Hello Greg, IBIS list:
Ummm... this is a place where a picture is worth a thousand
words, so I will try a little ANSI art.
Ground Clamp Data = *
Pulldown Data = +
(I)
|
|
| ++++++++++++++++
| +
| +
|+
-------********************** (V)
* +|
* + |
* + |
* + |
point A --> * |
* |
|
The picture above is an attempt to show a pulldown curve and ground
clamp curve on the same set of axis. What I've seen many time is that
at point A, the pulldown curve and the ground clamp become very close to
one another, even converging so that they lie on top of each other.
Physically, I'm not exactly sure what is happening, but apparently
when the pulldown FET starts acting as the clamp diode is stops acting
like a FET. Note that I've seen this convergence (or near to it) even in
devices that have ESD protection diodes. Apparently the current in the ESD
structure is small compared with the main clamping action of the
pulldown FET. At any rate, when you subtract the two curves you end up
with the curve shown in the cookbook. Hope this helps.
Regards,
Stephen Peters
Intel Corp.
> We got into a discussion the other day about the origin of the
> non-monotonic IV curve depicted on p. 20 of the cookbook. In my mind, the
> "output enabled" IV curve should be a sum of the clamp (ESD diode and/or
> parasitic well diode) and the FET characteristics. Therefore, when you
> subtract the clamp curve from it, you should get the FET curve, with the
> possible exception of some round-off errors. However, the "hump" you
> really see in the pull-up or pull-down curve is much bigger than round-off
> error. Is this easy to explain, or is it fairly subtle? Where did my
> reasoning go wrong?
>
> Greg Edlund
> Advisory Engineer, Critical Net Analysis
> IBM
> 3650 Hwy. 52 N, Dept. HDC
> Rochester, MN 55901
> gedlund@us.ibm.com
>
Received on Thu Dec 10 09:37:17 1998
This archive was generated by hypermail 2.1.8 : Fri Jun 03 2011 - 09:52:29 PDT