The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
Reply-To: "Fred Bartoli"
From: "Fred Bartoli"
References: <3E07D4B7.BA697DF5@earthlink.net> <email@example.com> <3E0A751B.93347EF4@earthlink.net>
Subject: Re: SMT boards: assembly and potting Q
Date: Thu, 26 Dec 2002 21:17:50 +0100
X-Newsreader: Microsoft Outlook Express 5.00.2919.5600
Organization: Guest of ProXad - France
NNTP-Posting-Date: 26 Dec 2002 21:17:40 MET
"Robert Baer" a écrit dans le message news:
> Fred Bartoli wrote:
> > "Robert Baer" a écrit dans le message news:
> > 3E07D4B7.BA697DF5@earthlink.net...
> > > I am hand assembling surface mount parts on a PCB.
> > > It is not easy to get chip resistors to align nicely, and SOT-23
> > > are the worst.
> > > 1) What is the best way to keep alignment of these parts for hand
> > > soldering?
> > > 2) If hand soldering is not recommended, what is the least expensive
> > > to assemble these parts at home (ie what do i get and where)?
> > > When completed, i would like to pot them, but need to keep ions and
> > > moisture, as these units are for voltages up to 2KV.
> > > I also am worried about mechanical stresses due to temperature
> > > from (say) -55C to 150C.
> > > I think a conformal coating first may be best.
> > > 3) What is the best potting compound and method to use?
> > Robert,
> > the site fab house that mention use of a toaster oven is
> > I recently came across and found the idea and the performances
> > as I'm balancing about using BGAs for homebrewing.
> > They have "selected" 3 or 4 toaster ovens and as far as I understand the
> > important thing is that they have the power resistors on the bottom side
> > the oven. This way the IR heat the PCB and then the solder past. This
> > a good temperature uniformity across all the PCB's surface (even under
> > component package) and a good temperature control.
> > To ensure an optimum soldering quality, i.e. follow as closely as
> > the recommended temperature profile, you could one or several
> > on a dummy board and make some measures to optimize the process.
> > There's still the problem of solder past dispensing which is normally
> > by printing through a laser cut steal stencil. That's of course out of
> > question for personnal use (cost).
> > Solder past dispensing can be done with a small syringe for the
> > resistors/caps/sot23... small packages, even maybe PLCCs. For the fine
> > QFPs or BGAs the solder past quantity have to be well controled so I
> > think it can be done that way.
> > Maybe someone solve that problem or have experience with this for
> > quantities ?
> > Oh, also don't use the oven you have in you kitchen but rather one you
> > dedicate to that use because of the presence of lead (and maybe other
> > nastiness) in the solder past.
> > regards,
> > Fred.
> 1) I would worry about temperature uniformity and hot spots, unless the
> air was stirred, or a metal plate was placed between the heater below
> and the rack/PCB. The plate makes the heating more uniform and fets rid
> of heat plumes from the heating element.
Sure the plate will provide better temperature uniformity but:
- it will imply additionnal thermal inertia which is *not* welcomed
- I don't think there will be hot spots problems (at least at the amateur
level) with a board heated from the bottom side : hots spots (the
interesting ones are those seen on top of the PCB) will be reduced by the
ability of the PCB to spread the heat across all its surface. This is mainly
because PCBs are anisotropics for heat transfert and is especially true for
boards with supply/ground planes, i.e. the *vast majority* of todays boards
and particularly those populated with SMDs.
- the main heat transfert mode is by IRs, not convection. Influence of heat
plumes, as you call them, is minimized by standing the board sufficently
high above the heating resistors.
- as you note, air stirring might be a good idea. It will also help heat
transfert, which can't be a bad thing.
> 2) One does *not* need to laser-cut anything, much less steel for
> screening. Hundreds of thousands of people have been making cards,
> artwork, putting logos on panels, and making PCBs (*note* NO apostrophe,
> NO ownership!) via the use of SILK screens. And if one needs to re-use a
> silk screen, NaOH or KOH works well to remove even the nastiest of
> resists and the silk screen is only slightly weakened. Many resists can
> be removed with less harsh materials.
Yes, silk screen of course, for *ink* printing.
But I'm sorry, AFAIK, steel stencils are the norm for the purpose of *solder
pastE* printing. The solder pastE quantity has to be tightly controlled,
particularly when soldering BGAs with other packages, and it is common to
have "thick" stencils selectively thinned on some regions for the sole
purpose of printing the right solder pastE quantity on the right pad.
Some are laser cut, some are electroformed by selective metal deposition,
some are chemically etched. They also sometimes have there apertures
electropolished in order to improve the solder paste printing quality. That
says a lot about the required printing quality (again for reliable
I also believe that there is some problems between the silk screen mesh size
and the granular content of the solder past that prevents using silk screens
for that purpose.
> 3) The term is "paste", not "past".
Yes, of course.
Generally I'm very happy to be corrected, i.e. not *when done in such an
abrupt manner* just to take the point.
FYI I'm not english native, neither I speak or write it as fluently as you
do. It takes much more time for me to write my posts (for example a
constructive answer to your question) than it will take for you just because
I consider I *have* to *correctly* express myself for the respect of the
BTW I challenge you to translate your nice answer in a "français académique"
without making any fault in less than twice the time you think I've spent to
answer you. I promise to correct you politely :-)
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup