This thread is provided to not derail the original.
There needs to be no flame war. I have no significant complaints about the KLC.
What happened is that I realized I wasn’t actually restricted by the KLC if I did not plan to officially share my footprints.
My biggest difference of opinion is probably in the visibility of the silkscreen after part placement. I prefer that the silkscreen not be visible when the part is placed.
This lets me easily read the address of this board.
The silkscreen lines are exactly as long as the body of the part is. If the part is not centered, or is rotated, these allow visual guides to help determine the extent.
The KLC also requires the footprint name on the Fab layer. This is unneeded visual clutter, as the footprint name can be checked in the bottom status bar.
The KLC is only for library contributors. You have plenty of rope if you design your own libraries.
This is for assembly people who have only Gerbers/PDF plot and BOM.
Please note that footprint name in FP Editor will become a part value in the output files!
There was a bruhaha over the state of the kicad footprints over a year ago (?) . Mostly arguing that the footprints weren’t ipc compliant. The number one rule for ipc silkscreens is no silkscreen under a component. It also specifies placement of pin 1 markers and more.
So thays where certain rules come from. The libraries have improved considerably over the past year.
The reason for no silk under the footprint is not (only) because it is in the ipc but because silk under (small) SMD parts is a problem for reflow soldering. (This is why we allow it for THT footprints if it aids manual assembly)
[quote=“Sprig”]
The KLC also requires the footprint name on the Fab layer. This is unneeded visual clutter, as the footprint name can be checked in the bottom status bar.[/quote]
I don’t fully understand your point. How would knowing the name of the footprint I used help for assembly?
Which layer output file will have the MPN as needed for board assembly?
It makes good sense for KiCad developers to comply with Industry Standards; which is why I have NO complaint. It would be what “I” would do if the one in charge of KiCad.
However, I’m not in charge of KiCad, and I feel I have a bit of inner pirate in me.
You are the one that asked me why that prompted me to create this thread.
A possible problem in reflow soldering is a very good reasoning. I already sent out a “test” board to OSH Park and have examined the boards and the parts and have formed the opinion that the silkscreen items are not likely to interfere with soldering with the parts that I am using. I am not using extremely small SMD parts; actually all are 1206 imperial.
I have also “hand soldered” the same device from several different manufactures onto the same pads (with the same silkscreen). It is my opinion that this is going to be the bigger problem in the end.
I am a very much NOOB when it comes to making my own decisions for a Pcb design. However, as a curious kid since kindergarten taking things apart, and as an electronics technician joining the USAF at the age of 18, I have seen and worked with nearly countless Pcb designs; actually some things that pre-dated Pcbs.
It just comes down to me wanting more visual information as I work on my prototypes then the IPC compliant KLC allows.
ON EDIT: Ink does have a cost per drop. If things get to be fully automated with near zero drop-out rate, then the silkscreen that will never be seen by humans will never be printed by robots in the first place.
The first prototypes that I have will be required to be hand-soldered. The smallest parts I’m currently using, and hand-soldering myself, is 1206 imperial sized SMD parts.
Figured this was a good starting point for SMD parts. Playing “Shrinky Dink” while keeping fallout to a minimum will come a bit later.
Please note that footprint name in FP Editor will become a part value in the output files!
The assembly people got a sheet where each component has a simplified outline and value from Fab layers. This way they may check if board is properly assembled.
