HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
-
Upload
sirjole7584 -
Category
Documents
-
view
214 -
download
0
Transcript of HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
1/8
Evaluation of Oxide Reliability
Using V-Ramp / J-Ramp Test
Application Note 4156-8
HP 4155B/4156B
Semiconductor Parameter
Analyzer
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
2/8
Introduction
With the size of gate area and the
thickness of the oxide shrinking
along with device geometries, cre-ating a reliable thin oxide has be-
come an important issue. The
integrity of the thin oxide in a MOS
device is a dominant factor deter-
mining overall reliability of a
micro-circuit.
This application note describes V-
Ramp and J-Ramp tests using the
HP 4155B/4156B semiconductor
parameter analyzer, based on the
JEDEC Standard No. 35.
They can promptly give useful
feedback to the manufacturing
process about oxide reliability.
What are V-Ramp and
J-Ramp?
Both Voltage Ramp (V-Ramp) and
Current Ramp (J-Ramp) are one of
so called Wefer Level Reliablity
(WLR) tests. In V-Ramp test, volt-
age across an oxide is increased at
a constant linear rate until failure
occurs. On the other hand, J-Ramp
test increases current through an
oxide logarithmically in linear time.
They both record total charge
(Qbd) and breakdown voltage
(Vbd). These parameters are meas-
ured on a large number of test sam-
ples and usually plotted as
cumulative breakdown distribution
versus breakdown charge density
or breakdown electrical field on
probability chart. The manufactur-
ing process should be maintainedso that this distribution becomes
closer to the ideal shape. V-Ramp is
more sensitive at detecting break-
down at lower voltages (i.e. logic,
micrprocessors). J-Ramp which is
more sensitive at higher voltages
(i.e. EPROM, flash). V-Ramp test is
more commonly applied to test
structures which have relatively
large oxide areas. On the other
hand, J-Ramp test is more often ap-
plied to small scribe-line test
structures.
V-Ramp Procedure
First, the test structure is checked
for integrity by measuring the leak-
age current after applying a con-
stant voltage (Vuse).
Second, voltage is applied to the
test structure in steps until the ox-
ide ruptures, compliance occurs, or
voltage exceeds specified limit This
ramped test should be performed
in accumulation mode.
Finally, a post stress analysis is
performed to record the failure
2
Structure Breakdown?
T
F
F
T
I leak > 1A
Current Compliance
F
Force Vuse
Measure I
Force Vstress
Measure I
Qbd = Qbd + I*T
T Initial Failure
Stop Test
Determine Failure
Type
Post Stress Test
START
END
Initial Failure
Stop Test
Determine Failure
Type
Post Stress Test
Qbd = 50C/cm^2
or Compliance?
T
F
F
T
Vmeas < Vuse
Vmeas < .85*Vprev
F
Force IstressMeasure Vmeas
T
Force I0
Measure V
Qbd = Qbd + Istress*T
Istress = Istress * FEND
START
Figure 1. Flowchart of V-Ramp
Figure 2. Flowchart of J-Ramp
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
3/8
category. Figure 1 shows the flow
of the V-Ramp Test.
J-Ramp Procedure
First, an initial test is made to de-
termine the quality of the oxide ca-
pacitor by monitoring the voltage
after applying a current (typically
1A). If the voltage does not reach
the use condition voltage within
specified time, it fails the initial
procedure.
Then, current is applied to the test
structure in steps until oxide rup-
tures, or total charge exceeds thespecified value. The current is in-
creased by multiplying a certain
factor (F) by the previous stress
current. JEDEC No.35 specifies
that F should be less than square
root of 10. Breakdown is defined by
a 15% drop in voltage. This ramped
test should be performed in accu-
mulation mode.
Finally, post stress test is per-
formed to determine the failure
category. Figure 2 shows the flowof the J-Ramp Test.
Problems with V-Ramp and
J-Ramp
Since oxides can rupture suddenly,
changing the step, monitoring the
voltage, and judgement of the exit
condition must be done very
quickly to insure reliable measure-
ment results. Automation using a
computer is mandatory for both of
these tests. Several JEDEC require-
ments make creating the test pro-
gram difficult.
It is desirable to keep the step time
as uniform as possible to measure
Qbd precisely and consistently.
During the ramped test, maintain-
ing a constant step time may be dif-
ficult. Using the clock in the
external computer, with its resolu-
tion and accuracy limitations, you
cannot set small and consistent
step durations. Even if you use the
sweep function of the instrument
and set the delay time, any range
changing that occurs during sweepwill add unpredictable step varia-
tion as shown in Figure 3.
HP 4155B/4156B Solution
The HP 4155B/4156B has built-in
HP Instrument BASIC. The whole
V-Ramp or J-Ramp tests can be
automated using HP Instrument
BASIC which is suitable for instru-
ment control and scientific calcula-
tion. Thus, an external computer is
no longer necessary.
Initial stress measurement, ramped
test, and post stress test are done
by loading the pre-saved measure-
ment setup files. These setup files
can easily be created by fill-in-the-
blank manner from the front panel.
This eliminates one of the most dif-
ficult parts of a program, instru-
ment control.
Of course, each setting parameter
can be changed from the program
later. So, any setting parameter
which is determined by compli-
cated calculation or by measured
results can be updated later.
You can make use of the HP
4155B/4156B's analysis function to
extract the resultant parameters.
Extraction of Vbd and calculation
of Qbd is possible by setting the
USER FUNCTION and moving the
marker to the breakdown point.
Complicated programming to ana-
lyze the data is eliminated.
The requirement of maintaining a
precise and constant time step
throughout the ramp test, is accom-
plished by using a fixed range, and
by using the output trigger function
(see Figure 4). The output trigger
function gets its name from the
ability of the HP 4156B to synchro-
nize with external instruments.
3
Measure Time
Range change
Delay Time
Figure 3. Duration Varies Due to Range Change and Inconsistent Measurement
Output Trigger
Step Delay Time
Measure Time
Delay Time
Overhead
Time
Figure 4. Constant Duration by Using Output Trigger Function
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
4/8
Step delay time for external instru-
ments is required to provide consis-
tent settling time for reliable
measurements. You can make use
of this function to keep the step du-ration constant during the ramped
test. Precise Qbd values can be ob-
tained by using this function by set-
ting time steps to 100s resolution.
V-Ramp with the HP4155B
/ 4156B
V-Ramp test
The entire procedure described be-
low is fully automated by the HP
Instrument BASIC.
First, an initial test checks leakage
current when Vuse is applied to the
test structure. To make a single
spot test, the HP 4155B/4156B is
set to sampling mode and the num-
ber of steps is set to "1". The post
stress test, which is made at the
end of entire test, is identical to the
initial test. Both measurements are
performed by loading presaved
measurement setup files. The setup
file can easily be created or modi-
fied from the front panel.
Then the ramped test is performed
using the sweep measurement
function. Figure 5 and 6 show the
SWEEP SETUP and MEASURE
SETUP pages.
Constant duration through out the
ramped test is desirable. The fol-
lowing setup procedure maintains
a constant step time interval during
the ramped test.
1 Enable trigger output function
in OUTPUT SEQUENCE page
(Figure 7).
2 Set DELAY TIME in SWEEP
SETUP page.
3 Set STEP DELAY time in OUT-
PUT SEQUENCE page.
4 Set measurement range to
FIXED range in MEASURE
SETUP page.
4
ME A S U R E : S WE E P S E T U P 9 4 J U N 2 5 0 6 : 3 1 A M
V o l t a g e R a mp S w e e p M e a s u r e me n t
* V A R I A B L E V A R 1 V A R 2
U N I T S MU 1 : MP
N A ME V g
S WE E P M O D E S I N G L E
L I N/ L O G L I NE A R
S T A R T - 5 . 0 0 0 V
S T O P - 4 0 . 0 0 0 V
S T E P - 5 0 . 0 mV
N O O F S T E P 7 0 1
C O M P L I A N C E 5 0 . 0 0 m A
P O WE R C O M P O F F
* T I M I N G
H OL D T I ME 0 . 0 0 0 0 s
D E L A Y T I ME 4 0 . 0 ms * S WE E P S T OP A T C OMP L I A N C E S t a t u s
* C O N S T A N T
U N I T S MU 4 : MP V S U 1 V S U 2
N A ME V s u b V S U 1 V S U 2
MO D E V V V
S O U RC E 0 . 0 0 0 0 V 0 . 0 0 0 0 V 0 . 0 0 0 0 V - - - - - - - - -
C O MP L I A N C E 1 0 0 . 0 0 mA - - - - - - - - - - - - - - - - - - - - - - - - - - -
ME A S U R E : ME A S U R E S E T U P 9 4 J U N 3 0 0 9 : 1 0 A M
V o l t a g e R a mp S w e e p M e a s u r e me n t
* M E A S U R E M E N T R A N G E
U N I T N A ME R A N G E Z E R O C A N C E L OF F
S MU 1 : MP I g F I X E D 1 0 0 mA O F F [ 1 n A ]
S MU 4 : MP I s u b F I X E D 1 0 0 mA O F F [ 1 n A ]
V MU 1 V MU 1 A U T O - - - - - O F F
V MU 2 V MU 2 A U T O - - - - - O F F
( * : Ol d d a t a i s u s e d . )
* I N T E G T I M E
T I ME N P L C
S H O R T @ 6 4 0 u s 0 . 0 3 2
ME D 2 0 . 0 ms 1L ON G 3 2 0 . ms 1 6
* WA I T T I M E
0 * ( D E F A U L T WA I T T I ME )
ME A S U R E : O U T P U T S E Q U E N C E 9 4 J U N 3 0 0 9 : 1 3 A M
V o l t a g e R a mp S w e e p M e a s u r e m e n t
* O U T P U T S E Q U E N C E * T R I G G E R S E T U P
U NI T N A ME MO D E E N A B L E / D I S A B L E E N A B L E
1 S MU 4 : M P V s u b V F U N C T I O N T R I G O U T
2 S MU 1 : M P V g V S T E P D E L A Y 2 1 . 4 ms
3 S MU 2 : M P P O L A R I T Y P O S I T I V E
4 S MU 3 : M P
5 V S U 1 V S U 1 V
6 V S U 2 V S U 2 V
7 S MU 5 : M P
8 S MU 6 : M P
9 P G U 1
1 0 P G U 2
Figure 5. SWEEP SETUP Page
Figure 6. MEASURE SETUP Page
Figure 7. OUTPUT SEQUENCE Page
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
5/8
5 Set WAIT TIME to zero in
MEASURE SETUP page.
JEDEC No.35 specifies that the
ramp rate should be from 0.1 to 1.0MV / cmsec. If you define step volt-
age (Vstep), the duration of each
step is calculated as follows.
Duration=Vstep/(Ramp rate Tox)[sec]
Tox: Oxide thickness [cm]
The delay time and step delay time
should be chosen so that they sat-
isfy the following equations.
In case of V-Ramp:
Duration=1.2 ms + (DELAY TIME)
+(STEP DELAY - 100 s)
In case of J-Ramp:
Duration=1.5 ms + (DELAY TIME)
+ (STEP DELAY - 100 s)
where: 1.2 ms and 1.5 ms are meas-
urement overhead.
DELAY TIME should be set large
and according to the above formula
so that the measurement is per-
formed close to the end of eachstep. STEP DELAY time must be
larger than sum of setting INTEG
TIME and measurement overhead.
To keep the step duration constant,
don't use USER FUNCTIONs while
voltage is swept. Analysis with
USER FUNCTIONs can be done af-
ter the ramped test procedure is
completed.
Select SWEEP STOP AT COMPLI-
ANCE Status in the SWEEP SETUP
page so that sweep aborts immedi-ately if compliance occurs.
The calculation of Qbd in equation
below, can be performed by a HP
4156B USER FUNCTION. The set
up of USER FUNCTION should be
made after the ramped measure-
ment to keep the duration con-
stant. Time is defined as
multiplication of measurement data
(@INDEX) and step duration. When
the marker is moved to the break-
down point, Qbd appears on the
GRAPHIC page.
Qbd =i(t) dt
The search for breakdown voltage
is made in the HP Instrument BA-
SIC program. JEDEC defines
breakdwon voltage as the voltage
when current becomes ten times
larger than the expected current or
the slope (log Ig /Vg) changesby between 1.5 to 3. Expected cur-
rent is decided by either a pre-
measured curve or theoretical
curve. Theoretical current is called
the Fowler-Nordheim current,
which is given as follows.
J = A E2 e(-B / E)
where
A , B : Constants in terms of effec-
tive mass and barrier
height.
E : Electric field
Or, since current starts flowing
right after oxide rupture, it will be
OK to detect the breakdown volt-
age one step before the Ig compli-
ance as a simplified alternative.
Another V-ramp test for
studying ultra thin oxide
leakages
The main purpose of WLR V-ramp
tests is to monitor the qualitative
tendency of the process. Parame-
ters such as Qbd give prompt proc-
ess control feedback by quickly
stepping the stress voltage to
5
C H A N N E L S : U S E R F U N CT I ON D E F I N I T I ON 9 4 J U N 1 8 1 0 : 5 7 A M
V o l t a g e R a mp S w e e p M e a s u r e me n t
* U S E R F U N C T I O N
N A M E U N I T D E F I N I T I O N
T i me s e c @I N D E X * . 0 6 2 5
V b d V @MY 2
Q b d C I N T E G ( I g , T i me )
G R A P H / L I S T : G R A P H I C S ME D I U M 9 4 J U N2 2 0 3 : 4 8 A M
o
*
V o l t a g e R a m p S w e e p M e a s u r e m e n t
M A R K E R ( 2 8 . 0 6 2 5 0 0 0 s e c - 5 . 2 4 2 5 m A - 2 7 . 4 0 0 V )
V b d - 2 7 . 4 0 0 0 0 0 0 0 0 V
( A ) Q b d - 1 1 . 3 3 3 5 0 9 3 7 5 m C ( V )
- 1 . 0 0 - 4 0 . 0
I g V g
d e c a d e - - - - - -
/ d i v / d i v
- 1 . 0 0 p - 5 . 0 0
0 . 0 0 T i m e ( s e c ) 5 . 0 0 / d i v 4 2 . 0
Figure 8. USER FUNC SETUP Page
Figure 9. V-Ramp Example Measurement Result
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
6/8
breakdown, in uniform time
increments.
For thin oxides (less than 100 ang-
stroms thick), low voltage "directtunneling" leakage must be accu-
rately studied. The HP 4156B is ide-
ally suited for this purpose due to
its ability to measure currents
down to fA levels at high speeds.
Use auto ranging in the MEASURE
SETUP page to increase accuracy.
Add a hold time only if you sweep
from a non-zero start point. Add de-
lay time only for large voltage step
size or if probes are not fully
guarded. Leakage sweeps, such asshown in Figure 10 can be made in
less than one minute using the HP
4156B with a well guarded probe
station such as the Cascade Sum-
mit series prober.
J-Ramp with the HP
4155B/4156B
The entire procedure described be-
low is fully automated by the HP
Instrument BASIC.
The initial test is performed usingthe sampling mode function. Speci-
fied current is forced and voltage is
repeatedly measured with constant
time interval. The measurement
stops when the voltage reaches the
specified threshold value, in this
case 2V. Figure 11 and Figure 12
show CHANNEL DEFINITION and
SAMPLING SETUP page
respectively.
After the initial test, the ramped
test is performed using the sweepmeasurement function. Figure 13
and Figure 14 show SWEEP SETUP
and MEASURE SETUP page
respectively.
Also with the J-Ramp test, constant
step time duration throughout the
ramped test is desirable. The dura-
tion is fixed constant by setting the
output trigger function, delay time,
6
G R A P H / L I S T : G R A P H I C S S H O R T 9 4 J U N 3 0 0 8 : 4 5 A M
o
. 0 0 1 C M 2 C A P T O X = 1 6 0 A n g s t r o m s
M A R K E R ( - 2 2 . 7 5 0 V - 2 7 . 5 9 0 m A )
( A )
- 1 0 0 . m
I g
d e c a d e
/ d i v
- 1 0 f
- 5 . 0 0 V g ( V ) 2 . 0 0 / d i v - 2 5 . 0
C H A N N E L S : C H A N N E L D E F I N I T I O N 9 4 J U N 1 9 0 0 : 1 0 P M
* M E A S U R E ME N T M O D E
S A MP L I NG
* C H A N N E L S
ME A S U R E S T B Y S E R I E S
U N I T V N A ME I N A ME MO D E F C T N R E S I S T A N C E
S MU 1 : MP V g I g I C O N S T 0 o h m
S MU 2 : MP 0 o h m
S M U 3 : M P
S MU 4 : MP V s b I s b V C O N S T
S M U 5 : M P
S M U 6 : M P
V S U 1 V S U 1 - - - - - - - V C O N S T
V S U 2 V S U 2 - - - - - - - V C O N S T
V MU 1 - - - - - - - - - - - - - - - -
V MU 2 - - - - - - - - - - - - - - - -
P GU 1 - - - - - - -
P GU 2 - - - - - - -
G N D U - - - - - - - - - - -
ME A S U R E : S A MP L I N G S E T U P 9 4 J U N 1 9 0 0 : 1 2 P M
* S A MP L I N G P A R A ME T E R * S T O P C ON DI T I O N
MO D E L I N E A R E N A B L E / D I S A B L E E N A B L E
I N I T I A L I N T E R V A L 1 0 . 0 0 ms E N A B L E D E L A Y 0 . 0 0 0 0 0 0 0 s
N O . O F S A MP L E S 1 0 0 N A ME V g
T O T A L S A MP . T I ME 1 . 0 0 s T H R E S H O L D 2 . 0 0 0 0 0 0 0 V
E V E N T | V a l | > | T h |
H O L D T I M E 0 . 0 0 0 0 0 0 s E V E N T N O . 1
F I L T E R O N
* C O N S T A N T
U N I T S MU 1 : MP S MU 4 : MP V S U 1 V S U 2
N A ME I g V s b V S U 1 V S U 2
MO D E I V V V
S O U R C E 0 . 0 0 0 0 A 0 . 0 0 0 0 V 0 . 0 0 0 0 V 0 . 0 0 0 0 V
C O MP L I A N C E 1 0 0 . 0 0 V 1 0 0 . 0 0 mA - - - - - - - - - - - - - - - - - -
Figure 10. Low Level Ig-Vg Characteristics Using Auto Ranging
Figure 11. CHANNEL DEFINITION Page (Initial Test)
Figure 12. SAMPLING SETUP Page (Initial Test)
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
7/8
step delay time, and zero wait time
as explained in the V-Ramp test.
The current is increased as IFn (n
is integer). To increase the currentwhile satisfying this equation, use
the LOG sweep function of the HP
4155B/4156B. Three kinds of F val-
ues can be chosen, which are
10(1/10), 10(1/25), and 10(1/50). The stop
current is calculated by solving the
following equation.
where
MaxQ : Maximum charge density
given by user (50 C/cm2 in
JEDEC No.35)Area : Oxide area [cm2]
In case of the HP 4155B/4156B, the
sweep measurement won't stop
right after the breakdown occurs.
After the specified sweep is over, it
can quickly search for the break-
down point. Figure 15 is an exam-
ple of J-Ramp ramped test.
Conclusion
The HP 4155B/4156B provides you
with features that greatly simplify
V-Ramp / J-Ramp tests. Pre-saved
setup files and auto analysis drasti-
cally reduce the programming ef-
fort. The built in HP Instrument
BASIC controller eliminates the
need for an external computer. Pre-
cise Qbd value is acquired by main-
taining a constant step time
duration through out the ramped
test.
7
ME A S U R E : S WE E P S E T U P 9 4 J U N 1 8 1 1 : 0 8 A M
* V A R I A B L E V A R 1 V A R 2
U N I T S MU 1 : M P
N A ME I g
S WE E P M O D E S I N G L E
L I N / L O G L O G 1 0
S T A R T - 1 . 0 0 0 u A
S T O P - 5 0 . 0 0 mA
S T E P - - - - - - - - -
N O O F S T E P 4 7
C O MP L I A N C E 4 0 . 0 0 0 V
P O WE R C O M P O F F
* T I M I N G
H OL D T I ME 0 . 0 0 0 0 s
D E L A Y T I ME 2 5 . 0 ms * S WE E P S T O P A T C OMP L I A N C E S t a t u s
* C O N S T A N T
U N I T S MU 4 : M P V S U 1 V S U 2
N A ME V s b V S U 1 V S U 2
MO D E V V V
S O U RC E 0 . 0 0 0 0 V 0 . 0 0 0 0 V 0 . 0 0 0 0 V - - - - - - - - -
C O MP L I A N C E 1 0 0 . 0 0 m A - - - - - - - - - - - - - - - - - - - - - - - - - - -
ME A S U R E : ME A S U R E S E T U P 9 4 J U N 3 0 0 9 : 0 7 A M
* M E A S U R E M E N T R A N G E
U N I T N A ME R A N G E Z E R O C A N C E L OF F
S MU 1 : MP V g F I X E D 4 0 V O F F
S MU 4 : MP I s b F I X E D 1 0 0 mA O F F [ 1 n A ]
( * : O l d d a t a i s u s e d . )
* I N T E G T I M E
T I ME N P L C
S H O R T @ 6 4 0 u s 0 . 0 3 2
ME D 2 0 . 0 ms 1
L O N G 3 2 0 . ms 1 6
* WA I T T I M E
0 * ( D E F A U L T WA I T T I ME )
G R A P H / L I S T : G R A P H I C S S H O R T 9 4 J U L 0 5 0 8 : 5 1 A M
o
*
C U R S O R ( - 2 . 3 5 0 0 0 0 0 0 m s e c 1 6 . 6 6 6 6 8 8 6 6 7 G A - 1 0 0 . 8 7 0 8 3 1 7 5 V )
M A R K E R ( 2 . 1 5 0 0 0 0 0 0 s e c - 1 9 . 5 1 m A - 2 7 . 1 1 2 4 V )
Q b d - 4 . 1 6 6 7 9 4 7 5 0 0 m C
( A ) V b d - 2 7 . 1 1 2 4 0 0 0 0 0 V ( V )
- 1 . 0 0 - 4 1 . 0
I g V g
d e c a d e - - - - - -
/ d i v / d i v
- 1 0 0 . n 0 . 0 0
0 . 0 0 T i m e ( s e c ) 2 0 0 . m / d i v 2 . 3 5
Figure 13. SWEEP SETUP Page
Figure 14. MEASURE SETUP Page
Figure 15. J-Ramp Example Measurement Result
-
8/14/2019 HP AN4156 8_Evaluation of Oxide Reliability Using v Ramp J Ramp Tests
8/8
For more information about Hewlett-Packard test and measurement products,applications, services, and for a currentsales office listing, visit our web site:http://www.hp.com/go/tmdir
You can also contact one of the followingcenters and ask for a test andmeasurement sales representative.
United States:Hewlett-Packard CompanyTest and Measurement Call CenterP.O. Box 4026Englewood, CO 80155-4026(tel) 1 800 452 4844
Canada:Hewlett-Packard Canada Ltd.5150 Spectrum WayMississauga, OntarioL4W 5G1(tel) (905) 206 4725
Europe:Hewlett-Packard CompanyEuropean Marketing CentreP.O. Box 9991180 AZ AmstelveenThe Netherlands(tel) (31 20) 547 9900
Japan:Hewlett-Packard Japan Ltd.Measurement Assistance Center9-1, Takakura-Cho, Hachioji-Shi,Tokyo 192, Japan(tel) (81) 426 56 7832(fax) (81) 426 56 7840
Latin America:Hewlett-Packard CompanyLatin American Region Headquarters5200 Blue Lagoon Drive9th FloorMiami, Florida 33126U.S.A.(tel) (305) 267 4245/4220(fax) (305) 267 4288
Australia/New Zealand:Hewlett-Packard Australia Ltd.31-41 Joseph StreetBlackburn, Victoria 3130
Australia(tel) 1 800 629 485 (Australia)(tel) 0 800 738 378 (New Zealand)
(fax) (61 3) 9210 5489
Asia Pacific:Hewlett-Packard Asia Pacific Ltd17-21/F Shell Tower, Times Square,1 Matheson Street, Causeway Bay,Hong Kong(tel) (852) 2599 7777(fax) (852) 2506 9285
Hewlett-Packard Company 1998Data subject to changePrinted in Japan 1/19985963-1248E
8