Service Manuals, User Guides, Schematic Diagrams or docs for : Agilent 5991-2803EN Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties c20140721 [10]
<< Back | HomeMost service manuals and schematics are PDF files, so You will need Adobre Acrobat Reader to view : Acrobat Download Some of the files are DjVu format. Readers and resources available here : DjVu Resources
For the compressed files, most common are zip and rar. Please, extract files with Your favorite compression software ( WinZip, WinRAR ... ) before viewing. If a document has multiple parts, You should download all, before extracting.
Good luck. Repair on Your own risk. Make sure You know what You are doing.
Image preview - the first page of the document
![5991-2803EN Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties c20140721 [10]](/pics/5c4d6cb51d797.gif)
>> Download 5991-2803EN Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties c20140721 [10] documenatation <<
Text preview - extract from the document
Analysis of Test Coupon Structures
for the Extraction of High Frequency
PCB Material Properties
White Paper
Heidi Barnes and Robert Schaefer
Keysight Technologies
Santa Rosa, USA
Jose Moreira
Advantest Corporation
Boeblingen, Germany
02 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Abstract
The driving forces behind low cost RF/Microwave mponents and the ever increasing
data rates of the digital world have resulted in a proliferation of printed circuit board
(PCB) materials and manufacturing techniques to provide cost effective solutions for
modern day applications. This has resulted in the need for practical ways of identifying
the frequency dependent loss properties of the after fabrication or "as-fabricated" PCB
designs for accurate pre- and post- layout simulation. Variations in key parameters such
as dielectric constant, loss tangent, dielectric height, etched trace width, surface rough-
ness, glass weave, moisture content, etc. can easily reduce the effectiveness of simula-
tions to predict the final design performance. Few companies have the time, money, or
equipment to fully dissect a fabricated PCB and determine all of these key dimensions
and material loss properties for a design. There is a strong need for simpler techniques,
such as measurable test structures, that can enable extraction of material properties for
improved accuracy of simulations. Recent papers using two transmission line lengths do
demonstrate measurements of dielectric constant and complex loss tangent [1,2,3,4], but
this method makes it difficult to predict the as-fabricated PCB trace width and dielec-
tric height. The intent of this paper is to explore the addition of Beatty series resonant
impedance structures [5] to improve the accuracy of extracting the as-fabricated PCB
material properties for the purpose of constructing 3D-EM simulations.
Keywords--material properties, PCB test structures, high speed digital, 3D-EM
Introduction
The proliferation of epoxy based multi-layer laminate PCBs for high speed digital circuits
running at microwave frequencies puts new demands on the world of EM simulations.
Recent publications have clearly shown that assuming dielectric loss to be constant with
frequency may have worked well for narrow band microwave applications, but for wide
band digital simulations it produces non-causal results. Surface roughness which could
be ignored with high quality thin film circuits is sparking new debates [6,7] to understand
the physics and accurately predict the variations seen between standard copper foil and
low profile versions. One issue that has not been widely addressed is the as-fabricated
transmission line etched trace widths and the dielectric height which are critical for the
set-up of accurate multi-Gigahertz EM simulations.
PCB fabrication can be very repeatable for a specific PCB design and fabrication
process, but accurately predicting the nominal value for a given fabricator's process is
best done with the measurement of simple test structures. This is especially true when
designing to the limits of PCB "lines and spaces" with narrow transmission lines that are
sensitive to fabrication tolerances and require the fabricator to adjust the original CAD
data in order to meet impedance targets. This is becoming a common practice in the
PCB industry, and yet few EM simulations are done with the after fabrication as-fabri-
cated trace widths and dielectric height since it is far easier to obtain the pre-fabricator
CAD data and laminate manufacturer's generalized data sheets. It is proposed that by
using the 2-Line test structure method for material properties with the addition of a
simple impedance varying Beatty resonant test structure, one can determine the PCB
transmission line dimensions as well as benefit from additional insights into the electrical
performance of the as-fabricated PCB material.
The Beatty structure [5] utilizes a simple change in trace width to create a measurable
delta change in impedance. The measured impedance change for a given increase in
physical transmission line trace width helps to define unique dielectric height and trace
width dimensions for improved EM simulation accuracy of the as-fabricated structures.
This paper provides a description of the methodology and then demonstrates the pro-
cess utilizing measurements from fabricated test coupons.
03 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Description of The Methodology
The two key factors in determining the performance of a transmission line are the imped-
ance and the loss. The IPC recommended approximation for the impedance of an internal
stripline trace with grounds on either side help to provide insight into the controlling
features that are needed for an accurate EM simulation [8].
(1)
Where:
Z o , characteristic impedance (Ohm)
b, the dielectric height between reference planes (mil)
t, copper thickness of the PCB trace (mil)
w, trace width (mil)
er , dielectric constant
The approximation for loss when a sine wave travels down a stripline with 1oz copper for
the conductor is: [8]
(2)
(3)
(4)
Where:
Z0, characteristic impedance (Ohm)
f, is the sine wave frequency (GHz)
w, trace width (mil)
co , is the speed of light in vacuum
tan d, loss tangent
er , dielectric constant
The equations show how impedance relates to the physical dimensions of the dielec-
tric height between the two reference planes and the trace width for a given dielectric
constant.
04 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Two Transmission Line Segments
To calculate the loss tangent and the dielectric constant one can use the 2-Line method
that enables the removal of connector fixturing losses and impedance mismatches so
that the total attenuation for a given length L of transmission line can be measured. At
high frequencies impedance reflections from the fixturing can be significant and must be
removed in order to see the true loss of the transmission line [9].
Figure 1: 2-Line Test Coupon Structures for measuring the S-Parameters of the Length L of trans-
mission line with the connector fixturing removed.
Since the transmission line is typically low loss, so it is helpful to select L such that it
is always longer (higher loss) then the connecting fixture and thus less sensitive to the
fixture removal quality. There are numerous methods for removal of the connector fixture
effects including NIST multi-line cal [3], optimization to split the 2x fixture into two equal
S-Parameters, and mathematical closed form expressions like the Automatic Fixture
Removal (AFR) used by Keysight Technologies PLTS software [10]. The AFR is a fast tech-
nique for obtaining the S-Parameters for Fixture A and Fixture B which can then be de-
embedded from the longer line to obtain the S-Parameters for the length L transmission
line. The S-Parameter data can then be used to find the dielectric constant as a function
of frequency by using the definition of how fast a signal travels in a dielectric material for
a given length L.
(5)
(6)
(7)
Where the dissipation factor
By enforcing causality, the real and imaginary permittivity must satisfy the Hilbert
transform relationship between the two [1,2] which enables one to create a model that
only requires measurement at two frequencies f1 and f2 where f1 is the main frequency
of interest and f2 can be set arbitrarily high at 1 Terahertz. This wide bandwidth causal
model or the Djordjevic model then determines the behavior of the dielectric constant
versus frequency [11].
05 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
(8)
And since "to know the real part is to know the imaginary part" then the Kramers-Kronig
relationship [1] gives:
(9)
The measured S-Parameters for the Length L of transmission line are then used to deter-
mine er'( f ) and tan d.
Figure 2: Plotting the loss in dB shows linear loss vs. frequency. At higher frequencies it may be
difficult to separate out trace loss from the accuracy of the fixture removal. (11.25 cm of 17.8 mil
stripline with R4350/R4450B 36 mil dielectric height. Dielectric constant at 10 GHz=3.78, and
Tand= 0.006)
Figure 3: The measured vs. simulated for unwrapped phase shows how the correct selection of
dielectric constant insures accurate prediction of the phase or electrical delay. (11.25 cm of 17.8
mil stripline with R4350/R4450B 36 mil dielectric height. Dielectric constant at 10 GHz=3.78, and
Tand= 0.006)
This 2-Line method works well for determining the dielectric properties and loss models,
but this does not address the need for accurate as-fabricated dimensions for creating
accurate 3D-EM models.
06 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Beatty Resonant Standard
To find the remaining as-fabricated variables including dielectric height, etched trace
width, and copper thickness. The proposed Series Resonant Beatty structure [5] has the
advantage of allowing additional data which includes the delta change in impedance for
a given delta change in physical trace width along with a resonant ripple in the insertion
loss which is affected by dielectric constant and transmission line losses.
Figure 4: The series resonant Beatty structure provides a delta impedance change for additional
information in determining the as-fabricated dielectric height and trace width.
The Beatty standard is constructed by increasing the trace width of the transmission line
for a specific distance that results in reflection resonances and broad band ripple on the
insertion loss.
Simulation of the methodology
To test out the methodology and explore the sensitivities of the simulated Beatty stan-
dard to the dielectric height and the trace width it is useful to use a closed form model
such as equations 1, 3, and 4 or a 2D-Planar EM model. It will also be shown that this is
a quicker way of predicting the settings needed for the full 3D-EM simulation and avoids
trying to tune the material properties with the longer simulation times of the 3D-EM
simulator.
Figure 5: Frequency Domain Simulation for tuning the transmission line parameters. The use of
2D-Planar EM models have improved accuracy over closed form models for wide bandwidth simu-
lations and yet still provide fast optimization and tuning of the variables to match with measured
data.
The initial simulation starts with the vendor material data sheets and PCB CAD data to
see how far off this method is in predicting the as-fabricated measured performance.
The prediction is okay at a few GHz, but for applications going beyond 10 GHz it begins
to lose accuracy.
07 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Figure 6: The simulation using CAD design data and vendor supplied data sheets works fine up to
a few GHz, but beyond 10 GHz it is helpful to make use of test structures for measurement based
modeling.
Updating with the measured loss properties and then tuning to the measured results of
the Beatty test structure for both insertion loss and time domain TDR impedance results
in much better correlation between simulation and measurements.
Figure 7: The 2D-Planar model is optimized to match the measurements.
Optimizing the impedance to match the measured Beatty standard helps to understand
how the fabrication tolerances impact the PCB performance.
Figure 8: The 2D-Planar model is also tuned in the time domain to match with the measured Beatty
test structure impedances.
08 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Measurement vs. simulation
The same settings for the transmission line properties that were determined from the
tuning of the 2D-Planar EM model can now be entered into the 3D-EM simulation. These
settings immediately provide a very good approximation to the measured data.
Figure 9: Successfully simulating a wide bandwidth resonant Beatty standard with a 3D-EM simu-
lator using measurement based as-fabricated properties.
The 2D-planar optimized settings also provide a good match with measured impedance
data in the time domain for the 3D-EM simulator. It is useful to note that if the meshing
density is too low in an EM simulator it can produce inaccurate results like lower then
expected impedance.
Figure 10: The 3D-EM simulator using measurement based as fabricated properties also predicts
the correct impedances of the Beatty test structure.
09 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
Conclusion
The well-established 2-Transmission Line or multi-line method of determining dielectric
and conductor losses goes a long way towards improving the causal time domain simula-
tions that are needed in today's high speed multi-gigabit systems. However, full 3D-EM
simulations also require information on the physical dimensions of the as-fabricated PCB.
It has been shown that with the addition of a series resonant Beatty test standard one
can improve the accuracy of using a 3D-EM simulator to predict measured performance.
The stepped impedances and resulting frequency domain broadband resonances provide
additional data points for optimizing to the fabricated dimensions of the transmission line
structure. There is also the added benefit of having a single structure which can verify
the PCB fabrication process and run to run repeatability. Future plans are to explore a
double resonant structure stepping from 1x trace width, up to 3x width, back down to 2x
trace width, and then back to the reference 1x trace width.
Acknowledgment
Don Sionne, Sharon Trueman, and Bill Wong for their wealth of experience in PCB layout
and fabrication. Mike Resso for his enabling resources for measurement based modeling
successes.
References
[1] C. Morgan, "Solutions for Causal Modeling and a Technique for Measuring, Causal, Broadband
Dielectric Properties," DesignCon 2008.
[2] E. Bogatin, D. DeGroot, C. Warwick, S. Gupta, "Frequency Dependent Material Properties: So
What?" 7-TA1, DesignCon 2010.
[3] D. Degroot, P. Pupalaikis, B. Shumaker, "Total Loss: How to Qualify Circcuit Boards",
DesignCon 2011.
[4] A. Rajagopal, B. Achkir, M. Koledintseva, A. Koul, J. Drewniak, "Material Parameter Extraction
Using Time-Domain TRL (t-TRL) Measurements", 978-1-4244-4267, IEEE EMC 2009.
[5] R.W. Beatty, "2-Port 1/4L Waveguide Standard of Voltage Standing-Wave Ratio" Electronics
Letters 25th January 1973, Vol 9, No. 2.
[6] Y. Shlepnev, C. Nwachukwu, "Practical Methodology for Analyzing the Effect of Conductor
Roughness on Signal Losses and Dispersion in Interconnects", 14-WA2, Designcon 2012.
[7] A. Horn, J. Reynolds, J. Ratio, "Conductor Profile Effects on the Propagation Constant of
Microstrip Transmission Lines", IEEE IMS 2010.
[8] E. Bogatin, Signal and Power Integrity- Simplified, Prentice Hall, 2010.
[9] H. Barnes, J. Moreira, M. Resso and R. Schaefer, "Advances in ATE Fixture Performance and
Socket Characterization for Multi-Gigabit Applications", DesignCon 2012.
[10] R. Schaefer, "Comparison of Fixture Removal Techniques for Connector and Cable
Measurements" Signal Integrity Workshop, IMS 2010.
[11] A.R. Djordjevic, R.M. Biljic, V.D. Likar-Smiljanic, and T.K. Sarkar, "Wideband Frequency-
Domain Characterization of FR-4 and Time-Domain Causality," IEEE Transactions on
Electromagnetic Compatibility, vol. 43, no. 4, November 2001.
10 | Keysight | Analysis of Test Coupon Structures for the Extraction of High Frequency PCB Material Properties - White Paper
myKeysight For more information on Keysight
Technologies' products, applications or
www.keysight.com/find/mykeysight
services, please contact your local Keysight
A personalized view into the information most relevant to you.
office. The complete list is available at:
www.axiestandard.org www.keysight.com/find/contactus
AdvancedTCA◦ Jabse Service Manual Search 2024 ◦ Jabse Pravopis ◦ onTap.bg ◦ Other service manual resources online : Fixya ◦ eServiceinfo