by Andrew Davie » Sun May 28, 2017 8:59 pm
Could use paint, but I would like to see if I can learn how to use Eagle.
Some notes:
Eagle manages a "schematic" and a "board". These are inextricably linked together. When you change one, the other changes.
The schematic is a conceptual view of the circuit; the board represents the physical implementation.
The approach seems to be to do the schematic first. The UI is not exactly intuitive, particularly on OSX.
Basically the first thing to do is add components. There are a gazillion different components available in the inbuilt library, and you can also add new libraries. You can search for your component name - I'm at a bit of a disadvantage because I don't know the generic names for the components (e.g, a 2-pin connector into which I screw incoming wires is a .... ?). So you need a bit of experience here.
First, then, lay down the components.
The next thing is to use the "NET" tool (NOT the line tool!) to connect components together. An important concept is that you design sub-parts of your circuit as stand-alone - the schematic is NOT the place to design/connect the whole circuit, traces and all. What you are doing is a conceptual design.
If you have a part with pins, you CANNOT or at least should not connect another part to it. You put the two parts down, and then you connect them with a NET connection. This is the correct way to do it. At the end of a NET connection, if it is not connected to anything, you can add a label/tag to it. This is essentially a short descriptor of where it connects. Any/all NET connections with the same label are connected together when the board view is engaged. So, for example, I can have a dozen GND lables/tags all over the schematic. These are "connected" by virtue of using the same label.
Similarly, if I have something connected to pin 6 of the Arduino, I can put a NET connection onto pin 6, and then tag it as "SOUND" for example. Then in the circuit subpart that handles the sound, somewhere else... I can put a NET connection to a resistor/capacitor subsection, and label the NET connection SOUND. These are implicitly connected on the board, even if not connected in the schematic.
One gotcha - you right-click (or double-finger-tap on OSX) for bringing up properties. A tag, though it looks like you can click on it -- CANNOT be clicked on. Tags are nothings - it's the NET connection that's actually the clickable object.
Once you have all your components in your schematic and you've defined through net connections and/or tags where everything logically hooks up (and note, you pay NO ATTENTION to physical location in the schematic) then you can go to the board view.
The board view lets you shift components around. It shows the components connected with lines, as defined in the connections (the NET stuff) in the schematic. You can shift, rotate, etc. When you're basically happy with component location, then you can do routing of the lines. You can do double-sided, or n-layers, or single-sided. The routing does what it can automatically (and yes, you can control the thickness of traces and set various spacing constraints). Things that it doesn't manage to connect remain as those connection lines, and you can manually fixup those issues.
That's the basic gist of it. It's frustrating finding names for components, and knowing the "correct" way to do things. For example, you need to add a "SUPPLY" which can be 5V or whatever. But I'm not quite sure how to handle this properly yet.
Still, I can see it's extremely powerful and versatile. Hopefully I'll pick up some skills in the next week or two that lets me build something.