If I modify a component then update the schematic the component ends up on top of any wires. Is there a way to always force the wires on top. I tried many things but nothing worked other than recreating the wires.
Do you mean you are routing wires across a symbol ?
That’s an unusual operation, probably untested, and I’d suggest you check carefully how that plots, as you seem to be relying on the render order to get the ‘on top’ effect.
I do hobby stuff and only have one sided copper so on the schematic I run all tracks where I can minimising wire jumpers so for example on the image I included (if it works) I have a flip flop with all tracks under the chip as it would be when I do the pcb layout.
That’s a little unusual, so you may need to experiment around the render on both screen and plots.
Did you check plot/print behaviour ?
If you change the background to non-colour, does that give a better graphical ‘OR’ ?
This could get very confusing if the symbol shows connections between internal blocks, looking like wires
Hehe, yes, it will be harder to read in Black and White…
I don’t see any plot on eeschema but print behaviour? I think the way for me is to have no background colour where you can see through, will make it hard to distinguish components, but it what it is… I’ll have to remove colour from components that I add. O’well.
There should be both Print and Plot.
Plot can be to PDF,SVG,DXF,Postscript & even HPGL if you have a real plotter …
It’s just for hobby stuff, If it ever gets printed, I’m the only one that will ever read it, know what it is, yet alone understand it.
O’ ok, I think some confusion. Nothing to do with printing or ploting, it’s the schematic itself on reloading a modified component it then sits itself on the wires. I wouldn’t worry about it, I was hoping that it could be something simple as selecting a different layer for wires, forcing components to the background or making component foot prints transparent but still distinguishable from the background.
Thanks for the suggestions.
Don’t do that.
Really, just don’t.
The schematic is a plan for the circuit and there for you (or someone else) to understand what the circuitry does… it’s not a visualization of ANY layout.
Most often the physical appearance is counterproductive to understand what the circuit does.
Drop that habit while you can.
PS: check out any sensible symbol for an MCU or a more complex IC than 4 pins and compare it with the real world device pin arrangement - they usually are not matching as the symbol is drawn and arranged so, that it’s easy to understand and connect to.
Example:
3 Likes
Believe me, I would love to have a nice looking schematic utilising the default symbols and have tried but using single layer copper and not knowing how to or being able to force KiCad to use one layer I still have to add many pad links. If I do follow the standard symbol I then have to constantly revise the schematic by removing unnecessary pads and or adding them to different wires. As such, In my case, I found that by applying the real world chip layout to the schematic I can then apply it to the pcb layout. In a single layer pcb running tracks under a chip can save a lot of links.
The draw back is that I have to do the layout manually but I just follow the schematic, as it also allows me to pick errors that I otherwise would not have noticed. I learnt a bit about global labels that way, and how they create links… I also can’t use any automation for the layout.
***It just occurred to me that If I allow KiCad to use both layers the top layer can be used as the assumed links layer and alter the schematic with pads at those points then create a vias? Assuming the links are kept at minimal it should work, also allowing for automation of the layout and the use of standard symbols. I’ll give that a try actually.
That’s usually how single-sided is managed, and it gives multiple choices
- Single Sided with Links added, top side traces are ‘jumper’ in nature.
- Double sided, non plated, with top/bottom soldering of thru-pins (track-pins etc)
(this can relax the jumper-nature of top-side routing) - Std PTH prototypes, with no design changes. Volumes mean PTH prices are now quite low.
You can use rules on the top-side/links to mostly keep (eg) 0.1" multiples, and I’d think a script to check/report the jumper-lengths used, would be fairly simple to code.
1 Like
I have a vague memory of an ancient pcb design package that was designed for single sided pcbs with jumpers.
KiCad could achieve similar designs by using two layers with very high priority to use the solder side. The component side “tracks” can then be replaced by jumpers. Even better if the component side is constrained to only route on one axis
1 Like
I always create schematic symbols that closely resemble the physical part. When I assign functions to microcontroller pins, I do so arbitrarily. Once I do the part placement I go back to the schematic and change the functions to suit. It can take a few iterations before I have good pin choices that make an acceptable layout (for board density, EMC, etc).
If you use a scrambled up schematic symbol, it will take you longer to revise the schematic based on a change in layout. You are more likely to make silly mistakes, such as place a signal with fast edges next to a sensitive ADC input. It’s also harder to catch mistakes in the symbol itself because you cannot directly compare it to the datasheet diagram. If you make commercial products, your technician (either in the factory or after sales service) won’t like it either as it takes much longer to find a part’s pin on a printed schematic.
There are sometimes good reasons to make an abstract schematic symbol, but for parts like microcontrollers, logic gates, programmable logic, etc. which have a very high amount of configuration possibilities it is I’ve found it better not to do so.