Condensation a problem on PCBs w/o a solder mask?

Reply to Thread

We just got back a few boards from APCircuits (great job, BTW) for
prototyping a circuit....2-sided, no solder masks or silkscreen.

Some traces are separated by only 10 mils and it just occurred to us
that condensation might be a problem (shorting out traces) without the
solder mask. The product that these proto boards are part of could
easily be brought inside from below-freezing temperatures and used
immediately.

Has anyone heard if this can be a problem? If so, is there a
recommended spray that we can use after assembly to protect against
this? I suspect clear Krylon might not be the best
solution...something removable would be great.

Thanks!

John Muchow
-- remove SPAMMENOT for e-mail responses --

 

I know of the stuff you want.. but not off the top of my head .. sorry.. we
used to use a clear or red spray on varnish.. had a really nice Z for low
speed stuff and small micros. There's also a range of 'dip' conformal
coatings. If there's a PCB assembler near you.. you might want to have a
talk with them and they might be able to help.

But IMO why the hell are you getting boards made that might possibly
condensate and not getting the soldermasks ?
its going to cost you more to get the boards coated then the cost of the
soldermask

Simon

 

Furthermore, if moisture absorption and/or leakage is a problem, then
do not use FR4 or FR5; use Matsushita's Megtron R5755.
The cost is not that much more than FR4, but the leakage
characteristics seem to be in the same region as Teflon.

 

Furthermore, if moisture absorption and/or leakage is a problem, then

Don't know if absorption or leakage is a problem (probably not, very
low speed circuitry) but thanks for the tip!


John Muchow
-- remove SPAMMENOT for e-mail responses --

 

But IMO why the hell are you getting boards made that might possibly

You're probably right...we had to ask though.

They're just boards for the prototypes...saving a few bucks (and time)
by not getting the masks. We didn't think of the condensation
possibility until after the boards arrived. We'll certainly be
getting masks for the production boards.

We could spray (masking off a DIP switch), but I think it's probably
going to be a lot easier to just tell everyone to let the device come
to room temperature before hooking it up. As easy as it is to do, we
can't think of a reason that someone might hook up the device so
quickly after coming inside from the cold. But, that unlikely
scenario is a possibility so we were hoping for an easy fix for the
proto boards.

I'll probably bring one of the devices home, freeze it, bring it into
a steamy bathroom, and then hook it up and turn it on. It can be
temporarily operated by a current-limited supply, so any shorts
shouldn't be *too* much of a problem. They might damage the device
(no problem) but won't shatter components all over my bathroom (a bit
of a problem).

Thanks for the PCB assembler tip!


John Muchow
-- remove SPAMMENOT for e-mail responses --

 

Rockwell Collins uses Humiseal, but I know that it's not the only mask.

 

Alan already mentioned the product I was going to suggest - Humiseal. Great
stuff!

Here's a link to them: http://www.humiseal.com/

I used to work on a tailgun system on B-52H aircraft. When deployed to
places like Guam, the aircraft would do a high-altitude missions in extreme
cold and then return to tropical climate. Some of our guys would have to go
out after a mission and open up some of the equipment and drain the water
out of them (a good bucket worth of water!). We used the humiseal as
additional protection for the circuit cards in those boxes and for
anti-corrosion on areas that weren't treated - like the EMP filters.

Glenn

 

Alan already mentioned the product I was going to suggest - Humiseal. Great stuff!
Thanks guys! I'll check it out.

John Muchow
-- remove SPAMMENOT for e-mail responses --

 

Actually, if the board is clean (free of ions) then moisture woll give
little or no problems.
Cleaning the surface is not too bad, but if ions are trapped in the
base/matrix material, and it can absorb moisture (read: FR4 and
phenolic), then you may be SOL.
Depends a lot on voltages, temperature, and gap(s) between voltage
stressed traces.
At 175C, things get rather nasty; the Megtron 5755 was a life saver.

 

Are you sure about this? Unless the coating materials are much more
expensive than soldermask materials, I can understand how it could be
more expensive. We are also considering skiping photo soldermasks and
go with coastings after assembling. In the worst case, we just have
to spray with UV cured soldermasks and hit the beach (for sun light),
right?

 

On Thu, 11 Sep 2003 02:43:38 GMT, John Muchow

You could also put a small incadescent bulb inside the device and put
it across the power supply. It will generate enough heat to keep the
condensation out.

 

Of course I'm sure.. but depends if you want to get your coating tested..
Your PCB supplier will have UL accreditation.. does your coating supplier ?
and how many cans of spray on varnish will you use making sure the coating
is flat and even. What about impedance characteristics of the coating ?
Solder mask usually doesn't cost more than a buck or so a board. and unless
your going to dip the board later, it isn't a useless layer, it will also
allow the boards to be wave or reflow soldered, that in itself is a thousand
times cheaper than hand soldering and I don't know of an assembly place that
would accept boards without a coating, is not worth their time... unless
your talking millions. and if you have tight tolerance, you can go closer
to pads if there's a solder mask no worries about solder bridges to traces
with a good soldermask and good rules. you also don't have the same problem
of the boards tarnishing due to finger prints or that miss placed flux you
put on earlier.

then there's environmental.. if the board isn't coated at all.. well .. I
had a box come back.. a very early prototype .. where a mouse got inside..
its pee etched the tracks clean off the PCB.

Simon

 

Not necessary.

I don't think we need UL for 5V, 100mA device.

About 10 cents per sq. in.

Our board has to be hand assembled and soldered anyway.

They are OK with very low resolution board (20 to 30 mils min
spacing).

Solder bridges does not seems to be a problem. It will be done by the
same assembly house. The different is photo soldermask (more
expensive), then assembly; or assembly, then spray plastic (cheaper).

They just have to be sure not to pack extra baggage.

Thanks.

 

you'r right.. you only need UL to sell to Europe or the USA.. neither are
big markets
ANY device going into Europe should have a CE mark.. that requires PCB
testing.. The trusty UL mark gets you a rubber stamp.

thats not a bad price.. I seem to remember the spray costing more.

then it ain't a good design unless you only plan to sell 10.

well 20/20mil is a common spec for single sided PCB's.. but seriously .. if
you talk to the assembler and ask "If it could be made for automatic
insertion.. how much wuold it cost to assemble?" you might find they divide
the cost by 3.. or 10.. I don't know whats on the board so its hard to
comment. There might be some old tech assemblers who stuff all through hole
by hand still but automatic assembly is the only way to go for volumn.

If theres no real volumns.. then any old how is good enough just be sure to
do your best and don't go into red ink