<Text_1>
     For those of you who may be interested this is an extract of some recent
     dialog that Joe Fjelstad and I had on flex endurance testing.


     Andy Magee - Applications Engineer
     Rogers Corp - Circuit materials Unit
     Tel: (602) 917-5237
     Fax: (602) 917-5256
     E-Mail: [log in to unmask]


     ______________________________ Reply Separator

     Joe,

     I just read your excellent column in the Feb '96 Circuitree.

     We routinely operate the rolling flex test at 1000-1100 rpm to qualify
     various material constructions for use in dynamic applications. This
     high speed extension of the test you described was developed at Rogers
     many years ago to satisfy an IBM disk drive qualification requirement.

     We've always referred to the dynamic folding test you described as the
     MIT test (JIS P 8115). Recently it's been adopted as the test method
     most representative of notebook computer hinge circuit motion.

     We're also aware of a new type of flex test for high speed dynamic
     applications, developed by ADFlex Solutions Ltd. (formerly Xyratex).
     The circuit is held in a loop similar to the rolling flex test, but
     rather than rolling it back and forth the gap is varied dynamically.
     Essentially the ends are fixed to parallel plates that move in and out
     relative to each other. This is intended to best represent the kind of
     varying radius motion that modern disk drive circuits experience. To
     date, the results seem to be comparable to the rolling flex test.

                      _______________________ Reciprocating Drive Arm
                      |          |          |------------------------
                |11111|2222|33333|4444|55555|
     Fixed Base |__________|__________|

     This ASCII schematic represents an end view of five circuits in test.

     Viewed from the top the circuit forms a U captured between parallel
     plates. The sides of the U are moving in and out relative to each
     other at up to 1500 rpm.

     |*            *|*                       *|
     |*            *|*                       *|
     |*            *|*                       *|
     |*            *|*                       *|
     |*            *| *                     * |
     |*            *|  *                   *  |
     |*            *|    *               *    |
     |*            *|       *         *       |
     |*            *|            *            |
     | *          * |                         |
     |   *     *    |                         |
     |      *       |                         |

     |-----min.-----|-----------max.----------|



     Andy Magee - Applications Engineer
     Rogers Corp - Circuit materials Unit
     Tel: (602) 917-5237
     Fax: (602) 917-5256
     E-Mail: [log in to unmask]


     _________________________________ Subject: Re: Flex Endurance Testing
     Author:  Josephfjel::(JSPHFJ) at ~FABRIK


     From: [log in to unmask]
     Date: Thu, Feb 22, 1996 6:09 AM
     Subject: Re: Flex Endurance Testing
     To: Magee, Andrew P
     Hi Andy!

     Thank you very much for your kind comments! And thank you for the new
     information as well. There is still much for me to learn. I hadn't heard of

     this new test developed by ADFlex (Xyratex) you described. It is
     interesting that the results of such testing parallels the rolling test. It

     would seem that the rolling test might distribute the stresses over a wider

     area more uniformly working the metal foil (and "searching out" flaws over

     a wider area as well) whereas the compression test you describe might work

     the center of the bend radius a bit more. This simply a visceral feeling. I

     would need to study it a bit more. Clearly it appears more efficient
     allowing multiple circuits to be tested at once. What number of cycles is
     typical for each test?

     Very Best Regards!
     Joe Fjelstad


     _________________________________ Subject: Re: Flex Endurance Testing

     Joe,

     You're right about the rolling test evaluating a greater effective foil
     area in each cycle. By implication results by the ASL changing radius test

     are a subset of the rolling test population. The strain applied in each
     cycle is equivalent for any specific radius considered. The correlation of

     these two techniques indicates that there is very little effect from
     localizing the strain.

     ASL studied the changing radius of the circuit by high speed video to
     understand if it was generating any non-uniform motions. Our concern was
     that during the course of the test the circuit might take on a permanent
     set forming a small radius at the tip of the bend. Fortunately, other than

     a tendency for the tip to shift from the center to one side during each
     cycle (whip) the change in radius is quite uniform. The tip whip is caused

     by one leg of the U providing more of the deflection angle than the other.


     By the way our current rolling test equipment can test up to 40 samples at

     one time. The number of cycles to failure for these tests is very test
     configuration and construction dependent, but we've seen anything from 10E4

     to >10E8.

     We've just completed work on a new Technical Bulletin "Maximizing Flex Life

     in Flexible Printed Circuits". Please meet me at our IPC Expo Booth #1520
     (3/5 or 3/6) and I'll give you a preprint copy.




     Andy Magee - Applications Engineer
     Rogers Corp - Circuit Materials Unit
     Tel: (602) 917-5237
     Fax: (602) 917-5256
     E-Mail: [log in to unmask]