How might knowing about the inputs and outputs and side effects or errors produced within a system or device help you fix the device?

Answer 1

The question seems to most clearly be rephrased as "How does a device's output depend on it's input, and how can I troubleshoot it?".

As an EE, I can best answer from a generalized electronic approach.

In general, electronic systems consist of an input, some "box", and an output.

For the output to be correct, the "box" must know what the error is between it's input, and it's expected output. Though there are several methods, the most basic one is termed "negative feedback". The concept is simple. An input is applied, the "box" does it's thing, and supplies the output. But how does the "box" know it's "right"?

In general, the box must react to it's input and quickly produce an output, but how?

Using negative feedback means using a basic addition system where that system compares the input to the "box" to it's output.

For a system to be stable (and the theory is beyond this question), this feedback consists of adding a negative version of the output back into, and added to the input. Any difference is the error and is used to modify the output.

If this is hard to visualize, think of it like this: A camera sees a dark area and decides to flash a bright light. But it does not know how the light will reflect. So before the camera takes the picture, it measures how much light comes back and subtracts that from what it originally flashed. Now the subtraction of what it put out vs. what it sees (negative feedback) shows what the "error" or correction is required. Knowing the difference allows the rest of the processing know how it needs to adjust so that what "what went in" = "what needs to come out".

Now, how to fix a device? Well, it might not be apparent, but this stabilizing system of negative feedback can make it very difficult to diagnose a device or system. Fundamentally, it gets in the way while you try to see what's wrong, It is there altering the behavior of the "box" as it is malfunctioning, trying to make things right.

So the basic answer is that, in order to fix a device, you have to first know what the input and output characteristics are, WITHOUT feedback. Then, to SEE that characteristic, the feedback somehow, must be removed, or accounted for so that it's effect can be eliminated from the "output = input x something" characteristic. This can be a difficult job within any system. I myself have struggled in attempts to separate "what's wrong" with the correcting effects of feedback trying in vain, to fix an erroneous output.

Knowing the fundamental device's input and output, plus the "correcting or side effects" of feedback, will help you fix the device.