NPTH holes on Paste layer

s2x3 · May 7, 2021, 9:02am

Hello

I have one stupid question about NPTH holes

I noticed that in some footprints such holes are described with *.Paste layer

(pad “” np_thru_hole circle (at 3.05 27.19 90) (size 1.27 1.27) (drill 1.27) (layers *.Cu *.Paste *.Mask))

KiCad/kicad-footprints/blob/master/Connector_Samtec.pretty/Samtec_FMC_ASP-134602-01_10x40_P1.27mm_Vertical.kicad_mod#L438


  (pad K32 smd circle (at 5.715 14.605 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K33 smd circle (at 5.715 15.875 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K34 smd circle (at 5.715 17.145 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K35 smd circle (at 5.715 18.415 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K36 smd circle (at 5.715 19.685 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K37 smd circle (at 5.715 20.955 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K38 smd circle (at 5.715 22.225 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K39 smd circle (at 5.715 23.495 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad K40 smd circle (at 5.715 24.765 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad "" np_thru_hole circle (at 0 -27.19 90) (size 1.27 1.27) (drill 1.27) (layers *.Cu *.Paste *.Mask))
  (pad "" np_thru_hole circle (at 3.05 27.19 90) (size 1.27 1.27) (drill 1.27) (layers *.Cu *.Paste *.Mask))
  (model ${KISYS3DMOD}/Connector_Samtec.3dshapes/Samtec_FMC_ASP-134602-01_10x40_P1.27mm_Vertical.wrl
    (at (xyz 0 0 0))
    (scale (xyz 1 1 1))
    (rotate (xyz 0 0 0))
  )
)

and in others - without:

(pad “” np_thru_hole circle (at 3.05 27.19 90) (size 1.27 1.27) (drill 1.27) (layers *.Cu *.Mask)

github.com

KiCad/kicad-footprints/blob/master/Connector_Samtec.pretty/Samtec_FMC_ASP-134604-01_4x40_Vertical.kicad_mod#L353


    (clearance 0.1))
  (pad D07 smd circle (at -1.91 -17.14 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask)
    (clearance 0.1))
  (pad C07 smd circle (at -3.17 -17.14 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask)
    (clearance 0.1))
  (pad D02 smd circle (at -1.91 -23.5 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask)
    (clearance 0.1))
  (pad D01 smd circle (at -1.91 -24.77 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask)
    (clearance 0.1))
  (pad C01 smd circle (at -3.17 -24.77 90) (size 0.64 0.64) (layers F.Cu F.Paste F.Mask))
  (pad "" np_thru_hole circle (at 3.05 27.19 90) (size 1.27 1.27) (drill 1.27) (layers *.Cu *.Mask)
    (clearance 0.1))
  (pad C01 thru_hole circle (at 2 -31.19 90) (size 4 4) (drill 2.7) (layers *.Cu *.Mask))
  (model ${KISYS3DMOD}/Connector_Samtec.3dshapes/Samtec_FMC_ASP-134604-01_4x40_Vertical.wrl
    (at (xyz 0 0 0))
    (scale (xyz 1 1 1))
    (rotate (xyz 0 0 0))
  )
)

Could you please explain why there is such difference and which description is more correct ?

Thank you in advance
Cheers, Sanya

paulvdh · May 7, 2021, 9:43am

I had a look at:

Connector_Samtec:Samtec_FMC_ASP-134602-01_10x40_P1.27mm_Vertical

in the library editor and the properties of the top pad are:

NPTH with a drill diameter specified ( 1.27mm ) is a real hole, and then setting F.Mask and/or B.Mask seems very weird to me.
( [Edit] “Mask” is a typo here, I meant Paste! )

Paste layers are for making solder paste stencils, and putting solder paste into a NPTH is not a good idea. So it looks like a bug in the footprint to me.

RRPollack · May 7, 2021, 3:21pm

Setting solder mask layers on non-plated-through holes as an option makes sense to me. I might not want solder mask around a mechanical hole for some reason.

Setting paste layers on a hole does not make much sense as a default, but allowing the user to configure it shouldn’t cause problems. Perhaps I want a hole in the paste stencil so I can align the stencil with the board more easily during hand assembly?

eelik · May 7, 2021, 3:36pm

It’s actually possible to add paste on a small hole and put a THT component on it. For example a USB connector which has both SMD pads (for the connections) and THT pads (for the metal chassis). If the THT slots for the protrusions in the metal chassis are narrow enough the paste doesn’t flow through the holes totally but when the component is put in its place it’s possible to solder it with a rework station or in an oven.

s2x3 · May 7, 2021, 3:52pm

Well, it could be that holes in stencil help to align it with the board.
But this connector is not possible to solder by hands.
And question remains - why the same kind of connector is described differently in two cases.

SembazuruCDE · May 7, 2021, 4:09pm

But the question the OP asked was about an NPTH hole. No pad, not hole plating, nothing for the paste… Unless you can come up with a reason for solder paste on a non-copper hole, I’m inclined to think the paste layer definition in the footprint was an accidental inclusion, i.e. a bug.

eelik · May 7, 2021, 4:15pm

My comment was just continuation from RRPollack’s comment. A reason why somebody could want a hole pad with paste (although a NPTH is admittedly different). Not a comment on this specific footprint.

Rene_Poschl · May 7, 2021, 5:22pm

In this case it is definitely a mistake as the footprint comes from the official lib. The rules we have would not permit paste on a NPTH (not even on a PTH).

s2x3 · May 10, 2021, 4:10pm

Hello
Will somebody with write privileges correct this mistake in two connectors in git repository ? Files are:

Samtec_FMC_ASP-134486-01_10x40_P1.27mm_Vertical.kicad_mod
Samtec_FMC_ASP-134602-01_10x40_P1.27mm_Vertical.kicad_mod

Rene_Poschl · May 11, 2021, 6:45am

All changes to the official library must go via merge requests. Nobody has direct write access to the master branch.

This is done to ensure every change is reviewed by somebody other than the contributor. (Right now with the severe lack of resources it is easier to get something in made by somebody not on the team as a lone reviewer is enough in that case. If a team member wants something merged than a second reviewer is needed)

Additionally: it might be the case that these footprints are scripted. In which case the mistake should be fixed in the script and all affected footprints be updated.

s2x3 · May 12, 2021, 12:45pm

Thanks for explanation
I’ve made a merge request with corrections

system · August 10, 2021, 12:58pm

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