Gert Hohenwarter
GateWave Northern
Effect of ground return path on
timing accuracy
Company
Logo
June 8-11, 2008
San Diego, CA USA
Objective
• Review problem of ground path
• Identification of trouble spots
• Assessment of individual impact
• Demonstrate significance
• Summary
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
22
Approach
• Examine complete signal path in a probecard
• Develop models (FEA, SPICE)
• Use models to evaluate impact of changes
• Provide select measurement results to
corroborate models
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
33
Problem
•
Signal in transmission line needs a return path
•
This path must be located physically close to the signal
•
Disruptions/detours add inductance and therefore delay
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
44
Probe card signal path
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
55
Probe card signal path
• The signal path in the probecard begins at the
perimeter with a tester connection and ends at
the probe tip with a connection to the DUT
Alongside the signal path a ground return must be provided
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
66
Follow the signal…..
• Timing verification begins at the tester
connection with a time domain reflectometer
and a test probe…..
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
77
…through the probe…..
• Common probes often have a fixed signal
contact and movable connection at ground:
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
88
Consequences
• Ground path inductance changes
• Signal path capacitance changes
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
99
Consequences
•
•
Ground path inductance changes
Signal path capacitance changes
• Delay changes
FEA analysis
SPICE model
dT [ps]
dL probe [nH]
2
dT [ps]
1
0
‐1
‐2
‐3
60
50
40
30
20
10
0
0
‐50
0
50
4
6
Lg [nH]
Ang le
June
June 88 to
to 11,
11, 2008
2008
2
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
10
10
Verification
• Measurement at actual probecard PCB
• At first glance everything is fine….
Multiple TDR
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
1.000
straight
long inline
short inline
ortho 30
gnd2
gnd3
0.800
rho
rho
Multiple TDR
0.600
0.400
0.200
gnd4
0.000
0
1
2
3
t [ns]
June
June 88 to
to 11,
11, 2008
2008
4
3
3.5
t [ns]
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
11
11
Delay extraction
• When processing the data, differences become noticeable
Delay variations
Delay variations
2.0
8.0
0.0
6.0
-2.0
4.0
-4.0
dt [ps]
10.0
0.0
-6.0
-8.0
-10.0
-2.0
-12.0
-4.0
-14.0
meas 1
meas 2
meas 3
meas 4
meas 5
Measurement to measurement
(probe straight)
fa
vo
ra
bl
e
gn
d4
gn
d3
gn
d2
30
or
th
o
Condition
le
as
t
-10.0
nl
in
e
-8.0
sh
or
ti
in
lin
e
-6.0
lo
ng
dt [ps]
2.0
For different probe positions
dT counter: ->
dT =
dT tot=
5
5
ps
ps
(average, not max. dT)
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
12
12
Follow the signal…..
Ground via selection (only one ground via present for each case):
dL for 0.5 mm pitch
0.4
dL [nH]
•
0.3
0.2
0.1
0
near
side
side
far
Inductance difference for 0.5 mm pitch
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
3
8
ps
ps
13
13
Follow the signal…..
• The signal travels into the PCB through a
connector and returns through an array of vias:
Disorganized
Organized
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
10 ps
18 ps
14
14
Follow the signal…..
PCB via position (vias must be provided near level shifts):
dT [ps]
dT [ps]
•
12
10
8
6
4
2
0
0
50
100
dy [mils]
Inductance change as a function
of lateral via position
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
5 ps
23 ps
15
15
Follow the signal…..
• Ground planes are disrupted in LGA/interposer area
Inductance per unit length changes
If traces get close to via antipads,
capacitance per unit length changes,
too
Example (inductance only):
Uniform array of 1mm pitch vias
Ground path inductance changes by
12% for a 40 mm long path
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
16 ps
39 ps
16
16
Follow the signal…..
• Ground via positions in the interposer contact array
Inductance [nH]
5
4
0.5
1
3
2
1.5
2
1
mm
0
1
2
# g nds
3
4
Inductance as a function of via distance from signal and number of
ground vias
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
10 ps
49 ps
17
17
Follow the signal…..
Delay change as a result of branch imbalance in ground path
150
dT [ps]
100
dT top 10k
50
dT bot 10k
dT top 50
0
-50
dT bot 50
-100
-150
0
0.5
1
Lg [nH]
1.5
2
Different load conditions result in
different delay changes
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
30 ps
79 ps
18
18
….to the DUT
Delay change as a result of grounding configuration at the DUT
Dense ground
With typical dimensions assumed the
inductance difference for the two
different configurations is 0.88 nH
Sparse ground
dT =
dT tot=
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
5 ps
84 ps
19
19
Overall changes
Graphic representation of relative change as a function of
where change occurs
Delay as a function of location
Comment:
os
er
rp
vi
os
as
er
gp
Sp
ho
ac
le
e
s
TR
-S
D
pr
R
ob
es
/D
U
T
Line length matching to
within 20 mils results in
an inaccuracy of 3.5 ps
vi
as
s
rp
in
te
in
te
tra
ce
vi
a
TR
po
s
vi
a
ea
s
Sum
Individual
m
ps
90
80
70
60
50
40
30
20
10
0
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
20
20
The impact of location
Delay change as a result of inductance change at
different locations along the signal path
Approximate equivalent circuit of signal path
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
21
21
The impact of location
Delay change as a result of inductance change at
different locations along the signal path
dT at different positions (50%)
50
1n
2n
3n
dt [ps]
40
30
20
10
0
1
Signal at DUT (SPICE simulation)
June
June 88 to
to 11,
11, 2008
2008
2
Pos # 3
4
Delay change
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
22
22
Waveform change
Wave form changes as a result of inductance change
dT at different positions (85%)
50
dt [ps]
0
-50
1n
2n
3n
-100
-150
1
Signal at DUT (SPICE simulation)
June
June 88 to
to 11,
11, 2008
2008
2
Pos #
3
4
Delay change
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
23
23
Different loads
Delay change for 3 different cases (10kOhm 2pF, 150 Ohm 2pF, 50 Ohm)
dT all loads (50%)
50.0
dt [ps]
40.0
30.0
20.0
10.0
0.0
1
2
Pos # 3
4
Load has additional impact on skew
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
24
24
Shared resources
Delay change as a result of inductance change in
case of split lines/ shared drivers
Simplified equivalent circuit
June
June 88 to
to 11,
11, 2008
2008
Signal change
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
25
25
Shared resources
Delay change as a result of inductance change in
case of split lines/ shared drivers
Delay change at 85% level
Delay change at 85% level
(10kOhm/2pF load)
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
26
26
Summary
•
•
•
•
•
•
Overview to problem of ground path provided
Timing accuracy is sensitive to location of ground discontinuity
Timing accuracy depends on architecture of signal delivery system
Budgeting should be performed
Compensation seems possible, if trends are known
Other contributions to timing arise from line length matching, fabrication
variances such as dielectric constant changes, coupled sections,
transitions and other discontinuities on the signal itself and must be
accounted for separately. These contributions have NOT been taken
into account here, yet a significant skew has already accumulated …..
Thank you!
June
June 88 to
to 11,
11, 2008
2008
IEEE
IEEE SW
SW Test
Test Workshop
Workshop
27
27