kicad / kicad-footprints

Official KiCad Footprint Libraries for Kicad version 5

⚠️ 301 Moved Permanently

Location: https://gitlab.com/kicad/libraries/kicad-footprints

kicad-footprints's Issues

Polarized THT capacitors do not have a pin 1 mark that is visible after assembly

See discussion in #71
This should be fixed within the generator scripts

Scripts are found in: https://github.com/pointhi/kicad-footprint-generator

CP_Radial_SMD have fixed chamfer of 1mm

Hi
since the SMD radial capacitors have wrong models, I wanted to make sure the footprints are up to date before making the new models.
first thing I noticed is all the footprints has the same chamfer (polarity, not pin 1) of 1mm for all sizes.
should it be kept like this or wouldnt a percentage chamfer make more sense?
please pitch in some opinions so I can get started :)

JAE LY connectors

@SchrodingersGat
I assume KiCad/Connectors_JAE.pretty#3 had a scripted family of these connectors but it didn't get merged. While these might be a few more tweaks to make, it seems a waste to never add these footprints.

THT capacitors require rework

See notes here - #71

The scripts need to be reworked to place the polarity indicator on the Silkscreen layer outside the footprint, so that it is visible after component is placed.

Archive legacy repositories

@SchrodingersGat
What about marking the legacy repositories as read-only, to prevent contributors making pull requests on no longer maintained repositories? https://help.github.com/articles/archiving-a-github-repository/

Should THT components include a courtyard area on the bottom?

I think kicad now has a DRC check for courtyard violations.
But our THT footprints don't include courtyard on the bottom. So if one would place something too near on the bottom then this check would not notice it.

What about manufacturer specific pin headers and socket strips? (Samtec)

Currently the samtec lib holds some socket strip footprints. Should we transfer these specialized footprints over here? (They have slightly different drill sizes: 0.95mm compared to 1mm for the generic footprints)
Samtec creates quite a lot of slightly differing pin headers and socket strips. I am not sure it would be a good idea to include specialized footprints for all of them. (But i am also hesitant to simply "delete" contributions.)

Edit: these have been added by @SchrodingersGat in pull request KiCad/Connectors_Samtec.pretty#6

Who is working on what?

I think we should definitely coordinate our effort to avoid doing work twice.

The next thing i will do is look over the JST scripts to move these connectors over here. (Yes i know i did not create most of them but i want to do one complete repo at a time.)
Before that i will transfer the unchanged contents of the Resistors_THT, Capacitors_THT and Diodes_THT repos.

@SchrodingersGat, @jkriege2, @Shackmeister, @evanshultz

Issues with potentiometer footprints

This will be a living list as the footprints are finished up and merged. Some of them will have the script as the root cause and re-writing or extending the script wasn't in scope for v5.

General:

• Pin number ordering (CW/CCW/etc.) cannot be specified. This prevents Bourns PRS11S from being built to the datasheet.
• Pinxoffset variable controls body placement as well. (Pins, body, and shaft should be independent.) This prevents from being generated correctly.

SMD:

• SMD pad sizes are all the same. ACP CA6 VSMD has two different sizes, for example.
• Potentiometer_Vishay_TS53YJ_Vertical or Potentiometer_Vishay_TS53YL_Vertical.kicad_mod (forgot which one) isn't generated on-center.

Better communication about the fact that Connector_Coaxial:BNC_PanelMountable_Vertical is a missused panel mount.

See discussion in #111

wrong sizes for R_1020

@poeschlr
Hi Rene I noticed R_1020 and the handsoldering version both has a body size of 5x2.5mm, this seems quite big :)
can you look into it?

What should we do with the microwave library

I am not really sure what these footprints represent.

Some of them have smd pads with all copper layers set
Not sure they should include paste.

footprint for AS7261 (Package_LGA:AMS_LGA-20_4.7x4.5mm_P0.65mm) is missing

Symbol footprints seem a bit large

More details see old issue (archived old footprint lib repo) KiCad/Symbols.pretty#8

@pointhi I assume you have them as svg and exported them with svg2mod? Are the original svgs somewhere available on the net?

Missing Bottom Courtyard for Edge Mount SMA Amphenol 132289

SMA_Amphenol_132289_EdgeMount.kicad_mod is missing bottom backyard (you can compare with SMA_Molex_73251-1153_EdgeMount_Horizontal.kicad_mod).

Most footprints for packages that have Exposed pads to not conform to KLC with regards to paste layer implementation

Violated KLC rule: http://kicad-pcb.org/libraries/klc/F6.3/

This rule states that there should be one large copper pad (paste layer disabled. Mask enabled unless a specific mask size needs to be achieved.)
Overlayed with smaller pads for paste only. (This pads have no copper enabled and no pin number assigned.)

---

Currently a lot of footprints (nearly all of them) have multiple copper pads that have simply a negative paste clearance or solder paste ratio set.

https://github.com/KiCad/Connectors.pretty is not deprecated yet

Testing and travis ideas

I had a look to the kicad library utils repo the last weekend. I have some ideas that I am willing to contribute to but I don't know what you think about them.

1. Testing now requires cloning more than one repo (travis clones both this repo and the utils one). I think this could be avoided by moving the testing code to this repo. The reused code between the different testing benches for PCB, SCH and SCHLIB is minimal, thus being a good option. This would reduce check time. Keeping the test code in the same repo would also make easier point number 2.

2. Using relative paths in scripts would ensure that scripts can be run both localy and in travis environment. Currently some scripts can only be used by travis. My proposal is to include a batch and a shell script that the user can click or execute in a terminal that will test the last modified components (I was thinking on testing the whole lib if in the master branch and the modified components if in another branch or something like that).

3. Testing now requires sudo and there seems to be no obvious reason for that. Sudo makes travis testing significantly slower. Dropping sudo needs would improve check times a lot.

4. I was thinking in including a coverage report. A simple script that would tell in which percentage the whole lib complies with KLC and/or another script that would test how many 3D models are present and how many missing footprints looking at the symbol lib.

Please let me know what you think about the previous ideas and if you would like to see them implemented.

Thank you!

KA5H0265RCTU is missing a TO-220-5L package from Fairchild/ONSemi

Datasheet: http://www.onsemi.com/pub/Collateral/KA5L0265R-D.PDF

The package seems to be particular for that manufacturer. The name is TO-220-5L.

I would suggest the names Fairchild_TO-220-5L or ONSemi_TO-220-5L (since Fairchild has been acquired by ONSemi). Which one should we use?

Footprints with oval drill. Should we require alternative with circular drill?

We have a few footprints that use oval drills.
Some of them also have a second footprint that uses a larger circular drill instead. Should we require such an alternative footprint (where possible) such that users who have not access to a fab that can create such a hole can use them?

Inductor THT: Missing 3d model references, second reference (fab) interferes with fab and some silk lines on top of pads.

This is ideally fixed within the generator scripts directly. (Fixing the inductor references would also fix some of the resistor references because they share the same generator scripts.)
So #74 is related in part to this issue.

Script location: https://github.com/pointhi/kicad-footprint-generator

Footprint libs as submodules?

@SchrodingersGat i thought you where against having the footprint libs as submodules.
https://lists.launchpad.net/kicad-developers/msg31020.html

duplicate of wuerth card connector

We have three card connectors with the same manufactur number, but different writing:

• https://github.com/KiCad/kicad-footprints/blob/master/Connector_Card.pretty/microSD_HC_Wuerth_693072010801.kicad_mod
• https://github.com/KiCad/kicad-footprints/blob/master/Connector_Card.pretty/microSD_HC_Wurth_693072010801.kicad_mod
• https://github.com/KiCad/kicad-footprints/blob/master/Connector_Card.pretty/MicroSd_Wurth_693072010801.kicad_mod

What should we do with the wire pads footprint lib?

The test point lib has been transferred in #52
The question about the wire pads lib has been raised in that pull request already.

FSDH321L and FSDL321L are missing an 8LSOP package

Datasheet: https://www.fairchildsemi.com/datasheets/FS/FSDH321.pdf

Components FSDH321L and FSDL321L available at the symbols repo (Converter_ACDC) are missing a footprint.

BGA footprints: Paste and mask definition not ideal for many footprints

More details see old issue in (archived) old repo:
KiCad/Housings_BGA.pretty#16

TE connectivity repo is archived without notifying the contributors of open pull requests about that fact

@SchrodingersGat i don't know a better place to place this. (I can't open an issue nor can i comment on the still open pull requests.)
Can you notify them about the fact that they should commit their stuff to this repo if they are still interested?

Lower case "d" in Connector_Pin library

KLC defines that diameters are prefixed with upper case "D".
Fixing this will also require fixing the 3d model names.

This has been discovered in KiCad/kicad-packages3D#165

Auto-update of fp-lib-table by KiCad?

@SchrodingersGat
Will KiCad v5 auto-update fp-lib-table? If so, then footprint repo changes can be pushed out to users. Otherwise, they'll be unable to add or remove libraries.

The schematic has the same issue.

Discussed somewhere here and at https://bugs.launchpad.net/kicad/+bug/1705291. Here is the root of the question: If we have two files that capture all current libraries, whether they're manually- or automatically-generated, why shouldn't the GitHub plug-in also update those files so the users always has the latest and greatest. For users who don't want to use the GH libraries, that's fine, but those that do can be kept updated between KiCad releases.

I realize this isn't a core library issue. But if there are reasons this can't be done by the library team, or isn't ready to be done, then of course there's no reason to have KiCad do it. But if all the infrastructure in the library is ready, and it's a good idea, then it needs to be raised to the KiCad dev team.

Polarity markers for Crystal footprints is missing

For details refer the original issue over at the old repo:
KiCad/Crystals.pretty#23

---

It seems most of the footprints are scripted. So fixing this inside the scripts is preferred:
https://github.com/pointhi/kicad-footprint-generator/blob/master/scripts/Crystals_Resonators_SMD/make_crystal_smd.py
https://github.com/pointhi/kicad-footprint-generator/blob/master/scripts/Crystals_Resonators_THT/make_crystal.py

---

In addition, add information about the scripted nature of these footprints to the readme of the lib.

Some of the second references (on fab layer) for resistor THT footprints interfere with other graphical elements on fab

Some oft he references need to either be scaled down or moved a bit.
Preferably done with the original generator scripts.
But can also be done by hand. (Only a few components are affected.)

Script can be found here: https://github.com/pointhi/kicad-footprint-generator

LEDs and color

Hi @evanshultz @poeschlr @SchrodingersGat ...

When I recently moved the LED footprints and 3D mdoels to the new repos, I stumbled upon a small problem: All the THT LED 3D models just have any color (most 2-pins red or green, 3/4/...-pin-version are white). That's mostly fine, but there are 3 problems with that:

1. If you want your 3D model to have the correct color, you have to make new 3D models and matching footpritns ... should we provide a list of such? And if so, how do we denote color in the names (e.g. as modifier _Red, _Green, ...)?
2. We should have clear/white and maybe grey/black variants at least of the 3mm and 5mm, as they are often used for Phototransistors/-diodes and IR-LEDs ... what do you think? Whould I add them and how should I denote the color in the name (see item 1).

Best,
JAN

Questions: Which pin is pin1 on a diode bridge rectifier?

Hi!

I'm currently adding several diode bridges to the symbols, footpritns and 3D models. With the typical SIL-packages, I'm a bit stuck, as I'm not sure which pin to designate pin 1:

Could you quickly state your preference?

For now I took the left pin, but currently I tend towards using the pin on the right, marked by the package bevel! This would also prevent us from the situation that pin1 is marked on the silkscreen, but the package has a mark at pin 4 ...

Best,
JAN

Should KLC specify the paste coverage ratio of the exposed pads?

Currently we only state that the paste should be broken into multiple squares.
Should we specify the percentage of the area that should be covered?

Rule in question: http://kicad-pcb.org/libraries/klc/F6.3/

SIP3_11.6x8.5mm is not compatible to RECOM R-78

AFAIK the footprint kicad-footprints/Package_SIP.pretty/SIP3_11.6x8.5mm.kicad_mod is the replacement of Converters_DCDC ... RECOM R-78 footprints.
Both, the original footprints in the Converters_DCDC package and the SIP3_11.6x8.5mm are not compatible with the RECOM R-78 parts.

What's wrong:
The courtyard and silk have the large part upwards, while is should be downwards. The datasheet R-78Bxx-1.5 has a recommended footpring graphik on the last page, with pin 1 marked.

Connector_IEC_DIN

This legacy library has now been archived but the footprints still need to be transferred across to here..

Diodes THT: Polarity marker on silk not ideal for most footprints and second (fab) reference interferes with other fab graphics

This should be fixed with the generator scripts.

Nets on wing pads of connectors

Not sure if this belong here or on the forum, let me know if its inappropriate here.

How are people handling connecting the wing pads on connectors to different nets (typically GND)?
usually I just edit the pads and add the net I want, but this is lost when updating the netlist, which usually makes me forget to change some of the 20ish pads. Unless somebody has a good way, we should maybe consider making a standard padnumber for these. This way the user could use a symbol with or without the wings depending on the nescessity.
Thoughts?

Shouldn't this repo be pinned to frontpage?

See above :)

Should we create a separate library that captures all FFC/FPC connectors?

A lot of manufacturer seem to create such connectors. Currently we have footprints for connectors by at least Molex, TE and Hirose.

I would suggest a new lib called Connector_FFC-FPC and group them in there.
Naming of footprints would then be [man]_[mpn]_{1xN or 2Rows_Npins}_[orientation]

KA5H0265RCYDTU is missing a TO-220-5L(Forming) package from Fairchild/ONSemi

Datasheet: http://www.onsemi.com/pub/Collateral/KA5L0265R-D.PDF

The package seems to be particular for that manufacturer. The name is TO-220-5L(Forming).

I would suggest the names Fairchild_TO-220-5L_Staggered or ONSemi_TO-220-5L_Staggered (since Fairchild has been acquired by ONSemi). Which one should we use?

Related to #240.

Missing audio connectors

https://github.com/KiCad/kicad-footprints/blob/master/TODO.md doesn't include (or doesn't include many) various professional audio connectors:

• 1/4" / phono
• XLR
• speakON

Some of these footprints exist in the old library but are outdated. I'll update and expand this collection of audio-specific connectors.

Incorrect scaling for THT diode models

Many of the THT diodes have a non-unity scaling for 3D models.

x:y:z must be 1:1:1

repository name

will it be nice and unique to have this repository name as "kicad-footprints" in sync with the schematic symbols repository name "kicad-library" ?

Footprint naming and placement of second reference designator not ideal in OptoDevice lib

Examples for badly placed second references:

Naming convention is a bit harder. We need to find a system for part prefixes that would group similar devices together

Missing fuse holders

Looks like good footprints for fuse holders were deleted. They were described as "Fuse_Holders_and_Fuses:Fuseholder5x20_horiz_open_universal_Type-III", type II and I (also other footprints with 5x20 and 5x25 in name were good). Please restore them.

Button_Switch_SMD.pretty naming convention not continuous

We have some mixture of rotary encoder without a prefix, switches without a SW_SPST prefix and switches with/without manufactur name

Some RF shielding footprints miss an inner courtyard.

Reported by @pointhi in the pull request that added these components here. #75

Polarity markers for Oscillator footprints

For details see original issue in the old repo: KiCad/Oscillators.pretty#6

adding a Conn_Coaxial library?

Hi guys
Wouldnt it make sense to have a library for all SMA, MCX, UFL etc connectors?