Thru-hole headers on SMD boards...

[[email protected]]

How are thru-hole headers with heat sensitive plastic handled during
PCB assembly these days? Are they soldered manually after reflow?
Basically, I'm wondering whether the board spec (drill hole diameter)
should call for the header pins having a tight or loose fit. A tight
fit would make the manual soldering process easier IMO.

 

[John Devereux]

That's what we do. We experimented with stencil-pasting the thru-holes
and running them through the oven with the rest of the parts, but QC
didn't like the results.

Through hole parts are normally flow (wave) soldered, as a separate
production process. With the plastic part of the header on the
opposite side of the PCB from the solder wave, of course. Normal
plastic headers will take this amount of heat without melting - that
is how they are designed to be soldered. They don't float up in my
experience.

Tight fit keeps the pins aligned better.

You can also get surface-mount pin headers. Samtec has some nice ones,
with alignment pins too.

The through hole ones are much, much cheaper. Stronger too.

 

[John Devereux]

I'll give you the "stronger" -- although the SMT ones seem "more than strong
enough" unless you go for the really low pin-count parts -- but are
through-hole ones that much cheaper if you stick with the same manufacturer
such as 3M or Amp or FTC? I'm just thinking here that I've gotten dual-row
through-board headers from, e.g., Jameco, and while they were quite
inexpensive, the quality was definitely several notches down from the
considerably spendier offerings from 3M.

I'm thinking of the "standard" single or dual row 0.1 inch pitch,
0.64mm(?) square pin types here. They were about $0.003 per pin last
time we bought a load. I have never had any quality problems with the
ones we get. Not sure whose they are right now but they were bought
from a largish UK distributor. The SMT connectors I have looked at
have been a lot more expensive, although of course it can save an
extra production step.

 

[[email protected]]

That's what we do.

Me too. But the headers fit so loosely when the board is made to the
manufacturer's (Molex etc.) header spec that it's a PITA to solder.

 

[[email protected]]

Through hole parts are normally flow (wave) soldered, as a

separate production process.

I know but what do they do when the board is predominantly SMD? Video
boards in particular tend to have SMD components on both sides.

 

[[email protected]]

I'm thinking of the "standard" single or dual row 0.1 inch pitch,

0.64mm(?) square pin types here.

Breakaway headers...Yeah, they're also great for jumper pads, just
break off the amount you need. I buy them in bulk too. Don't see
myself designing for SMD headers anytime soon.

 

[krw]

I know but what do they do when the board is predominantly SMD? Video
boards in particular tend to have SMD components on both sides.

We use a selective solder tool for through-hole parts. It's
basically a solder fountain on an X-Y-Z stage.

 

[[email protected]]

We use a selective solder tool for through-hole parts. It's

basically a solder fountain on an X-Y-Z stage.

Okay thanks. Do you place thru-hole headers by hand or is a thru-hole
automated PnP used? I want to spec my boards so that headers have a
tight fit to make manual soldering by me easier for the time being but
wondering whether this is a no-no for say an assembly house using
automated PnP machines and something like the solder fountain you
mention. Following say Molex's spec for the header holes results in a
very loose fit and I'm wondering if there is a good reason for it.

 

[[email protected]]

I have used a piece of foam and also just some cloth on the headers.

flip the board, solder the headers.

Yes, that works but the easiest route is smaller holes allowing the
headers to remain in place by themselves. IIRC, .033" holes instead
of .035" does the trick for my boards. Pins are square so solder flow
along and around the boards thru-hole plating is not an issue.
Wondering what are the downsides associated with a tight fit.

 

[krw]

What's the rough diameter of the solder "tip" of the pool? (I.e., just how
"selective" is it?)

It needs a .200" component keepout area on the reverse side, around
the thru-hole components (thru-hole components can be closer to
each other).

BTW, it's fun to watch. ;-)

 

[[email protected]]

Wondering what are the downsides associated with a tight fit.

Should also mention that the tolerances that I've encountered do not
justify such a loose fitting. Seems to be overly conservative IMO.

 

[John Devereux]

I have used a piece of foam and also just some cloth on the headers.
flip the board, solder the headers.

Blu-Tack.

 

[John Devereux]

What's the rough diameter of the solder "tip" of the pool? (I.e., just how
"selective" is it?)

I have seen a guy doing this with the main solder wave!

Scary - it was bad enough when I got hot melt glue on my finger.

 

[krw]

Okay thanks. Do you place thru-hole headers by hand or is a thru-hole
automated PnP used? I want to spec my boards so that headers have a
tight fit to make manual soldering by me easier for the time being but
wondering whether this is a no-no for say an assembly house using
automated PnP machines and something like the solder fountain you
mention. Following say Molex's spec for the header holes results in a
very loose fit and I'm wondering if there is a good reason for it.

I think they're placed by hand, but I don't go back there very
often. There aren't very many thru-hole parts (connectors and
transformers, mainly).