Moving components and wires

Schematic
1

I have only just started with kiCAD. I am trying to use eeschema to draw a schematic of a vintage circuit board, but not sure it is going to do what I wanted. This is a little bit different than its intended use and I think maybe I’m wasting my time?

My approach has been to layout the components in the same place and orientation as they appear on the pcb and draw wires where the traces go. So far so good.

Next step is to reorganise everything to make a readable schematic. The problem is that when I move components, the wires don’t always move with them, they just disconnect. I was expecting them to behave in much the same way as ratsnesting in the layout program.

Is there a way to do this, or a better approach to the problem?

2

“Move” (M) will move components without keeping the wires attached (I think this is what you’re doing).
“Grab/Drag” (G) will move components but keep the wires attached (I think this is what you want).

You’ll probably have to clean up the wires after “grabbing” but at least the connections will be maintained.

3

I assume you have an old PCB, and you want to make the schematic for it and re-create the PCB.
Correct?

This falls under the part “reverse-engineering” and KiCad does not have much integrated capabilities for it.

Re-drawing the schematic as if the wires were layed out as on the PCB is not a great idea. The schematic represents function, while the PCB layout is driven by the form of packages and ability to route tracks.

If you want to do this now with KiCad, then the workflow is sort of as follows:

  • Make good quality pictures, or preferable a scan on a flatbed scanner of the PCB.
  • Clean that up in a graphics program.
  • Convert the result with “Bitmap to Component” into something you can load in KiCad.
  • Or: convert it do a .dxf or .SVG (I think only in the nightly V5.99) and load it as a graphics layer in PCBnew.

Then go back to the schematic.

  • Put a bunch of the schematic symbols on the schematic
  • Assign footprints, and put those footprints in Pcbnew with [F8]
  • Place the footprints on locations that match the imported graphic layer.
  • Re-draw a track on copper with the graphic layer as a guide.
  • If you get to a connection, then first make that connection in the schematic, and then:
  • Update the PCB so Pcbnew knows the newly made netlist connection.
  • Repeat until finished.

Here is an example with more info:

4

I was trying Grab, but have since found it works ok from the right click menu, but doesn’t seem to with the hotkey. Rotate doesn’t seem to want to co-operate at all.

However, tidying up the wires after seems harder than it it should be.

5

(EDIT: you posted your later comment while I was writing. Seems that you don’t want to use the dragging feature. 5.99 may have some features for you but you have to try it yourself.)

The schematic is meant to be human readable and as clear as possible. Handling the wires like a ratsnest, just having them from point A to point B, would defeat the purpose. Intelligent automatic handling would require an autorouting algorithm, and it’s notoriously difficult.

That said, the dangerous and unstable development version 5.99 (nightly builds) has this feature. You can drag a symbol and it keeps the wires attached. You can clean it up afterwards. If you have a non-critical hobby project and are a daredevil you can try the nightly builds.

6

Thanks for the detailed explanation. I just want to make a schematic at this point in time. I think that is more than I want to get into.

I think I will go back to pen and paper, then use eeschema to make a tidy copy.

7

I’m another that initially installed EESchema to draw circuit boards, but now I’m in the process of designing a board for a project.

Here’s my system for documenting an electronic assembly to make a drawing in EESchema:

The first list to be made consists of all the components, followed by a series of numbers / letters for each, corresponding to how many connections the part has (used or not). Example: Resistor (1K) - 1, 2; transistor (2N3904) - E, B, C; capacitor (10uF) - Pos, Neg; IC (4011) - 1, 2, … 13, 14

Once all the parts, connectors, etc. are thus documented, you can start on the second list - the netlist.

Pick a point (it usually works best to start at a connector - power or signal) and start with the first pin, write it down, follow it to the next connection, write it down, and continue until you get to the last connection. Any ‘branches’ along along the way, as long as it is a trace or wire lead, means all the components thus tied together are on the same ‘net’.

Once you document one ‘net’, go back to the first list and cross off the corresponding pin number for each part documented.

Repeat with the next selected point, then work your way through the list until you’ve crossed off all the available connections. A few unconnected pins may be left over at the end, go back and make sure they are so, then put that in a separate list.

It should then be easier to draw a schematic when working off a prepared netlist than doing it on the fly.

8

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.