Re: [SI-LIST] : Diode Termination of Transmission Lines

Larry Smith ([email protected])
Wed, 10 Jun 1998 12:49:31 -0700 (PDT)

Diode terminations can be a quick and dirty way to solve a problem
but they can make problems also. Fast diodes at the end of a transmission
line will clamp the line and prevent the voltage from shooting more than
a diode drop from either rail. But they essentially make a short circuit
to AC ground. This reflects a big current back down the line. The
switching event is not over until all currents have died out on the line.
The presence of clamp diodes at some position on the net can cause
large voltage swings on another part of the net, sometimes several
time-of-flights later, when the diode finally comes out of conduction.

Resistive terminations are more desirable because they absorb energy
rather than transfer it to a different part of the line. The energy
from the driver eventually has to be dissipated in a resistance somewhere.

regards,
Larry

> From: [email protected]
> X-Lotus-FromDomain: NATURAL MICROSYSTEMS
> To: [email protected]
> cc: [email protected]
> Date: Wed, 10 Jun 1998 09:06:17 -0400
> Subject: Re: [SI-LIST] : Diode Termination of Transmission Lines
> Mime-Version: 1.0
>
>
> My experience with Diode termination:
>
> ? Schottky Diode Termination
> In some cases it can be expedient to use Schottky diodes or
> fast-switching silicon diodes to terminate lines. The diode
> switching time must be at least four times as fast as the signal
> rise time. Where line impedances are not well defined, as in
> breadboards and backplanes, the use of diode termination is
> convenient and can save time.
> The Schottky diode's low forward voltage, V(subscript: f)
> (typically 0.3 to 0.45V), clamps the input signal to a V
> (subscript: f) below ground (lower diode) and V(subscript: cc)+V
> (subscript: f) (upper diode). This significantly reduces signal
> undershoot and overshoot. Some applications may not require both
> diodes.
> ? The advantages of diode termination are:
> 1. Impedance matched lines are not required
> 2. The diodes replace terminating resistors or RC termination
> 3. The diodes' clamping action reduces overshooting and undershooting
> 4. Although diodes cost more than resistors, the total cost of layout
> might be less because a precise, controlled transmission-line
> environment is not required
> 5. If ringing is discovered to be a problem during system debug, the
> diodes can be easily added.
>
> Wojciech
>
>
>
>
>
> [email protected] on 06/10/98 06:09:39 AM
>
> To: [email protected]
> cc: (bcc: Wojciech Babij)
> Subject: [SI-LIST] : Diode Termination of Transmission Lines
>
>
>
>
>
>
> Hi,
>
> Might anyone be able to point me in the direction of some PRACTICAL
> application information on using Diode Termination on Transmission
> Lines. In particular for PCB1/Ribbon Cable/PCB2 interconnection as
> well as one side of a PCB to the other side.
>
> Most of the information that I have seen so far doesn't do anything
> more than give a brief overview. I was hoping on getting information
> on the diodes used and any of the pitfalls.
>
> Thanks
>
> Peter Baxter
>
>
> **** To unsubscribe from si-list: send e-mail to
> [email protected]. In the BODY of message put: UNSUBSCRIBE
> si-list, for more more help, put HELP. si-list archives are accessible at
> http://www.qsl.net/wb6tpu/si-list ****
>
>
>
>
> **** To unsubscribe from si-list: send e-mail to [email protected].
In the BODY of message put: UNSUBSCRIBE si-list, for more more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****

**** To unsubscribe from si-list: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list, for more more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****