[SI-LIST] : diode terminations(?)

Fred Rosenberger (fred@gpc.ibc.wustl.edu)
Tue, 15 Sep 1998 16:30:22 -0500 (CDT)

There was some discussion a few weeks ago about diode terminations.
I want to point out something thats frequently overlooked.

What has been said is true for zero-time transitions. Consider more
realistic case of a ramp transition. Now the diode termination is
only perfect at the end of the input transition. During the first
half of the input transition (from 0 to 50% of incident transition, this
is doubled by the open circuit impedance at the receiving end and
produces a 0 to 100% signal at receiving end) the diode (termination)
is open-circuit. During the last half (50% to 100%) it is
short-circuit. The result is a half-amplitude triangle-shaped pulse
that is reflected back to sending end. The initial return to receiving end
is in phase with the signal, but the following one is out of phase and
will subtract from signal. Not to mention what it may do to logic at
intermediate points on the line. Also, if the signal changes soon
enough, the first re-reflection to reach receiving end (3T with T
as one-way delay) will be out of phase (with the new receiving end
voltage). Loss is required to attenuate these reflections and make
diode terminations work.

Also, it cannot be over-emphasized that package or diode lead
inductance is critical and seldom considered.

Regards,

Fred

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****