Turns out #KiCad plugins are rather easy to write and easy to use - just drop a python script in your /scripting/plugins directory and refresh your plugin list.
-
Turns out #KiCad plugins are rather easy to write and easy to use - just drop a python script in your /scripting/plugins directory and refresh your plugin list. You don't even have to restart KiCad.
You can even call into wxWidgets to give your plugin a UI.
The applications for this are really endless. Here's a simple example, i wanted to add these fields for the RAM chips that are numbered on the PCB silkscreen. You could do that manually, but I figure if this is something I do a lot, this little script might save me time.
-
Turns out #KiCad plugins are rather easy to write and easy to use - just drop a python script in your /scripting/plugins directory and refresh your plugin list. You don't even have to restart KiCad.
You can even call into wxWidgets to give your plugin a UI.
The applications for this are really endless. Here's a simple example, i wanted to add these fields for the RAM chips that are numbered on the PCB silkscreen. You could do that manually, but I figure if this is something I do a lot, this little script might save me time.
I've gone on to make the entire silkscreen text field generation automatic - it is driven off a custom symbol field template - the value 'seq_silkscreen', if present, overrides the Value field, and is formatted at the specified size and line thickness.
If the text origin matches the anchor point of the footprint, implying it has never been moved, it will try to move it into rough position below the center of the chip, using the footprint pad bounding box as a reference.
If there's no 'seq_footprint' field at all the script tries to set one, using the reference field for resistors and capacitors, or else trimming off the "74" from the Value field and using that, matching Seequa's conventions.
This has saved me hours of tedious tweaking.
-
I've gone on to make the entire silkscreen text field generation automatic - it is driven off a custom symbol field template - the value 'seq_silkscreen', if present, overrides the Value field, and is formatted at the specified size and line thickness.
If the text origin matches the anchor point of the footprint, implying it has never been moved, it will try to move it into rough position below the center of the chip, using the footprint pad bounding box as a reference.
If there's no 'seq_footprint' field at all the script tries to set one, using the reference field for resistors and capacitors, or else trimming off the "74" from the Value field and using that, matching Seequa's conventions.
This has saved me hours of tedious tweaking.
Here's some fun field-driven ideas I can already think of.
- set a custom 'row' or 'column' field, and have a script vertically or horizontally align all footprints that share the same value. Yes you can align stuff manually, but now it is a policy you don't have to think about. you could have a custom 'align' field too, so you could say, ensure these three chips are in the same row, but aligned by their bottoms:
-
Here's some fun field-driven ideas I can already think of.
- set a custom 'row' or 'column' field, and have a script vertically or horizontally align all footprints that share the same value. Yes you can align stuff manually, but now it is a policy you don't have to think about. you could have a custom 'align' field too, so you could say, ensure these three chips are in the same row, but aligned by their bottoms:
The original motherboard did not have 'U' reference numbers - although my hope is that Chameleon owners might start to use the U numbers from my project as sort of a standard from this point forward, the board does have coordinates on its edge.
I have a 'seq_coord' field that I was intending to enter the coordinate for for each component. I was going to do that manualy, but it would be a bit tedious and possibly error-prone, and there would be a lot of fudging.
So of course, let's just set the coordinate by script! The numbers count by inches horizontally, and the letters count every two inches vertically. Again using the center of the footprint pad bounding box, we can calculate a very precise value for the coordinate using two decimal places.
So instead of say, "C3" you might see 1.95, C.62 and now you tell a Chameleon owner exactly where on the board they will find any given component.
-
The original motherboard did not have 'U' reference numbers - although my hope is that Chameleon owners might start to use the U numbers from my project as sort of a standard from this point forward, the board does have coordinates on its edge.
I have a 'seq_coord' field that I was intending to enter the coordinate for for each component. I was going to do that manualy, but it would be a bit tedious and possibly error-prone, and there would be a lot of fudging.
So of course, let's just set the coordinate by script! The numbers count by inches horizontally, and the letters count every two inches vertically. Again using the center of the footprint pad bounding box, we can calculate a very precise value for the coordinate using two decimal places.
So instead of say, "C3" you might see 1.95, C.62 and now you tell a Chameleon owner exactly where on the board they will find any given component.
I might be over-thinking that. Perhaps just the letter/number coordinate is simpler - if you can't find a component in a two square inch area, you probably have bigger problems.
-
R relay@relay.infosec.exchange shared this topic
