With our components pretty much resolved, it's time to answer the question of "how does everything go together?" Well, the step in that is to get a map of the nets on the board, and start assigning them names. Ideally all this info would be captured in schematic form, using a program like KiCAD's EESchema or (🤮) AutoCAD Electrical, but without a solid understanding of what we're looking at here, it's tough to know where to begin. 

The Process:

Borrowing a trick from my favorite random electronics disrepairman BigClive, I took high res photos of the front and back of the boards, isolated them onto a white background, and mirrored the backside. There aren't any designators there anyways, and this gives us a picture that's far easier to mentally process, similar to looking through all layers of a PCB in KiCAD's or Altium's layout modes. Since we have no top traces, I only needed to map out traces on one side, so I picked up my bucket and line tools and began filling in all attached nodes to individual nets.

Net Naming:

Unfortunately, we have a couple problems. The board designers weren't kind enough to give us a list of signals in the connector, a pin 1 marking, or even a ground plane. We'll have to make some assumptions:

1. Pin 1 on the connector is marked by the green line on the top layer
    - We can't possibly know this, but I need some sort of reference

2. There are/may be multiple power inputs and returns. We can't trust that only one ground exists

3. This board may be multi channel
     - While we don't know the exact function, we can't assume uniqueness or a lack thereof. All nets should be treated independently.

4. The board may be AC, DC, or a mix of both
     - I'll use the phrase 'power' and 'ground/return' a lot, but there's no actual reason to believe this is DC at this point. I just do it out of habit/convenience.

Not very helpful to our investigation, I know. But we have to start somewhere, so we'll name the gold fingers on the edge connector Pins 1 thru 21, name their nets accordingly, and index the remaining nets from 22 onward.

The Result:

After a lot of tedious drawing and labelling, the shape of the board starts to emerge:

It appears we're looking at 36 unique copper nets, with very few similarities or repeats in shape and design. Hard to deduce anything immediately, but this will give us a basis for a better diagram in KiCAD, and then a SPICE simulation. I have a few observations of my own:

- NET 21 connects to a large number of components and could be a power rail of some kind.
- NET15 thru NET21 are all similarly fanned out, and could be part of a common power or ground bus.
- NET2 connects to both sides of a trimpot, as does NET23 and NET 24. They all probably have the same function

The next step in the project is to turn this jumbled mess into a proper (or proper-ish) schematic. For that, it'd be a massive help to have a view where we can see all nets and components at the same time. Fortunately, with the magic of transparency...

With that, it's time to get stuck into some schematics.