This is a fairly complex question and I don't have much time today, but I'll take a stab at a quick answer.
To get a patent, the minimum you should describe is a basic, functioning implementation of your invention. The invention is defined by the claims, so you should start there -- with a rough cut at what you want your claims to look like.
To get at that, think of the elements of something someone else would make/use/sell to infringe your patents. Think of the simplest expression of an infringing circuit and/or combination of circuits. If your expression describes any existing circuits, you have to be more specific in your expression such that it doesn't describe any pre-existing things.
Once you're there, you have to describe a fully functioning implementation of that. And, you have to comply with the US's "best mode" requirement -- that is, you have to describe any preferences you might have. For example, lets assume your circuit uses an amplifier of some sort and you have found that a particular amplifier works well with the other components. You have to describe that. If it's an off-the-shelf component, name it, check it off, and move on. If it's an amplifier that you've designed for the specific purpose of working in this combination, you have to describe that design.
Now, if the amplifier is novel enough for its own patent application, you can file that separately from (but no later than) the other application for the combo and incorporate the amplifier application into the combo application by reference. So there are ways to minimize the redundancy of multiple applications.
If you have ideas that are improvements on the simplest implementation of your idea and that aren't really part of that implementation (as best mode), you can leave those out and evaluate them separately for possible patent applications -- i.e., do an independent cost/benefit analysis for each of those improvements.
I hope that helps.
Regards.