC

Ground Vias

BOARD STACK UP: CURRENT LOCATION:


Signal Trace Signal Trace
Ground Via IC
Ground Layer

Ground Layer Ground Layer

November 2008 Bruce Archambeault, PhD 44


Low Frequency Return Currents
Take Path of Least Resistance

Driver
Receiver

Ground Plane

November 2008 Bruce Archambeault, PhD 45


High Frequency Return Currents Take
Path of Least Inductance

Driver
Receiver

Ground Plane

November 2008 Bruce Archambeault, PhD 46


PCB Example for Return Current
Impedance
Trace

GND Plane

22” trace
10 mils wide, 1 mil thick, 10 mils above GND plane
November 2008 Bruce Archambeault, PhD 47
PCB Example for Return Current
Impedance
Trace

GND Plane

Shortest DC path

For longest DC path, current returns under trace


November 2008 Bruce Archambeault, PhD 48
MoM Results for Current Density
Frequency = 1 KHz

November 2008 Bruce Archambeault, PhD 49


MoM Results for Current Density
Frequency = 1 MHz

November 2008 Bruce Archambeault, PhD 50


U-shaped Trace Inductance
PowerPEEC Results

0.6

0.55

0.5

0.45
inductance (uH)

0.4

0.35

0.3

0.25

0.2

0.15

0.1
1.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08
Frequency (Hz)

November 2008 Bruce Archambeault, PhD 51


Traces/nets over a Reference Plane

Microstrip Transmission Line

Signal Trace

Reference Planes
Dielectric

Stripline Transmission Line

November 2008 Bruce Archambeault, PhD 52


Traces/nets and Reference Planes
in Many Layer Board Stackup

Signal Traces

Reference Planes
(Power, “Ground”, etc.)

November 2008 Bruce Archambeault, PhD 53


Microstrip Electric/Magnetic Field Lines
(8mil wide trace, 8 mils above plane, 65 ohm)

Electric Field Lines

Vcc

Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 54


Microstrip Electric/Magnetic Field Lines
Common Mode
8 mil wide trace, 8 mils above plane, 65/115 ohm)

Electric Field Lines

Vcc
Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 55


Microstrip Electric/Magnetic Field Lines
Differential Mode
8 mil wide trace, 8 mils above plane, 65/115 ohm)

Electric Field Lines

Vcc
Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 56


Electric/Magnetic Field Lines
Symmetrical Stripline

GND

Vcc

Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 57


Electric/Magnetic Field Lines
Symmetrical Stripline (Differential)

GND

Vcc

Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 58


Electric/Magnetic Field Lines
Asymmetrical Stripline

Vcc

GND

Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 59


Electric/Magnetic Field Lines
Asymmetrical Stripline (Differential)

Courtesy of Hyperlynx

November 2008 Bruce Archambeault, PhD 60


What About Pseudo-Differential
Nets?
• So-called differential traces are NOT truly
differential
– Two complementary single-ended drivers
• Relative to ‘ground’
– Receiver is differential
• Senses difference between two nets (independent
of ‘ground’)
• Provides good immunity to common mode noise
• Good for signal quality/integrity

November 2008 Bruce Archambeault, PhD 61


Pseudo-Differential Nets Current in
Nearby Plane
• Balanced/Differential currents have
matching current in nearby plane
– No issue for discontinuities
• Any unbalanced (common mode) currents
have return currents in nearby plane that
must return to source!
– All normal concerns for single-ended nets
apply!
November 2008 Bruce Archambeault, PhD 62
Pseudo-Differential Nets
• Not really ‘differential’, since more closely
coupled to nearby plane than each other
• Slew and rise/fall variation cause common
mode currents!

November 2008 Bruce Archambeault, PhD 63


Differential Voltage Pulse with Skew
1 Gbit/sec with 95 psec rise/fall time
1.2

0.8
Complementary -- Line1
Complementary -- Line 2
Skew=2ps
Voltage

0.6 Skew=6ps
Skew = 10ps
Skew = 20ps
Skew = 30ps
0.4
Skew =40ps
Skew =50ps
Skew =60ps
0.2

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (nsec)
November 2008 Bruce Archambeault, PhD 64
Common Mode Voltage
From Differential Voltage Pulse with Skew
1 Gbit/sec with 95 psec rise/fall time
0.6

Balanced
0.4 Skew=2ps
Skew=6ps
Skew =10ps
Skew =20ps
0.2
Skew =30ps
Skew =40ps
Skew =50ps
Voltage

-0.2

-0.4

-0.6
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (nsec)
November 2008 Bruce Archambeault, PhD 65
Common Mode Current
From Differential Voltage Pulse with Skew
1 Gbit/sec with 95 psec Rise/fall Time
100

80

60
Balanced
40 Skew=2ps
Skew=6ps
Skew =10ps
20
Skew =20ps
Level (ma)

Skew =30ps
0
Skew =40ps
Skew =50ps
-20 Skew =60ps

-40

-60

-80

-100
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (nsec)

November 2008 Bruce Archambeault, PhD 66


Common Mode Current
From Differential Voltage Pulse with Skew
1 Gbit/sec with 95 psec Rise/fall Time
150

Skew=2ps
140 Skew=6ps
Skew =10ps
130 Skew =20ps
Skew =30ps
120 Skew =40ps
Skew =50ps
110 Skew =60ps
Level (dBuA)

100

90

80

70

60

50
1.E+08 1.E+09 1.E+10 1.E+11
Frequency (Hz)

November 2008 Bruce Archambeault, PhD 67


Differential Voltage Pulse with Rise/Fall Variation/Unbalance
1 Gbit/sec with 95 psec Nominal Rise/Fall Time
1.2

0.8

Original Pulse rise=95ps


Level (volts)

Complementary Pulse Rise=90ps


0.6 Complementary Pulse Rise=80ps
Complementary Pulse Rise=105ps
Complementary Pulse Rise=115ps

0.4

0.2

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (ns)
November 2008 Bruce Archambeault, PhD 68
Common Mode Voltage
From Differential Voltage Pulse with Various Rise/Fall Unbalance
1 Gbit/sec with 95 psec Nominal Rise/Fall Time
0.2

0.15

0.1

0.05
Voltage

-0.05

Complementary Pulse Rise=90ps


-0.1
Complementary Pulse Rise=80ps
Complementary Pulse Rise=105ps
-0.15
Complementary Pulse Rise=115ps

-0.2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (ns)
November 2008 Bruce Archambeault, PhD 69
Common Mode Current
From Differential Voltage Pulse with Various Rise/Fall Unbalance
1 Gbit/sec with 95 psec Nominal Rise/fall Time
60

40

20
Current (ma)

-20

Complementary Pulse Rise=90ps


-40 Complementary Pulse Rise=80ps
Complementary Pulse Rise=105ps
Complementary Pulse Rise=115ps
-60
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (ns)
November 2008 Bruce Archambeault, PhD 70
Common Mode Current
From Differential Voltage Pulse with Various Rise/Fall Unbalance
1 Gbit/sec with Nominal 95 psec Rise/fall Time
90

85

80 Complementary Pulse Rise=90ps


Complementary Pulse Rise=80ps
Complementary Pulse Rise=105ps
75 Complementary Pulse Rise=115ps
Level (dBua)

70

65

60

55

50
1.E+08 1.E+09 1.E+10 1.E+11
Frequency (Hz)

November 2008 Bruce Archambeault, PhD 71


Antenna Structures
Dipole antenna

Non-Dipole antenna

PCB GND planes

November 2008 Bruce Archambeault, PhD 72


Board-to-Board Differential Pair
Issues
PC
B
Pl
an
e
2

Connecto Micros
r trip

Microstrip

V
PCB Plan Ground-to-Ground
e 1
noise
November 2008 Bruce Archambeault, PhD 73
Example Measured Differential
Individual Signal-to-GND

500 mV P-P (each) Individual Differential


Signals ADDED
Common Mode Noise
170 mV P-P
November 2008 Bruce Archambeault, PhD 74
Measured GND-to-GND Voltage

205 mV P-P

November 2008 Bruce Archambeault, PhD 75


Pin Assignment Controls
Inductance for CM signals

37.17 nH 25.21 nH
(a) (b)

16.85 nH 20.97 nH
(c) (d)
Signal Pin Related Ground Pins
November 2008 Bruce Archambeault, PhD 76
Different pins within Same Pair may
have Different Loop Inductance for CM
“Ground” pins Differential pair

4 3
pin 1 -- 26.6nH

2 1 pin 2 -- 23.6nH
pin 3 -- 31.8nH
pin 4 -- 28.8nH

November 2008 Bruce Archambeault, PhD 77


Pseudo-Differential Net Summary
• Small amounts of skew can cause
significant common mode current
• Small amount of rise/fall time deviation
can cause significant amount of common
mode current
• Discontinuities (vias, crossing split planes,
etc) and convert significant amount of
differential current into common mode
current
November 2008 Bruce Archambeault, PhD 78
Return Current vs. “Ground”

• For high frequency signals, “Ground” is a


concept that does not exist
• The important question is “where does
the return current flow?”

November 2008 Bruce Archambeault, PhD 79


Referencing Nets
(Where does the Return Current Flow??)
• Microstrip/Stripline across split in
reference plane
• Microstrip/Stripline through via (change
reference planes)
• Mother/Daughter card

November 2008 Bruce Archambeault, PhD 80


Microstrip/Stripline Across Split
in Reference Plane
• Don’t Cross Splits with Critical Signals!!!
– Bad practice
– Stitching capacitor required across split to
allow return current flow
• must be close to crossing
• must have low inductance
• limited frequency effect --- due to inductance
– Major source of Common Mode current!
November 2008 Bruce Archambeault, PhD 81
Splits in Reference Plane

• Power planes often have splits


• Return current path interrupted
• Consider spectrum of clock signal
• Consider stitching capacitor
impedance
• High frequency harmonics not
returned directly
November 2008 Bruce Archambeault, PhD 82
Split Reference Plane Example

PWR

GND

November 2008 Bruce Archambeault, PhD 83


Split Reference Plane Example
With Stitching Capacitors

Stitching Capacitors
Allow Return
current to Cross
PWR Splits ???
GND

November 2008 Bruce Archambeault, PhD 84


Capacitor Impedance
Measured Impedance of .01 uf Capacitor

100.0

10.0
Impednace (ohms)

1.0

0.1
1.E+06 1.E+07 1.E+08 1.E+09
Frequency (Hz)

November 2008 Bruce Archambeault, PhD 85


Frequency Domain Amplitude of Intentional Current Harmonic Amplitude
From Clock Net
160

140

120
level (dBuA)

100

80

60

40
0 200 400 600 800 1000 1200 1400 1600 1800 2000
freq (MHz)

November 2008 Bruce Archambeault, PhD 86


MoM Microstrip Model Current
Distribution Example

November 2008 Bruce Archambeault, PhD 87


MoM Microstrip Model Current
Distribution Example

November 2008 Bruce Archambeault, PhD 88


Emissions From Board
• Far field emissions not important unless it
is an unshielded product
• Near field emissions above board ARE
important
• Example of emissions from board with
critical net crossing split reference plane

November 2008 Bruce Archambeault, PhD 89


Near Field Radiation from Microstrip on Board
with Split in Reference Plane
Comparison of Maximum Radiated E-Field for Microstrip
With and without Split Ground Reference Plane
120

110

100
Maximum Radiated E-Field (dBuv/m)

90

80

70

60 No-Split

Split
50

40

30

20
10 100 1000
Frequency (MHz)

November 2008 Bruce Archambeault, PhD 90


With “Perfectly Connected” Stitching Capacitors
Across Split
Comparison of Maximum Radiated E-Field for Microstrip
With and without Split Ground Reference Plane and Stiching Capacitors
120

110

100
Maximum Radiated E-Field (dBuv/m)

90

80

70

60
No-Split
50 Split

Split w/ one Cap


40
Split w/ Two Caps

30

20
10 100 1000
Frequency (MHz)
November 2008 Bruce Archambeault, PhD 91
Stitching Caps with Via Inductance
Comparison of Maximum Radiated E-Field for Microstrip
With and without Split Ground Reference Plane and Stiching Capacitors
120

110

100
Maximum Radiated E-Field (dBuv/m)

90

80

70

60
No-Split
50 Split
Split w/ one Cap
Split w/ Two Caps
40 Split w/One Real Cap
Split w/Two Real Caps
30

20
10 100 1000
Frequency (MHz)

November 2008 Bruce Archambeault, PhD 92


Example of Common-Mode Noise Voltage Across Split Plane
Vs. Stitching Capacitor Distance to Crossing Point
25

20

15
Gap Voltage

100MHz
200MHz
10 300MHz
400MHz
500MHz
600MHz
700MHz
5
800MHz
900MHz
1000MHz

0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Distance (mils)

November 2008 Bruce Archambeault, PhD 93


Are Stitching Capacitors
Effective ???
• YES, at low frequencies
• No, at high frequencies
• Need to limit the high frequency current
spectrum
• Need to avoid split crossings with ALL
critical signals

November 2008 Bruce Archambeault, PhD 94


Pin Field Via Keepouts??

s
d

Return Current must go around Return current path


entire keep out area --- just as bad deviation minimal
as a slot
Recommend s/d > 1/3

November 2008 Bruce Archambeault, PhD 95


Changing Reference Planes
Six-Layer PCB Stackup Example

Signal Layer

Plane
Signal Layer

Signal Layer
Plane
Signal Layer

November 2008 Bruce Archambeault, PhD 96


Microstrip/Stripline through via
(change reference planes)

Via Trace

November 2008 Bruce Archambeault, PhD 97


How can the Return Current Flow
When Signal Line Goes Through Via??

What happens to Return Current


in this Region?

Return Current

November 2008 Bruce Archambeault, PhD 98


How can the Return Current Flow
When Signal Line Goes Through
Via??
• Current can NOT go from one side of the
plane to the other through the plane
– skin depth
• Current must go around plane at via hole,
through decoupling capacitor, around second
plane at the second via hole!
• Use displacement current between planes

November 2008 Bruce Archambeault, PhD 99


Return Current Across
Reference Plane Change

What happens to Return Current in


this Region?

Reference Planes

Displacement Current

Return Current Trace Current

November 2008 Bruce Archambeault, PhD 100


Return Current Across
Reference Plane Change
With Decoupling Capacitor
Decoupling Capacitor

Displacement Current
Reference
Planes
Return Current

November 2008 Bruce Archambeault, PhD 101


Return Current Across
Reference Plane Change
With Decoupling Capacitor (on Top)

Decoupling Capacitor

Common-Mode Current

Displacement Current
Reference Planes
Return Current

November 2008 Bruce Archambeault, PhD 102


Location of Decoupling
Capacitors (Relative to Via) is
Important!

• One Decoupling Capacitor at 0.5”


• Two Decoupling Capacitors at 0.5”
• Two Decoupling Capacitors at 0.25”

November
June 20072008 Bruce Archambeault, PhD 103103
RF Current @ 700 MHz with One
Capacitor 0.5” from Via

November
June 20072008 Bruce Archambeault, PhD 104104
RF Current @ 700 MHz with One
Capacitor 0.5” from Via
(expanded view)

November
June 20072008 Bruce Archambeault, PhD 105105
RF Current @ 700 MHz with Two
Capacitors 0.5” from Via

November
June 20072008 Bruce Archambeault, PhD 106106
RF Current @ 700 MHz with One
Capacitor 0.5” from Via
(Expanded view)

November
June 20072008 Bruce Archambeault, PhD 107107
RF Current @ 700 MHz with Two
Capacitors 0.25” from Via

November
June 20072008 Bruce Archambeault, PhD 108108
RF Current @ 700 MHz with Two
Capacitors 0.25” from Via
(expanded view)

November
June 20072008 Bruce Archambeault, PhD 109109
RF Current @ 700 MHz with One
REAL Capacitor 0.5” from Via

November
June 20072008 Bruce Archambeault, PhD 110110
RF Current @ 700 MHz with Two
REAL Capacitors 0.5” from Via

November
June 20072008 Bruce Archambeault, PhD 111111
RF Current @ 700 MHz with Two
REAL Capacitors 0.25” from Via

November
June 20072008 Bruce Archambeault, PhD 112112
Possible Routing Options
Six-Layer Board
Bad
Signal Layer
Reference Plane
Signal Layer
Signal Layer
Reference Plane
Signal Layer

Bad
Signal Layer
Reference Plane
Signal Layer
Signal Layer
Reference Plane
Signal Layer

Good
Signal Layer
Reference Plane
Signal Layer
Signal Layer
Reference Plane
Signal Layer
November 2008 Bruce Archambeault, PhD 113
Compromise Routing Option for
Many Layer Boards

Good Compromise

Vcc1

Gnd
Reference
Plane

Lot’s of Decoupling caps


near ASIC

November 2008 Bruce Archambeault, PhD 114


Typical Driver/Receiver Currents
VCC
IC switch
IC load
driver
VDC
Z0, vp CL

logic 0-to-1 GND


logic 1-to-0
VCC VCC
IC load IC load
IC
IC
driver charge discharge
driver
Z0, vp Z0, vp
VCC 0V

GND GND
November 2008 Bruce Archambeault, PhD 115
Suppose The Trace is Routed Next
to Power (not Gnd)
TEM Transmission
Vcc1
Line Area

Vcc1
“Fuzzy” Return
Path Area “Fuzzy” Return
Path Area
Return Path Options:
-- Decoupling Capacitors
-- Distributed Displacement Current
November 2008 Bruce Archambeault, PhD 116
Suppose The Trace is Routed Next
to a DIFFERENT Power (not Gnd)
TEM Transmission
Vcc1
Line Area

Vcc2
“Fuzzy” Return
Path Area “Fuzzy” Return
Path Area

Return Path Options:


-- Decoupling Capacitors ??? May not be any nearby!!
-- Distributed Displacement Current – Increased current spread!!!

November 2008 Bruce Archambeault, PhD 117


Via Summary
9 Route critical signals on either side of ONE
reference plane
9 Drop critical signal net to selected layer close
to driver/receiver
9 Many decoupling capacitors to help return
currents
9 Do NOT change reference planes on critical
nets unless ABSOLUTELY NECESSARY!!
9 Make sure at least 2 decoupling capacitors
within 0.2” of via with critical signals
November 2008 Bruce Archambeault, PhD 118
Mother/Daughter Board
Connector Crossing

• Critical Signals must be referenced to


same plane on both sides of the connector

November 2008 Bruce Archambeault, PhD 119


Mother/Daughter Board
Connector Crossing

Signal Path Connector

GND

PWR

Signal Layers

November 2008 Bruce Archambeault, PhD 120


Return Current from Improper
Referencing Across Connector
Decoupling
Displacement Capacitors
Current

Signal Path Connector


Return current

GND

PWR

Signal Layers

November 2008 Bruce Archambeault, PhD 121


Return Current from Proper
Referencing Across Connector

Signal Path Connector

GND

PWR

Signal Layers
Return current

November 2008 Bruce Archambeault, PhD 122


How Many “Ground” Pins Across
Connector ???
• Nothing MAGICAL about “ground”
• Return current flow!
• Choose the number of power and “ground”
pins based on the number of signal lines
referenced to power or “ground” planes
• Insure signals are referenced against
same planes on either side of connector

November 2008 Bruce Archambeault, PhD 123


Think about Return Currents!!
9Reference plane should be continuous
under all critical traces
9When Vias are necessary make sure there
are two close decoupling capacitors
9When crossing a connector to a second
board, make sure the critical trace is
referenced to the same reference plane as
the primary board

November 2008 Bruce Archambeault, PhD 124


Ground-Reference Plane Noise
(Voltage Difference Across Plane)
• Connection of large PC ‘ground’ planes
to chassis important
– ESD current can result in voltage difference
across ‘ground’ plane
– Looks like input pulse to circuits
– More connection to chassis will reduce this
voltage difference

November 2008 Bruce Archambeault, PhD 125


Connection to Chassis
PCB gnd plane
Good connection in
I/O area important
for emissions
control!! Chassis

Screw post
Connection to chassis away
from I/O area NOT important
for emissions control

November 2008 Bruce Archambeault, PhD 126


Connection to Chassis
for ESD Control
PCB gnd plane

Chassis

Screw post
Distributed Connection to chassis
away from I/O area very
important for ESD control

November 2008 Bruce Archambeault, PhD 127


Contacts for Chassis
Connection

Want this!
Screw head contact pad
on top of PC Board

Screw head

NOT this!

Copper pad Vias to “Ground” plane

November 2008 Bruce Archambeault, PhD 128


Model for Current Simulations
PCB gnd plane
Screw post
Trace Source

Chassis
ESD Voltage
Between Chassis and
gnd plane
Trace Load

November 2008 Bruce Archambeault, PhD 129


Comparison of Trace Load Noise Voltage
for 1 Kv ESD Pulse from PCB GND to Chassis
2
No Connection to Chassis
One connection to Chassis (Near I/O)
Four Connections to Chassis (Near I/O)
1.5
Eight Connections to Chassis
16 Connections to Chassis
20 Connections to Chassis
1
Load Voltage (volts)

0.5

-0.5

-1

-1.5
0 0.5 1 1.5 2 2.5 3 3.5
Time (ns)

November 2008 Bruce Archambeault, PhD 130


Comparison of Trace Load Noise Voltage
for 1 Kv ESD Pulse from PCB GND to Chassis
2.5

1.5
No Connection to Chassis

1 One connection to Chassis (Near I/O)


Load Voltage (volts)

0.5 Four Connections to Chassis (Near I/O)

0 Eight Connections to Chassis

-0.5 16 Connections to Chassis

20 Connections to Chassis
-1
Eight Connections to Chassis (4 @
-1.5 each end)

-2

-2.5

-3
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Time (ns)
November 2008 Bruce Archambeault, PhD 131
Current Flow w/One Screw Post

November 2008 Bruce Archambeault, PhD 132


Current Flow w/Eight Screw Posts

November 2008 Bruce Archambeault, PhD 133


Current Flow w/20 Screw Posts

November 2008 Bruce Archambeault, PhD 134


Current Flow w/Eight Screw Posts (4 each end)

November 2008 Bruce Archambeault, PhD 135


Number ONE Problem

• Intentional signal return


current

November 2008 Bruce Archambeault, PhD 136


Where to Go for More?

• Limited selection of EMC design books


– Beware of some popular books!!!
– “PCB Design for Real-World EMI Control” (good choice)
• Bruce Archambeault

• EMC ‘experts’
– Experience is important
– Again, beware ---- ask questions and understand WHY
• Cookbooks do not work! Every case is
special and different

November 2008 Bruce Archambeault, PhD 137

You might also like