Re: Sign of current

Kellee and all the rest -

I think the proposed method of current sign determination might run into
trouble in at least two cases. One is when there are exactly the same
number of rising and falling segments, which could happen quite easily,
and the other is the same problem one runs into testing for monotonicity
of the independent variable. That problem is that very small oscillations
occur on the "flat" part of a numerically generated curve, not so much
because they are real, but because of the nature of small numbers, finite
precision, and floating point round off. It is almost impossible to do
enough smoothing to get rid of this effect without destroying the shape of
the curve in more "normally behaving" regions. These oscillations would
very possibly lead to false conclusions. I think it better, read more
reliable, to test the slope of the straight line between the zero volt, or
therabouts, current and the rail voltage, or thereabouts, current. ( It
is likely that there will not be a point at EXACTLY 0 or rail voltage. )
And now that I think about it again, this ought to work for ECL also, with
the terms 0 and rail redefined as appropriate. That should tell at a
glance if we are looking at a pull up or pull down curve, and the sign of
the current can be adjusted accordingly.

______________________________________________ ____________________
| | Seeking to return to | / __________________O
| Bob Ward | the North Country, MN, | / /|________________
| bward@neosoft.com | WI, ND, SD, or MT. Also | / /_/________________O
| 713.568.4122 | considering Mtn. West. | /_____________________
|___________________|_________________________| |_____________________O
Received on Wed Mar 16 04:58:00 1994