Doug's picture
High Frequency Measurements Web Page
Douglas C. Smith

 Address:  P. O. Box 1457, Los Gatos, CA 95031
 TEL:      800-323-3956/408-356-4186
 FAX:      408-358-3799
 Mobile:   408-858-4528
 URL:      www.dsmith.org
 Email:    doug@dsmith.org



Technical Tidbit - February 2004
Measuring Printed Circuit Board Characteristic Impedance without a TDR
(when you need a close answer fast!)

by
Charles Grasso
Senior Compliance Engineer
Echostar Communications Corp.
Tel: 303-706-5467
Fax: 303-799-6222
Cell: 303-204-2974
Email: charles.grasso@echostar.com
Email Alternate: chasgrasso@ieee.org

Abstract: In a crunch, being able to adapt test equipment to get a close answer fast can be a life saver. The characteristic impedance of  a prototype circuit board was suspected as the cause of a 64 bit DDR SDRAM  bus failing initialization tests and a method to confirm the board design was needed urgently. This Technical Tidbit describes the method by which a network analyzer was pressed into service and its use outlined. This technique has been successfully used for analyzing other assemblies such as EIDE cables and USB unshielded twisted pairs.

Discussion: The relationship between inductance and capacitance for any transmission line is defined as (L/C)1/2. What was needed then was a way to measure the electrical components of the board fabrication and estimate the characteristic impedance by using the transmission line formula. Typically when boards that have impedance sensitive designs are manufactured, coupons are provided with test data that describes the board electrical details. Test coupons are small PCBs with exactly the same layer and trace construction as the main PCB. For example, a coupon will include traces of the same line width and copper weight on the same layer as the controlled impedances on the main PCB thereby providing an easy way to confirm the board design. The coupon is manufactured in the same process as the main board and with the same aperture code as used for the critical traces. An example of a coupon is shown is Figure 1.

 

  Printed Circuit Board Coupon
 Figure 1. Printed Circuit Board Coupon
 
Test Method. The test coupon is a very simple pwb construction with traces that reflect the board stack-up and trace construction design. Test coupons will also have, typically, a set of vias connected to ground and the signal, in close proximity at one end of the trace, to allow a test probe to be attached. The test coupon shown in Figure 1 was ideal for testing in the impedance analyzer as not only are the test points are close to each other but there is a set at each end of the trace making the modification for the inductance measurement very simple indeed. The coupon is prepared by attaching wires to one end of the set of test points as shown in Figure 2 and the wires are used to attach the coupon to the analyzer adapter as shown in Figure 3. Two measurements are now taken with the test coupon in two different configurations: with the second set of vias open (used in the capacitance measurements) and shorted ( used for the inductance measurement) as shown in Figures 4. After getting the capacitance and inductance numbers, use the transmission line formula to estimate the characteristic impedance of the trace.
 
 

Coupon with wires attached

Figure 2. Coupon with wires attached.
[note the desired impedance!]

Coupon under test

Figure 3. Coupon under test

Coupon showing shorted signal

Figure 4. Coupon showing shorted signal.

(click on pictures for a larger image)

 
 
Test Results: Network Analyzer Inductance data (click here), Network Analyzer Capacitance data (click here)
 
From the plots: L =  37.5954 nH and C = 39.1054 pF
        Zo (estimated) = 31 Ohms
 
 
TDR: Impedance data (click here)
          Zo (measured) = 29 Ohms
 
 
Conclusion: When you need a close answer fast, a network analyzer can be pressed into service in place of a TDR. Typically the error for measuring the characteristic impedance of a circuit board  is about 5% which can really add up so if you need a more accurate measurement then a TDR is the best instrument to use.
 
Editor's note: Be sure to select a test frequency high enough to easily measure the L and C values, but low enough to avoid standing waves on the coupon's signal traces. For this article, Charles picked 3 MHz, a suitable frequency for this case. If you are working with circuit boards 2 cm in length, a higher frequency would be in order and control of the connection parasitics becomes more important. Thanks goes to Charles for contributing this article.

Top of page
Home


Questions or suggestions? Contact me at doug@dsmith.org
Copyright © 2004 Douglas C. Smith