more musings on EFI controllers

[tom cloud]

>At 11:00 AM 8/6/96 -0500, you wrote:
>>The following is an excerpt of a missive I just sent to Michael
>>Kasimirsky and is a continuation of my rambling thoughts on EFI
>>control.
>>
>>>
>>>   [ snip ]
>>>
>
>>TPS and couple it through a capacitor (differentiate it), process this
>
>
>Don't we INTEGRATE by coupling through a capacitor & DIFFERENTIATE by 
>coupling through an inductor?
>
>
>                                                                           Bill
>
>

No,

Ignoring inductors (for practical purposes, they are too large
and too expensive, so most circuits just use R-C combos)
a circuit with R connected between the source and load and the C
connected in shunt across the load is called either a low-pass
filter or an integrator -- depending on what you're trying
to do.

   ----- R -------
IN             |     OUT
               C                lo-pass / integrator
               |
   ---------------


A circuit with C connected between the source and the load and the R
shunted across the output is called either a high-pass filter or
a differentiator -- depending on what you're trying to do.

   ----- C -------
IN             |     OUT
               R                hi-pass / differentiator
               |
   ---------------


If the integrator were perfect, a square wave input would yield
a triangular wave output -- but it isn't and it doesn't.

If the differentiator were perfect, a square wave input would
yield an infinitely large pulse with an infinitely short width
at each transition, in the direction of the transition -- but it
isn't and it doesn't.

With a square wave input, the simple R-C integrator yields a sawtooth
output while the R-C diff. gives a pulse no greater than the applied
voltage and a trailing decay (time constant).

thanks,

Tom Cloud