[SOLVED] How to globally put component pads on B.Cu back copper only?

I think he doesn’t use wires but tracks at top. As home-made thermo-transfer has only not plated holes he will have bigger clearance if there will be no pads at top. Limited accuracy of technology is the reason he wonts to not have pads at top (you make the envelope from printed top and bottom and insert PCB into it and then go with that all through modified laminator).
I don’t know why but it looks also that all of these top layer tracks ends at vias and not pads.

I’m not sure if at that layer you will not have the pads also.
May be you can design the PCB as normal - 2 layer (with pads at top) but before printing top layer just delete elements. It can be simpler then modifying all pads ar footprints.

Or just leave the pads in there. They do not hurt and if soldered might improve the longevity of the pcb.

Also, if the top layer is already present why not use it to its fullest? For example to get a proper ground plane.
(I kind of assumed that the top layer is only an imaginary one and that it is only used to document where jumper wires are placed from the original description)

I think they hurt.
For example DIP-14. Pads 1.6mm with hole 0.8mm.
When you have pads you have 0,94mm space and when you have no pads you have 1,74mm space.

From original description:

I understand - not wires but tracks.

Assume technology is track=0.5mm, clearance = 0.5mm - you need 1.5mm space to go with track.
With pads not possible. Without - no problem.

Removing F.Cu from all TH pads should be easy with a bit of python scripting.

Indeed you have it pretty much spot on.

The only reason I have some jumper tracks is If I can keep those down to just half a dozen, I will just use wires and not bother with with top layer. Otherwise there will be a top layer with the equivalent to a silk screen showing board connections / pins assignments (not component designations) in copper and the jumper tracks.

The reason I prefer to use jumper tracks on a prototype design like this is there is no through hole plating on the homebrew board. So, having a pad on the top of the board on a component would need soldering on the top and the bottom pads. This is not very practical or possible in a lot of cases even with long component pads so it is easier to keep all the component pads on the bottom layer and use the jumper tracks in a more accessible place where both sides of the track can be soldered.

I know this seems like a faff but it is still a quick turn around for me to get a few parts of my breadboard design onto homespun PCB rather than waiting a few weeks and keep costing money each time that I do not have.

The final designs will be a production PCB of course.

My apologies, I had also missed that your “jumpers” are actually tracks of copper etched out of the top side.

If this is a common technique that you use, you may want to start making your own versions of the footprints. This will require some significant work to start, but progressively less work the larger your own library of footprints grow. My suggestion is to modify the footprints that you use from the standard library into versions of the footprints with only B.Cu selected for the pads. If you aren’t sure of how to do this, check out this FAQ:


If you need more of a primer and some good suggestions on how to organize your own set of libraries, check out this FAQ:

Edit: Images in the FAQ are working again.

2 Likes

@castingflame You have generated a nonsensical question.

My advice is: Do NOT do this.;Front or Back for a board with traces on one side only is effectively a 1 layer board.

Chose a side, Back(B.Cu) is recommended, and disable the other copper layer in the Board Setup/Layers Menu.

You misunderstood the intent. @castingflame makes two layer boards but wants to have more freedom on the top side. (Removing pads allows more space for traces that would otherwise not fit. They also want to have the top side as empty as possible.)

And most importantly: This is not a nonsensical question! Nearly every user concern is valid. (Some feature requests might be of negligible priority but they are still valid requests. Even requests not fitting with the general plan of kicad are valid in their own way. They show corner cases of what will be possible.)

1 Like

SembazuruCDE

Re your suggestions about new footprints.
Yes, yesterday I did just that. I created a new library for these and saved edited versions (ie B.Cu pads only) of existing footprints and saved them there. It worked perfectly and I thank you for your help. It is much appreciated.

Sprig
I understand that most users no not use homemade PCBs anymore but this does not make this an irrelevant question. A lot of people still make home made boards. The only way to get a plated through hole that is accessible to solder properly is to make a vias that are connected with a pin that is soldered on both sides. I have already explained this above.

Also use the top layer for the jumper tracks (that can be soldered top and bottom) and I use it for labels of pins where needed as I have no silk screen. I have also explained this above.

so in answer to your quote of me…

“It is a 2 layer board but I only use the front layer…” … for Jumper tracks and labels.

“The component pads do not connect to anything on the front layer” because there is no through hole plating so the component have to act as the connection to both layers. This is not always possible to solder the component to the top layer when the component is fitted to it even with long pads! That is why I make those connections Off pin in the form of top layer jumper tracks where the top and bottom layers can can accessed and soldered. I also use the top layer like a silk screen to show pin values/connections as I have no silk screen.

Yes a small percentage of people make home made PCBs now but I am making valid boards that work well and can be made within a day. So this IS a valid question.

The question is resolved now so I will mark it so. Thanks all for your Input.

I get what is you perspective as I also make a couple of homemade PCBs a year, but I think your approach is a bit heavy for lack of a better word. By removing all footprint pads from F.Cu, you make sure that the produced PCB will work 100% as there is no chance that you will connect track from the top side to the THT component, where you will not be able to solder (e.g. IDC connector, terminal block, radial leaded electrolytic capacitors, …)

But you deprive yourself of additional degree of freedom, where you can use THT components that allow soldering from the top side (DIP packages, resistors, …) and you can also use them as a via if you solder a pad from top and bottom side.

Obviously this approach is error prone and sometimes it happens that the designer connects a track from the top side to component where soldering is impossible, but such mistakes are rare (at least in my case)

Removing F.Cu pads from the footprint libraries is just too aggressive step in my opinion if you also design PCBs for fabrications. As you then have to keep two sets of libraries which I imagine is a major PITA.

Writing all this gave me an idea for an action plugin which would:

  • check and brighten (highlight) all the tracks which are on F.Cu and which connect to pads which are on F.Cu. Then the designer can evaluate if there are any errors or not.
  • check the locations of all vias and brighten (highlight) vias which are under components. A via under DIP is not a problem, while a via under IDC connector is a problem when the vias are mad manually on home made board without any metalization.
2 Likes

MitjaN

That is a fair point and it maybe something I may do in the future if appropriate. For certain components like horizontally placed resistors, diodes etc it would be fine but the majority of my parts are digital ICs , DIL resistor packs, IDC connectors, etc at the moment.

I will bear it in mind though and thank you for your input.

I have not looked at any of the scripting in KiCAD yet but I think I will investigate a bit over the next week. My python is poor but as it is C based I should pick it up quickly.

Your plugin sounds useful!

DIL packaged ICs I have soldered top and bottom at 2 layer PCB with no plated holes 40 years ago having no precision soldering iron.
If you only can go with one track between DIL pins I would left pads at top.

I have read that using thermo-transfer people reach to made PCBs with 10 mils (0.25mm) tracks.

Added later:
Being at your place I would probably consider not deleting pads at top but replacing circular pads with oval. It will left enough copper for soldering but allow more space between pads.

1 Like

I really do not think I have a reading comprehension problem. Also, I never stated that the concerns were irrelevant.

And, I’m not the only one:

After considerable effort on my part, it is still not clear to me exactly what the problem actually is.

One. Using leaded through hole parts, the leads can be bent to fit the board, the board does not have to fit the part. It would be trivial to alter a few existing Footprints.

Two. If only wanting to use one side of a board, use SMT parts.

Three.

No good explanation, in my opinion, was provided for wanting to do home spun PCBs significantly different than most common DIY methods. Better results may be obtained by using the more common methods.

Five.

As the Footprints are significantly different, this design process will end up requiring the time to create a second production prototype board design in KiCad.

Finally.
Good Luck!

You sure? (Either you contradict yourself or my english is worse than i fear)

1 Like

@Rene_Poschl Yep, I’m certain. :slight_smile:

The word “nonsensical” is similar in meaning to “illogical”.

I mentioned such to indicate that my following answer may not be the answer to the issue at hand. Yet, was meant to further the dialog to work towards a solution.

ON EDIT: The word “nonsense” is typically used to indicate that a comment has no meaning; gibberish or rubbish. Isn’t English fun?!

Hmm, then it seems strange to not focus on that final design ?
Will the final design also be made not plated thru, top and bottom soldering ?
If no, and you will use the common PTH flows, then you should start the design targeting those flows.
Part of what a prototype does, is prove the overall design flow, before you hit the ‘production’ button.
That mans most designers try to diverge the least, from that final flow.

If you really want a hybrid flow, one means to do that could be to allocate a specific and unique PAD SIZE to the Top layer to-be-removed-for-home-etch pads.
You then edit the Gerber file and change the DCODE for that special size to shrink/remove all those pads.
A non-edited file is used for production.

1 Like

At no point did I ask for a critique! I asked one question. So thank you for the ‘you are doing it wrong’ advice. Maybe if I give even more WHY you may understand a little more. Not that I have to or is anything to do with the [SOLVED] question … but there you go.

These boards are just a way of moving my working circuit off of the breadboard onto something less flimsy with less chance of pulling wires or getting lost. As you all know, breadboards have their own foibles/ weaknesses. Cracked /broken solid wires, pull-outs, no ctrl D after moving stuff!, thin resistor legs in a enlarged hole etc, there are a few, let alone wrong hole placement after rejigging. At some point I like to ‘save’ my physical circuit by making a homebrew PCB that is much more hardy. That will then plug into another breadboard for the remaining part of the design, that may take weeks or months, depending on other projects. I am then much more confident to carry on and have additionally freed up some breadboards.
The final design never has any bearing on the home made PCB component placement so I have no issue about “wasting time doing it twice”. This may not be how other people work. But it reduces my stress levels and gives me the ability to draw a line under a ‘completed’ working example of the building blocks of a part of my circuit. Personally, I get much less ‘issues’ doing it this way and it suites me and how I work and learn.

Everyone else is free to work their own way without issue from me!

1 Like

I think that problem is:
To use top layer for making connections when:

  • using DIL packages,
  • having technology 0.5mm/0.5mm,
  • needing to go with tracks betwean DIL pads.

Not all prototypes are the destination circuit prototypes. And I think such prototypes are the subject here.
I am ordering prototype PCBs from PCB manufacturer and wait each time 5 working days. Actually (to save costs) I have always ordered them when my PCB manufacturer had a promotion day - about 80% discount of initial costs. Since some time they announce “PCB day” typically once per month, but you are never sure. So really my time for prototypes is much longer. Fortunately I am doing some tests for future use so no hurry.
I understand that if someone mastered the thermotransfer technology then he wonts to use it to make prototypes fast.
Isn’t it great to have a PCB 30 minutes after finishing its project?

2 Likes

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