|
|
Back in the the pre-stone age, when ICs didn't exist, I worked for Beckman
Instruments.
All components on PC boards were discrete, individual parts- transistors,
resistors, capacitors, diodes, etc. Some of the instruments were designed
for airborne use.
Finn's description of parts snaking about on the boards under test
conditions is accurate (jogged some of what's left of my memory).
The components were mounted flush with the circuit board. One of my duties
was to "pot" these components to the PC boards using a clear, flexible
epoxy. To replace a component (say a resistor), the potting compound was
burned from around it with a 25 watt soldering iron.The entire PC board was
covered, not just the components.
This stuff was somewhat stiffer, but very similar to clear silicone caulking
compound. One might consider using such a material (the silicone) to make a
pretty bullet-proof electronic assemby. After mounting the components flush
to the PC board. And giving them a good long burn-in.
I think the mechanical failure mode of aircraft electronic components is
more likely to be vibration than g-stresses (or dropped wrenches), and even
though your airplane feels dead smooth, it's probably vibrating at all kinds
of frequencies that can't be perceived by the human posterior, and it
doesn't take much amplitude over time to separate a lead from a capacitor or
make a two-piece diode.
FWIW
Jack Ford
----- Original Message ----- From: "Charlie England" <ceengland@bellsouth.net>
To: "Rotary motors in aircraft" <flyrotary@lancaironline.net>
Sent: Friday, February 18, 2005 7:46 PM
Subject: [FlyRotary] Re: Digital Fuel Monitoring System
Hey Ed, want a 2nd opinion? What Finn said.
Same reason that 13B through-bolts break.
Charlie
Finn Lassen wrote:
> Hi Ed.
>
> This of course all only applies to electronics in a vibration exposed
> environment.
> Good practice, as I was taught it many years ago, is to have hole
> spacings to allow components to lie flat on the board.
> Of course there are exceptions like power resistors that need airflow
> around them, but they should be mounted with "stand-offs" (small tubes
> on their legs). The rule used to be that components should not be held
> in place by the solder, but by their mechanical mounting. That's why
> the pins are typically bent 45 degrees where they stick through the
> PCB before being cut off and soldered. Heavy components are
> additionally secured with some kind of mechanical fastener or globs of
> goo (technical term). Now obviously the weight of the body of the
> component in relationship to number of pins/legs is a factor. An
> example would be an IC with oodles (another technical term) of pins.
>
> I "grew up" in a company that secialized in sound and vibration. It
> was very educational to watch a PCB board mounted on a "shaker"
> (basically a loudspeaker where there membrame was replaced with a
> mounting place) lit with a strobe light synchronized to the frequency
> sweep. Components dancing around and finally falling off if not
> properly supported (mounted).
>
> But that was then. Now we have surface mounted components which I know
> nothing about...
>
>
> Finn (Looks like I've been kicked off the list again)
>
>
> Geez, Finn, I was unaware that there was a "protocol" for resistor
> sizes on PC boards. Seriously, is there some reason (vibration?) for
> forcing resistors all the way to the board or is it just "esthetics?".
>
> Ed A
>> Homepage: http://www.flyrotary.com/
>> Archive: http://lancaironline.net/lists/flyrotary/List.html
|
|