Академический Документы
Профессиональный Документы
Культура Документы
April 2010
IBM
1.5
0.5
Voltage (volts)
-0.5
-1
-1.5
0 1 2 3 4 5 6 7 8 9 10
Time (usec)
April 2010 Dr. Bruce Archambeault, IBM 12
Frequency Domain Example
Sine Wave Spectrum
Sine Wave Spectrum
35
30
25
20
Level
15
10
0
10 100 1000 10000 100000
fundamental
0.5
Amplitude
-0.5
-1
-1.5
0 5 10 15 20 25
time (ns)
fundamental
1
3rd harmonic
(Normalized) Fundamental and 3rd harmonic
0.5
Amplitude
-0.5
-1
-1.5
0 5 10 15 20 25
time (ns)
fundamental
1 3rd harmonic
5th harmonic
(Normalized) Fundamental and 3rd & 5th harmonic
0.5
Amplitude
-0.5
-1
-1.5
0 5 10 15 20 25
time (ns)
7th harmonic
-0.5
-1
-1.5
0 5 10 15 20 25
time (ns)
fundamental
3rd harmonic
1 5th harmonic
7th harmonic
9th harmonic
(Normalized) Fundamental and 3rd,5th,7th,9th harmonic
0.5
Amplitude
-0.5
-1
-1.5
0 5 10 15 20 25
time (ns)
1 5th harmonic
7th harmonic
9th harmonic
-0.5
-1
-1.5
0 5 10 15 20 25
time (ns)
0.5
Amplitude
0
fundamental
3rd harmonic
-0.5 5th harmonic
7th harmonic
9th harmonic
-1 11th harmonic
13th harmonic
(Normalized) Fundamental and 3rd,5th,7th,9th,13th harmonic
-1.5
0 5 10 15 20 25
time (ns)
-10
-20
Normalized Amplitude (dB)
-30
-40
-50
-60
-70
-80
-90
0 100 200 300 400 500 600 700 800
Frequency (MHz)
0.5
Amplitude
Pulse
-0.5
Sinewave
-1
-1.5
0 10 20 30 40 50 60 70 80 90 100
Time (nsec)
Pseudo-random
-10 Squarewave
Relative Level (dB)
-20
-30
-40
-50
-60
0 100 200 300 400 500 600 700 800 900 1000
Frequency (MHz)
[something ]
d Changing with
respect to time
dt
[something ]
d Changing with respect
to position (x)
dx
April 2010 Dr. Bruce Archambeault, IBM 28
Partial Derivative
How fast is something changing for
one variable?
Changing with respect
to time (as x is
[something (t , x )] constant)
t
dl
stop
V = ( E dl )
start
April 2010 Dr. Bruce Archambeault, IBM 31
Line Integral
piece of E field
dl
stop
V = [ ( E x dx ) + ( E y dy ) ]
start
April 2010 Dr. Bruce Archambeault, IBM 32
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
April 2010 Dr. Bruce Archambeault, IBM 33
Line Integral -- Closed
Closed line integrals
find the path length
And/or the amount of
some quantity along
that closed path length
Area = da
da = dx dy
Area = dx dy
As dx and dy become
smaller and smaller, the
area is better calculated
April 2010 Dr. Bruce Archambeault, IBM 35
Closed Surface Integral
Find the surface area
of a closed shape
shape da
Volume = dv
dv = dx dy dz
Volume = [dx dy dz ]
D
across a small piece of space Density with
respect to time
H = J +
t
A difference in Electric Field
B
across a small piece of space
E =
A change in
Magnetic Flux
t
Density with
respect to time
Faradays Law
Gauss Law
Amperes Law
Stokes Theorem
Many others
Design intentionally
Pass the first time!
X L = 2fL
April 2010 Dr. Bruce Archambeault, IBM 52
Current Loop = Inductance
B
V = A
t
V
B
Loop #2
Loop #1 r
2 = B1 (r ) n dS 2
S2
L1 L3
Inductance is everywhere
Loop area most important
Inductance is everywhere
Via Separation
Inductance Depends
on Loop AREA
Height above Planes
Via
The Good Best
Capacitor Pads
The Bad
Better
The Ugly
Really Ugly
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)
60 mils
40 mils
86 mils minimum
106 mils typical
*Note: Minimum
distance is 10 mils but
more typical distance is
April 2010 Dr. Bruce Archambeault, IBM 20 mils 66
Connection Inductance for Typical Capacitor Configurations
X L = 2fL
April 2010 Dr. Bruce Archambeault, IBM 73
What we Really Mean when we
say Ground
Signal Reference
Power Reference
Safety Earth
Chassis Shield Reference
GND via
GND via
GND via
Metal Enclosu
re
GND via
GND via
GND via
Metal Enclosu
re
Teletype
Receiver
Teletype
Transmitter
Teletype
Receiver
Teletype
Transmitter
Lightning
Lightning
Lightning rod
A
D
Data Cable
Ground wire
System #2
System #1
Ground wire
System #2
System #1
IC
Ground Vias
Driver
Receiver
Ground Plane
Ground Plane
GND Plane
22 trace
10 mils wide, 1 mil thick, 10 mils above GND plane
April 2010 Dr. Bruce Archambeault, IBM 96
PCB Example for Return Current
Impedance
Trace
GND Plane
Shortest DC path
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)
Signal Trace
Reference Planes
Dielectric
Signal Traces
Reference Planes
(Power, Ground, etc.)
Vcc
Courtesy of Hyperlynx
Vcc
Courtesy of Hyperlynx
Vcc
Courtesy of Hyperlynx
GND
Vcc
Courtesy of Hyperlynx
GND
Vcc
Courtesy of Hyperlynx
Vcc
GND
Courtesy of Hyperlynx
Courtesy of Hyperlynx
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)
April 2010 Dr. Bruce Archambeault, IBM 115
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)
April 2010 Dr. Bruce Archambeault, IBM 116
Differential Voltage Pulse with Rise/Fall Variation/Unbalance
1 Gbit/sec with 95 psec Nominal Rise/Fall Time
1.2
0.8
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)
April 2010 Dr. Bruce Archambeault, IBM 117
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
-0.2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 118
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
April 2010 Dr. Bruce Archambeault, IBM 119
Example Measured Differential
Individual Signal-to-GND
205 mV P-P
Non-Dipole antenna
37.17 nH 25.21 nH
(a) (b)
16.85 nH 20.97 nH
(c) (d)
Signal Pin Related Ground Pins
4 3
pin 1 -- 26.6nH
pin 2 -- 23.6nH
2 1
pin 3 -- 31.8nH
pin 4 -- 28.8nH
bit rate
rise time
not enough information!
20 dB/decade
d
Amplitude (S) Log Scale
40 dB/decade
1/(d) 1/(t)
600
400
300
200
100
0
0 0.5 1 1.5 2 2.5 3
Rise Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 133
Time Domain comparison of Effect on Voltage Waveform
with Good and Poor Termination
3.5
2.5
2
Amplitude (volts)
1.5 Poor-Term
Good-Term
0.5
-0.5
0.0E+00 5.0E-09 1.0E-08 1.5E-08 2.0E-08 2.5E-08 3.0E-08 3.5E-08
Time (sec)
8.E-02
6.E-02
4.E-02
Amplitude (amps)
2.E-02
Poor-Term
0.E+00 Good-Term
-2.E-02
-4.E-02
-6.E-02
-8.E-02
0.0E+00 5.0E-09 1.0E-08 1.5E-08 2.0E-08 2.5E-08 3.0E-08 3.5E-08
Time (sec)
120
110
Amplitude (dBuv)
100
Term-Poor
90 Good Term
80
70
60
1.E+07 1.E+08 1.E+09
Frequency (Hz)
80
70
60
Amplitude (dBuA)
50
Term-Poor
40 Good Term
30
20
10
0
1.E+07 1.E+08 1.E+09
Frequency (Hz)
IL sin j 1 1
E = 2 + 2+ 3
4 0 c r cr j r
PWR
GND
Stitching Capacitors
Allow Return
current to Cross
PWR Splits ???
GND
100.0
10.0
Impednace (ohms)
1.0
0.1
1.E+06 1.E+07 1.E+08 1.E+09
Frequency (Hz)
140
120
level (dBuA)
100
80
60
40
0 200 400 600 800 1000 1200 1400 1600 1800 2000
freq (MHz)
110
100
Maximum Radiated E-Field (dBuv/m)
90
80
70
60 No-Split
Split
50
40
30
20
10 100 1000
Frequency (MHz)
110
100
Maximum Radiated E-Field (dBuv/m)
90
80
70
60
No-Split
50 Split
30
20
10 100 1000
Frequency (MHz)
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)
Split
Width
Distance to
Capacitor
8 d
0
0 100 200 300 400 500 600 700
Distance to Capacitor (mils)
1.5
h1
0.5
Split Plane
h2
Solid Plane
0
0 100 200 300 400 500 600 700
Distance to Capacitor (mils)
1.5
2 h1
Split Plane
h3
Solid Plane
Transfer Inductance (nH)
1.5
No Additional Plane
0.5
Ratio h3/h1 = 1.0
Ratio h3/h1 = .666
Ratio h3/h1 = 1.333
0
0 100 200 300 400 500 600 700
Distance to Capacitor (mils)
Via Separation
Inductance Depends
on Loop AREA
Height above Planes
s
d
Via Trace
Return Current
Reference Planes
Displacement Current
PWR
Displacement Current
Reference
Planes
Return Current
Decoupling Capacitor
Common-Mode Current
Displacement Current
Reference Planes
Return Current
Via
Good
Possible Routing
Layers Planes
Good Compromise
Vcc1
Gnd
Reference
Plane
VCC GND
VCC
IC
charge IC load IC load
driver discharge
IC
driver
Z0, vp VCC
Z0, vp 0V
GND
logic 0-to-1 GND logic 1-to-o
April 2010 Dr. Bruce Archambeault, IBM 189
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
April 2010 Dr. Bruce Archambeault, IBM 190
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!!!
April 2010 Dr. Bruce Archambeault, IBM 191
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 (or GND vias) 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 0.25 of
via (change reference planes) with critical signals
Signal
Signal
PWR
GND
Signal Path Connector
Signal
GND
PWR
Signal
Signal
Signal
PWR
GND
Displacement De-cap
Current
De-cap
Signal
GND
PWR
Signal
Signal
Signal
PWR
GND
Return Path
Connector
Signal Path
Signal
GND
PWR
Signal
Difference in
edge current
High current
Low current
Trace
120
110
Electric Field Amplitude (dBuv/m)
100
90
80
70
60
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Distance from Board Edge (mils)
Microstrip h
I Edge =
I Signal wg 2 (wg / 2 ) d
arctan
[ ]
2 2h
Isignal = the current amplitude (in linear scale) for each harmonic of the current waveform.
wg = the width of the ground-refer4ence plane (set to 1000 always)
d = the distance from the trace to the edge of the ground-reference plane
h = the height between the trace and the ground-reference plane
I Edge =
I Signal wg 2 (wg / 2 ) d
arctan
[ ]
2 2 2h
Isignal = the current amplitude (in linear scale) for each harmonic of the current waveform.
wg = the width of the ground-refer4ence plane (set to 1000 always)
d = the distance from the trace to the edge of the ground-reference plane
h = the height between the trace and the ground-reference plane
Asymmetrical Stripline h1
d
w
h2
I Edge Nearest =
I Signal h2
arctan
[
wg 2 (wg / 2 ) d ]
h1 2 2h2
I Edge Furthest =
I Signal h1
arctan g
[
w 2 (wg / 2 ) d ]
h2 2 2h1
Isignal = the current amplitude (in linear scale) for each harmonic of the current waveform.
wg = the width of the ground-refer4ence plane (set to 1000 always)
d = the distance from the trace to the edge of the ground-reference plane
h1 and h2 = the height between the trace and the ground-reference plane
April 2010 Dr. Bruce Archambeault, IBM 210
To Prevent/Reduce Intentional
Signal -- Common Mode
Emissions
9Keep traces away from board edge
Reduces Return current along edge of reference plane
9Improve termination to reduce high
frequency content
If the high frequency currents are not created, they can
not radiate!
High current
Low current
Trace
Split in Plane
Treat the Ground pin as a SIGNAL Pin
Serial, Parallel
Audio,
keyboard, mouse
High Speed
Circuits
Ferrite Beads
High Speed I/o, Video,
SCSI, etc
10000
1000
Impedance (ohms)
100
Split=.125cm
Split=.25cm
Split=.5cm
10 Split=.75
Split=1cm
1
10 100 1000
Frequency (MHz)
External Cable
External Cable
Split in Plane
80
Maximum Radiated E-Field (dBuv/m)
60
40
20
0 Without Wire
No Split
-20
Split I/O Ground Plane
-40
10 100 1000
Frequency (MHz)
Serial, Parallel
Audio,
keyboard, mouse
High Speed
Circuits
Ferrite Beads
High Speed I/o, Video,
SCSI, etc
To IC
unbalanced I/O
load signal
GND
Stand via
off
GND
plane
Metal
Enclosure
Ground
Bounce
20 cm
I/O Pin/Via
Signal Via
30 cm
FR4 r=4.2 Loss tan=0.02
-20
Transfer Function (dB)
-30
-40
-50
-60
-70
1.E+08 1.E+09 1.E+10
Frequency (Hz)
#1 #2 #3 #4 #5
10
9
#6 #7 #8 #9 #10
5
1
3
Figure 1
6
9
April 2010 Dr. Bruce Archambeault, IBM 241
S21 Used for Decoupling Goodness
Board
-10
Capacitance
Dominates
-20
-30
S21 (dB)
-40
-50
-60
-70
Physical Board Size Resonances Dominate
-80
0.0E+00 1.0E+08 2.0E+08 3.0E+08 4.0E+08 5.0E+08 6.0E+08 7.0E+08 8.0E+08 9.0E+08 1.0E+09
Frequency (Hz)
April 2010 Dr. Bruce Archambeault, IBM 243
Test Board Decoupling Capacitor
Placement for 25 .01 uf Caps
Possible Cap
Location
Populated
Cap
Locations
Possible Cap
Location
Populated
Cap
Locations
-10
-20
-30
-40
S21 (dB)
-50
-60
-70 No Caps
25 Caps
-80 50 Caps
99 Caps
-90
-100
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09
Frequency (Hz)
No Caps
-10 25 Caps
51 Caps
-20 99 Caps
S21 (dB)
-30
-40
-50
-60
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09
Frequency (Hz)
10000
1000
|Z| (Ohms)
100
10
0.1
1.00E+6 1.00E+7 1.00E+8 1.00E+9 1.00E+10
Freq (Hz)
Possible Cap
Location
Populated
Cap
Locations
-10
9 22pf Caps
-20 17 22pf Caps
21 22pf Caps
S21 (dB)
33 22pf Caps
-30
41 22pf Caps
99 22pf Caps
-40 99 Caps
-50
-60
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09
Frequency (Hz)
-10
-20
S21 Transfer Function (dB)
-30
-40
-50
No Caps
-60 99 0.01 uF Caps
0.01 uF and 330 pF Caps
-70
-80
-90
-100
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09
Frequency (Hz)
-10.0
-20.0
-30.0
-40.0
S21 (dB)
-50.0
Measured
-70.0
-80.0
-90.0
-100.0
0.0E+00 1.0E+08 2.0E+08 3.0E+08 4.0E+08 5.0E+08 6.0E+08 7.0E+08 8.0E+08 9.0E+08 1.0E+09
Frequency (Hz)
-10
-20
-30
S21 (dB)
-40
-50
99 .01uF caps
99 0.1uF caps
-60
99 0.33uF caps
-80
0.0E+00 1.0E+08 2.0E+08 3.0E+08 4.0E+08 5.0E+08 6.0E+08 7.0E+08 8.0E+08 9.0E+08 1.0E+09
Frequency (Hz)
Inductance Depends
on Loop AREA
Height above Planes
Via
The Good Best
Capacitor Pads
The Bad
Better
The Ugly
Really Ugly
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)
10 mils
0.1 uF Capacitor
Spacing Complex Simple rect loop Complex Formula Simple rect loop
between Formula (20 mils to plane) (10 mils to plane) (10 mils to plane)
Vias (20 mils to plane)
0805 + 2*10mil 3.0 nH 3.1 nH 2 nH 1.38 nH
Fan, Jun, Wei Cui, James L. Drewniak, Thomas Van Doren, and James L. Knighten, Estimating the
Noise Mitigating Effect of Local Decoupling in Printed Circuit Boards, IEEE Trans. on Advanced
Packaging, Vol. 25, No. 2, May 2002, pp. 154-165.
April 2010 Dr. Bruce Archambeault, IBM 263
Other Design Possibilities
So-called Buried Capacitance
Reduces high frequency transfer function
Allows less capacitors to be used
Really should be called increased distributed
capacitance
Lossy decoupling
Reduces high frequency transfer function
Allows less capacitors to be used
-10
-20
-30
Transfer function (dB)
-40
-50
-60
-70
2 mil - No Caps
-100
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Frequency (Ghz)
April 2010 Dr. Bruce Archambeault, IBM 267
Transfer function for Decoupling Board 10" x 12"
with Various Power/Ground Plane Separation
0
-10
-20
-30
Transfer function (dB)
-40
-50
-60
-20.0
-30.0
-40.0
Transfer Function (dB)
-50.0
-60.0 99 RC Terminations
99 .01uF caps
-30
-40
-50
-60
-70
-80
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09 2.0E+09
Frequency (Hz)
10
0
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09 1.0E+10
Frequency (Hz)
April 2010 Dr. Bruce Archambeault, IBM 273
Transient Analysis
(Time Limited)
GND
VCC VCC
IC IC
charge IC load discharge IC load
driver driver
Z0, vp VCC Z0, vp 0V
shoot-thru shoot-
current GND thru GND
logic 0-1 current logic 1-0
VCC
DC/DC SMT
converter electrolytic capacitors VCC/GND
capacitors plane
ultimate charge
source 100s uF 0.01-100s uF pFs-100 nF
100
.1uF
.01uF
.001uF
.0001uF
Impedance (ohms)
10
All in Parallel
0.1
0.01
1.00E+07 1.00E+08 1.00E+09
Frequency (Hz)
April 2010 Dr. Bruce Archambeault, IBM 280
Traditional Analysis #2
Calculate loop area Traditional loop
Inductance formulas
Which loop area? Which size conductor
IC Power Pin
Distributed
capacitors
ESR
Loop L
Note: L increases
Capacitance as distance from
source increases
Source
Power
Gnd
ESR
Lcap + LIC
Source
Nominal
Capacitance
0.1
0.01
1.E+07 1.E+08 1.E+09
Frequency (Hz)
April 2010 Dr. Bruce Archambeault, IBM 290
Effect of Distributed Capacitance
20
10
Impedance (dBohms)
-10
-20 no caps
300 mils
500 mils
-30 700 mils
1000 mils
-40
1.0E+07 1.0E+08 1.0E+09 1.0E+10
April 2010 Dr. BruceFrequency
Archambeault,
(Hz) IBM 293
Z11 Phase Comparison as Capacitor distance Varies for 35 mils FR4
ESL = 0.5nH
2
1.5
0.5
Phase (rad)
-0.5
100 mils
200 mils
-1 300 mils
400 mils
-1.5 1000 mils
2000 mils
-2
1.0E+06 1.0E+07 1.0E+08 1.0E+09
Frequency (Hz)
10
no caps
Impedance (dBohms)
300 mils
-10 500 mils
700 mils
1000 mils
-20
-30
-40
-50
1.0E+07 1.0E+08 1.0E+09 1.0E+10
April 2010 Dr. BruceFrequency
Archambeault,
(Hz) IBM 295
Cavity Resonance (EZ-PowerPlane)
Equivalent Circuit for HSPICE
Port i Lij Port j
Lii Ljj
Por
Nmni C0 Lmn Gmn Nmnj tn
10
0
Impedance (dBohms)
-10
-20
-50
1.0E+07 1.0E+08 1.0E+09 1.0E+10
April 2010 Dr. BruceFrequency
Archambeault,
(Hz) IBM 297
Current Source Pulse for Simulated IC Power/GND
750 ps Rise/Fall
1.2
0.8
Current (amps)
0.6
0.4
0.2
-0.2
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 298
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor (with 2 nH) at Various Distances (Fullwave Simulation)
2
1.5
0.5
-0.5
-1
-1.5
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 299
Time Domain Current through Capacitor From Simulated IC Power/GND (1 amp)
Single Capacitor (with 2nH) at Various Distances
50
-50
-100
Current (milliamps)
-150
-200
-250
750ps Rise, 35 mil planes,1uF @ 10mils
750ps Rise, 35 mil planes,1uF @ 400mils
-300 750ps Rise, 35 mil planes,1uF @ 800mils
750ps Rise, 35 mil planes,1uF @ 1200mils
750ps Rise, 35 mil planes,1uF @ 1600mils
-350
-400
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 300
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor (with No L) at Various Distances
2
750ps Rise, 35 mil planes,1uF @ 10mils
750ps Rise, 35 mil planes,1uF @ 400mils
1.5 750ps Rise, 35 mil planes,1uF @ 800mils
750ps Rise, 35 mil planes,1uF @ 1200mils
1 750ps Rise, 35 mil planes,1uF @ 1600mils
0.5
Level (volts)
-0.5
-1
-1.5
-2
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 301
Time Domain Current through Capacitor From Simulated IC Power/GND (1 amp)
Single Capacitor (with no L) at Various Distances
400
200
0
Current (milliamps)
-200
-400
-600
750ps Rise, 35 mil planes,1uF @ 10mils
750ps Rise, 35 mil planes,1uF @ 10mils
-800 750ps Rise, 35 mil planes,1uF @ 800mils
750ps Rise, 35 mil planes,1uF @ 1200mils
750ps Rise, 35 mil planes,1uF @ 1600mils
-1000
-1200
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 302
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor with Various Capacitor Connection Inductance
0.6
0.5
0.4
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 303
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor with Various Capacitor Connection Inductance
2
0.5
-0.5
-1
-1.5
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 304
Maximum Time Domain Noise Voltage Across Simulated IC Power/GND Pin
Single Capacitor at Various Distances (Fullwave Simulation)
0.8
0.7
0.6
0.5
Voltage (volts)
0.4
0.1
0
0 200 400 600 800 1000 1200 1400 1600 1800
Distance (mils)
April 2010 Dr. Bruce Archambeault, IBM 305
Maximum Time Domain Noise Voltage Across Simulated IC Power/GND Pin
Single Capacitor at Various Distances (Fullwave Simulation)
0.8
0.7
0.6
0.5
Voltage (volts)
0.4
0.1
0
0 200 400 600 800 1000 1200 1400 1600 1800
Distance (mils)
April 2010 Dr. Bruce Archambeault, IBM 306
Maximum Time Domain Noise Voltage Across Simulated IC Power/GND Pin
Single Capacitor at Various Distances (Fullwave Simulation)
1.8
1.6
1.4
1.2
Voltage (volts)
0.4
0.2
0
0 200 400 600 800 1000 1200 1400 1600 1800
Distance (mils)
April 2010 Dr. Bruce Archambeault, IBM 307
Maximum Time Domain Noise Voltage Across Simulated IC Power/GND Pin
Single Capacitor at Various Distances (Fullwave Simulation)
1.6
1.4
1.2
1
Voltage (volts)
0.4
0.2
0
0 200 400 600 800 1000 1200 1400 1600 1800
Distance (mils)
April 2010 Dr. Bruce Archambeault, IBM 308
Maximum Voltage vs Distance to Capacitor for 1 ns Rise/fall time
0.01 uF Capacitor with 0.5 nH ESL and 30 mOhm ESR
1.4
1.2
Maximum Voltage at source (volts)
1 35 mil FR4
10 mil FR4
2 mil FR4
0.8 1 mil FR4
0.5 mil FR4
0.6
0.4
0.2
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Distance From Capacitor (mils)
PWR
GND
PCB
PWR
GND
PCB
PCB PWR
GND
PCB PWR
GND
Cap
PCB PWR
GND
0.2
Level (volts)
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 316
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor (with 0.5 nH Connection L) with Various Capacitor Values
0.6
0.5
750ps Rise, 10 mil planes, (0.5nH) 1uF @ 400mils
0.2
Level (volts)
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 317
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor (with 1 nH Connection L) with Various Capacitor Values
0.6
0.2
Level (volts)
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 318
So Far
Frequency domain simulations not optimum
for charge delivery decoupling calculations
(phase not considered)
Time domain simulations using single pulse of
current indicate limited capacitor location
effect
Connection inductance of capacitor much higher
than inductance between planes
Charge delivered only from the planes
Vdc
a = 12
DC voltage used to
charge the power
plane
3.5
3
[V]
2.5
plane
Ls = 1nH
V
2
Ls = 10 nH
1.5 Ls = 50 nH
1
0 2 4 6 8 10
Time [ns]
April 2010 Dr. Bruce Archambeault, IBM 324
Charge Depletion vs. Capacitor Distance
Port1 (8,7)
(Ls 50nH)
C=1uF
ESL=0.5nH Decap
ESR=1 10
1 inch
Ground pin
Port2 (4,5)
(power pin)
r =4.5
12
April 2010 Dr. Bruce Archambeault, IBM 327
Effect of Multiple Capacitors While Keeping
Total Capacitance Constant
The decap locations are 800mils, 1200mils, 2700mils from the power pin
Port1 (8,7)
(Ls 50nH)
C=0.5uF
ESL=0.5nH Decap
ESR=1 10
1 inch
Ground pin
Port2 (4,5)
(power pin)
r =4.5
12
April 2010 Dr. Bruce Archambeault, IBM 328
Effect of Multiple Capacitors While Keeping
Total Capacitance Constant
The decap locations are 800mils, 1200mils, 2700mils from the power pin
Port1 (8,7)
(Ls 50nH)
C=0.25uF
ESL=0.5nH Decap
ESR=1 10
1 inch
Ground pin
Port2 (4,5)
(power pin) r =4.5
12
April 2010 Dr. Bruce Archambeault, IBM 329
Constant Capacitance
800 mil Distance
0.2
Level (volts)
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 333
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor (with 0.5 nH Connection L) with Various Capacitor Values
0.6
0.5
750ps Rise, 10 mil planes, (0.5nH) 1uF @ 400mils
0.2
Level (volts)
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 334
Time Domain Noise Voltage Across Simulated IC Power/GND Pin (1 amp)
Single Capacitor (with 1 nH Connection L) with Various Capacitor Values
0.6
0.2
Level (volts)
0.1
-0.1
-0.2
-0.3
-0.4
0 0.5 1 1.5 2 2.5 3 3.5 4
Time (ns)
April 2010 Dr. Bruce Archambeault, IBM 335
Noise Voltage is INDEPENDENT
of Amount of Capacitance!
As long as there is
Dist=400 mils enough charge
ESR=30mOhms
ESL=0.5nH
d +r
ln 2 1
d (R + r ) (d1 + r )
2 2
d +
ln R r
Courtesy of Jingook Kim,
4 r 3 (d 2 + r ) 4
d +r
ln 2
r
Jun Fan, Jim Drewniak
Missouri University of
=
d
ln
( R + r)
4
Science and Technology
4 (2 R sin( / 2) + r )r
3
750 mil
1400
1000 mil
1300
1200
1100
1000
900
800
700
600
500
0 50 100 150 200
Angle (degrees)
450
400
350
Inductnace (pH)
300
250
200
500mil
150 250 mil
750 mil
100 1000 mil
50
0
0 50 100 150 200
Angle (degrees)
350
300
500mil
250 mil
250
Inductnace (pH)
750 mil
1000 mil
200
150
100
50
0
0 50 100 150 200
Angle (degrees)
2
d1 = R2 + d2
1400
1300
1200
1100
1000
900
800
700
600
500
0 100 200 300 400 500 600 700 800 900 1000
Distance Between Capacitors (mils)
450
400
350
Inductnace (pH)
300
250
200
250 mil
150
500mil
0
0 100 200 300 400 500 600 700 800 900 1000
Angle (degrees)
350
300
250 mil
250 500mil
Inductnace (pH)
750 mil
1000 mil
200
150
100
50
0
0 100 200 300 400 500 600 700 800 900 1000
Angle (degrees)
10cm
10mm
20cm
10cm
2 via with degree 90 2 via with degree 180
20cm
[m] [m]
[m] [m]
A/m2 A/m2
[m] [m]
40 324
8
48 16
24 16
5468
8
64
564736422
6546
0.11
2
40
80
16 16
0.11
32
16
430
6754468
32
80
820
8
48
2
80 72 0.105
24
0.105 64 6840
56
24
64
6
4
0.1 7820 0
16
0.1 80 72 56448
16
72 50648
8
80
8
4 24 32
32 0.095 16
8
0.095 16 24 8
0.09 8 0.09 8
0.085 0.085
0.08 0.08
0.08 0.0850.09 0.095 0.1 0.1050.11 0.1150.12 0.08 0.085 0.09 0.095 0.1 0.105 0.11 0.115 0.12
0.12 0.12
0.115 0.115 8
8 21
56344206
8
24
0.11 0.11 56408
5766244
48
5664
4042
16
44302
6
16 1
48
7
16
8
8
0.105 0.105
8
241 8 48
68450468 56
3240
56
32
24
4782 60
40
546
0 0.1 6840
6
48
0.1
4032
782 6 7 564
40
24
324
7248 5 432
2
8
0 2 16 6472
808
4
8
16
0.095 24 0.095
16 8
0.09 0.09 6744208 2
55664 42304
8
8 16
48
0.085 0.085 8
0.08 0.08
0.08 0.0850.09 0.095 0.1 0.1050.11 0.1150.12
0.08 0.085 0.09 0.095 0.1 0.105 0.11 0.115 0.12
Case2 : 5 inches
Port1 Port2
Case3 : 2 inches
1 inch
Case1 : 10 inches
~ 330pH
Case2 : 2 inches
~ 250pH
~ 560pH
t1 t1 t
t2
0 T/2 T/2
C pd * m *Vcc
I p2 =
t 2
Cpd is specified for Clock drivers/buffers
m is number of I/O drivers
t2 = tr + tf
C L nVcc
IL =
t / 2
April 2010 Dr. Bruce Archambeault, IBM 357
More Accurate I/O Driver
Current
Use Signal Integrity tools to
find current waveform
Hyperlynx
Spectraquest
SPICE
-30
-40
-50
-60
-70
-80
-90
0.5 1 1.5 2 2.5 3
Frequency (GHz)
-40
-50
-60
-70
-80
-90
0.5 1 1.5 2 2.5 3
Frequency (GHz)
April 2010 Dr. Bruce Archambeault, IBM 362
CDC208 (8 loads)
-20
Theoretical results
-30
Measured results
Spectrum (dBm)
-40
-50
-60
-70
-80
-90
0.5 1 1.5 2 2.5 3
Frequency (GHz)
April 2010 Dr. Bruce Archambeault, IBM 363
Modeling results with different CPD
MPC905
-10 CPD=0 pF
-20 CPD=19.5 pF
CPD=100 pF
Spectrum (dBm)
-40
-50
-60
-70
-80
-90
0.5 1 1.5 2 2.5 3
Frequency (GHz)
-30
-40
-50
-60
-70
-80
0.5 1 1.5 2 2.5 3
Frequency (GHz)
April 2010 Dr. Bruce Archambeault, IBM 365
Modeling results for ICS9341 (load C only)
ICS9341
-30
Theoretical results
-40
Spectrum (dBm)
-60
-70
-80
-90
-100
-110
0.5 1 1.5 2 2.5 3
Frequency (GHz)
Note: The chip has a lot of different clocks: 8 CPU (133 MHz), 8 PCI (33.3
MHz), 2 USB (64 MHz and 32 MHz) and 2 REF (14.318 MHz).
April 2010 Dr. Bruce Archambeault, IBM 366
Other Sources on Active Board
Large ASICs do not specify Cpd
Measured power/ground plane noise for
large ASIC with and without decoupling
capacitors installed
Circuits operating with exerciser software
Examples for 1.5 volt and 2.5 volt supplies
1.9
1.8
1.7
Amplitude (volts)
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.0E+00 1.0E-07 2.0E-07 3.0E-07 4.0E-07 5.0E-07 6.0E-07 7.0E-07 8.0E-07 9.0E-07 1.0E-06
Time (seconds)
1.9
1.8
1.7
Amplitude (volts)
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.0E+00 1.0E-07 2.0E-07 3.0E-07 4.0E-07 5.0E-07 6.0E-07 7.0E-07 8.0E-07 9.0E-07 1.0E-06
Time (seconds)
60
50
Caps
Caps gone
Percentage %
40
30
20
10
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
Voltage
90
80
70
Amplitude (dBuv)
60
50
40
30
20
10
0
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09 2.0E+09
Frequency (Hz)
90
80
70
Amplitude (dBuv)
60
50
40
30
20
10
0
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09 2.0E+09
Frequency (Hz)
2.7
2.6
Amplitude (volts)
2.5
2.4
2.3
2.2
0.0E+00 1.0E-07 2.0E-07 3.0E-07 4.0E-07 5.0E-07 6.0E-07 7.0E-07 8.0E-07 9.0E-07 1.0E-06
Time (seconds)
2.7
2.6
Amplitude (volts)
2.5
2.4
2.3
2.2
0.0E+00 1.0E-07 2.0E-07 3.0E-07 4.0E-07 5.0E-07 6.0E-07 7.0E-07 8.0E-07 9.0E-07 1.0E-06
Time (seconds)
50
40 Caps
Caps gone
Percentage %
30
20
10
0
2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3
Voltage
90
80
70
Amplitude (dBuv)
60
50
40
30
20
10
0
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09 2.0E+09
Frequency (Hz)
90
80
70
Amplitude (dBuv)
60
50
40
30
20
10
0
0.0E+00 2.0E+08 4.0E+08 6.0E+08 8.0E+08 1.0E+09 1.2E+09 1.4E+09 1.6E+09 1.8E+09 2.0E+09
Frequency (Hz)
Area Equal
current
Current
for 1/3rd
Total Power
Slot interrupts
currents on inside of
enclosure
coating
Iinj
contact area
coating
Vgap
Shielded Box
Shielded Cable
Cable shield
EM TIGHT
April 2010 Dr. Bruce Archambeault, IBM 395
Antennas
All metal conductors are antennas!
Some are more efficient than others
especially at certain frequencies
The same antenna can radiate or receive
reciprocity works!
IL sin j 1 1
E = 2 + 2+ 3
4 0 c r cr j r
Monopole
h Antenna
h
Image
Metal Box
wire
EMC experts
Experience is important
Again, beware ---- ask questions and understand WHY
Cookbooks do not work! Every case is special
and different