Create a footprint for RF PCB antennas with "ENIG only"

[oberstet]

Yeah, after reading your article (which I am still digesting, it’s pretty dense, lots of new aspects for me), I came to the conclusion:

for Sub-GHz, the insertion loss regardless of these materials is tiny, likely practically irrelevant

for microwaves, completely different story, and immersion silver seems really good

---

what’s also interesting is the “skin depth” depending on freq … actually I don’t know how thick the nickel+gold finish is at my manuf. … typical seems 0.5 - 2um … and the article claims skin depth of 2.95um at 0.5GHz, which means it could be right within the Nickel, or slightly into the Copper

But, why not just have soldermask over the copper?

That’s a very good point.

Triggered by this, I tried to figure out how it actually would look like, and recognized that I simply don’t know enough about the details. I should probably get over it, just simulate, test & measure, and hope for the best;)

Anyways, now that I spent the time drawing, let me post my naive pictures for how my zig-zag helical planar antenna looks like physically.

Given the following data/definitions

the option you hint at (if I understood right):

and the option I had in mind:

---

I simply don’t know if above pics are correct, and even if so, I don’t know which one would be “better”, or if it matters at all at sub-GHz … so I basically know nothing to make an informed choice at this point

[BlackCoffee]

I reluctantly post this - look at the screenshot I posted re JLCPCB (above) - they tell user what to do with the VIA (read the Quality Standard text). My point (and please forgive me) is that the info/questions/etc may have been answered and if you hunt further, you may find all the answers you want. As I also mentioned, you need to check with the company making the PCB as they may have different rqmts. Some want to add value by doing some of the stuff…

I leave it here…

[oberstet]

np! thanks for all your hints and insisting I should RTFM;) fwiw, I’ve found a 1h video that explains it all:

How Does Copper Get Inside of a VIA Hole in Your PCB - Do you know this? (with Kailey Shara)

here is a 15min excerpt https://www.youtube.com/watch?v=2hvMgG6tKdg

of course my drawings above had bugs. I was missing the magic of “electroless copper deposition” which fills vias with copper. fwiw, here is the corrected drawing:

I will go with Option B since it provides a “constant stackup” throughout the antenna as it switches sides.

[retiredfeline]

@hermit

It occurs to me that CrapGPT needs to start doing citations.

They are already risking lawsuits for scraping copyrighted content.

[mousey]

I recently finished a project it was only 50mm by 50mm and I wanted ENIG for no other reason than “its my gold and I want to see it !” and the way I did it was to create a zone out of a section of trace or the entire trace and make that zone a ‘mask layer’ this was easy just followed the flow it was very intuitive. Here’s a screen shot in PCB Editor which shows that everything in purple is masked and will be gold ignore the blue its on the back !

Then I sent of my files to JLCPCB to be made and what came back was great and the via’s where uncovered. The boards looked exactly like the 3D rendering so that was great here’s a pic of the front.

and just for completeness here is the back.

So my first experience with JLCPCB and ‘ENIG’ went well for me and I hope you have a great result as well

[oberstet]

oh, nice! thanks for posting your design=) what’s the device doing?

---

btw:

• do you have a camera image of the final PCB as you received it from JLPCB?
• plus, ideally, a screenshot of the KiCad PCB view of your device with the corresponding solder mask layer selected … at least surrounding the SMA?

that would allow me to learn about the correlation between features as designed, colors in the KiCad views (both PCB and 3D), and the final product.

e.g. looking at only the pictures you posted, I’m not sure what’s exactly going on. eg I guess you are using “black” as PCB color, not standard “green”?

did you change any KiCad color settings?

---

fwiw, here are some screenshots of my current design (I don’t have a real PCB yet):

---

did you ever use “edge plating”? I don’t need that for my device (I think), but I’d be interested in learning … things like in this PCB:

[oberstet]

OT, but I’ve read your blog, and " 20%LOAF" made my day! =) maybe I will try your flowchart, might help.

[mousey]

The Board is a very low distortion Sine Wave generator (1Khz in my case). The bulb stabilised bit is reference to the use of an incandescent light bulb in the negative feedback. HP (Hewlett Packard) very first instrument was a low distortion sine generator based in this oscillator and Dr Hewlett did a lot of work on this too. Here are a couple of pics of the real board and I must apologise for the quality I rushed !

and the back

yes my choice of ‘solder mask’ was black and I’m not sure if I changed the background from default but no major changes where made. Here is a screen grab of the PCB showing the front copper (red) and I have reduced the opacity of the tracks so you can see the masks I put on the tracks to keep the gold. Finally I have no experience with getting the edges of the board coated.

I hope this is useful or interesting !

[oberstet]

thanks a lot!

yes, this is interesting … at least for a newbie like me;) I am trying to figure out the relation between the KiCad views (the definition via graphical elements and visual appearance), both layers in PCB editor and 3D viewer, and the final manufactured PCB.

two details that spring to my eye: the inner diameter/surface of the 4 big holes/pads is indeed also ENIG, but the board cutout (the outline / outer edge surface) is not ENIG … it seems … even though I cannot really see since the pics are straight top/bottom shots.

the bottom pic with the PCB editor shot is quite confusing to me: the traces are colored purple, but they must have solder mask removed, as they are ENIG.

the pads themself (e.g. the IC pads) are red, but they also have no solder mask and are ENIG

however, the areas in between the traces is mostly red as well, but is not ENIG.

finally, there is a small space between the purple traces and the red areas surrounding (the space is black in the PCB editor). I can’t see if there is a space/gap in the real PCB between the ENIG’ed traces and the surrounding areas - which are black, as you’ve chosen “solder mask black”.

thus, is the tiny space then “bare PCB material” without solder mask - I’d assume that would be a different color from black (since I’d assume you cannot change the color of the PCB base material itself).

such stuff could only be seen in a close up shot.

anyways, it did help, thank you, it’s all starting to make sense;)

---

very low distortion Sine Wave generator (1Khz in my case)

cool! I’d assume “low distortion sine wave” pretty much requires high-resolution digital generator … DAC based sine wave … is it a digital device? if not, how do you get “low distortion” using analog?