Exposed IC pads

[DL1SIG]

Hi Steve, all,

after seeing the recent discussion about TX issues some users are facing
and the mention of checking/re-soldering the exposed pad connection, I
checked my board from the January 2023 batch. Note that I haven't yet
tested TX because I haven't had the time yet to properly mount it in the
case; receiving seems to work fine. So I just checked because I was
curious and, since it's not mounted yet, the board is still easy to
access.

Based on the published design files the following ICs have a hole under
their footprint to access the exposed pad and the pad should be
soldered: U2, U3, U4, U6, U7, U8, U9, U12, U16

On my board U2 has a clearly visible drop of solder on one side of it's
hole and the solder joint looks fine. On the inner edge of the hole of
U7 there was something that looked like flux residues. After cleaning it
with ethanol it's gone. On the edge of the hole I think I now see the
metal of the pad, in the center it has a black/dark-grey circle (Maybe
from soldering? Definitively looks not like the photo LY2SS posted here,
there everything looks like metal from the pad). Can't see any solder
joint.

On all other ICs I only see the IC (pad metal and, for some, part of the
plastic package) through the hole. No indication that anything has been
soldered.

Of course there might be still a solder connection below the package on
the part of the pad that aren't exposed by the hole.

So my questions are:

- Is my list above with ICs that should have their pad soldered
correct?

- Should there be a solder joint visible through the holes below the
ICs?

- If that's not the case (as described above), is it recommended to
solder them?

Thanks, Simon DL1SIG

[Jiri Culak]

Hello Simon / Group...

I am sorry for late input but since I was offline and soldered my ICs already I cant put more images in, but looking at LY2SS Image that is exactly what I had on U2 U7 etc. I honestly can not say if the groundplane of the IC had or had not connection or if there was a tiny amount of solder which just wasnt visible through the "through hole" but since I manually reflew the GP with hotair and put small amount of solder/flux in I have to say that I dont see any TX issues at all....but I never had any.

Jiri 73 

Dne čtvrtek 23. března 2023 v 23:56:04 UTC+1 uživatel DL1SIG napsal:

[Steve Haynal]

Hi Group,

As I said before, only a thin layer of solder is expected on the exposed thermal pad in an automated assembly process. If Makerfabs were to add extra solder, then the chip would rise off of the PCB too much, solder would short pins, and there would be other manufacturing problems. Makerfabs does not manually solder these holes as there usually is no need. All HL2s pass a screening test which includes measuring full expected power on all bands before it is shipped. So you will at most see just a tiny bit of solder where the PCB meets the IC.

Then why am I telling people to solder these holes? The holes are a compromise between manual and automated assembly.  I think sometimes the holes cause problems for automated assembly. About 30% of the surface of the exposed pad no longer has an opportunity to abut to the PCB. Solder is a poor thermal conductor even if you fill the hole. It is more difficult to apply the right amount of solder paste. It may improve automated assembly results if we get rid of the holes. We decided to keep the holes for DIY assemblers, and we live with a slightly larger chance that they cause a problem during automated assembly.

**You do not need to solder these holes unless you encounter a specific problem and are advised to do so!**

73,

Steve

kf7o

[DL1SIG]

Steve Haynal:
[...]

> You do not need to solder these holes unless you encounter a specific
> problem and are advised to do so!

Thank you very much for your quick response and detailed explanation.
Seems I got unnecessarily worried after seeing that one of the pads had
a visible solder join while the others didn't.

In the mean time I mounted it in the case and based on a quick test TX
is working fine.

73, Simon DL1SIG