At 08:34 PM 1/2/2009, you wrote:
>Vern,
>
>When you say "Carling switches with loose rivets". How loose? Do
>you mean wobbly terminals?
> I've checked all mine and most of the terminals can pivot around
> the rivet until it hits a stop molded into the switch. I wonder if
> this is tight enough. I would expect them to be tighter.
Interesting! The riveted-tabs are obviously not
as rigid about the rivet axis. Some versions do
drop the tab base into a "pocket" molded into the
back of the housing. See:
http://tinyurl.com/96yqmg
I'm away from home for a week so I can't shuffle through
all the switches in my shop. But even with the recesses
shown, the tab isn't really "captured".
> I have a couple Mil-spec switches and I don't think they pivot
> like this but will check again next time I'm at the shop. Maybe
> this pivoting action is what is not compatible with a high
> vibration environment.
I'm aware of no mil-qualified switches that do not feature
terminals molded into the housing. Here's an exemplar
design (Microswitch) . . .
http://tinyurl.com/842eos
>
>I've had two strobe switch failures, one landing light switch
>(actually a wig-wag) failure, and one master switch failure (loose
>rivet but no charred terminals). I have also received brand-new
>Carling switches with loose terminals.
In retrospect, the only thing that resists rotation of
the tabs on the rivets is friction between tab and
housing established coefficient of friction between
plastic/brass and force of the riveted joint. Obviously,
there are limits to the force you can put on the plastic
housing . . .
>
>As an experiment, I replaced several Carling switches with Honeywell
>switches and have been running them, monitoring for failures. I
>proposed about 100 hours of testing before any conclusions could be
>made. This will probably take several more months. I have check
>them in the interim, with no problems.
>
>If the experiment shows failure of the Honeywell switches, then the
>problem may be attributed to my installation.
>If the experiment does not show failure of the Honeywell switches,
>then the problem may be attributed to the Carling switches
>sensitivity to vibration.
>
. . . and I would expect no failures of these devices.
They're exceedingly well designed and tested in environments
our airplanes will never see.
>Bob has concluded that the Carling switches are fine, but should
>have vibration decoupling loops to minimize vibration stresses on
>the terminals. Good advice no matter who's switches are
>installed. Unfortunately, I don't agree that the problem is that my
>(our) installations are incorrect. Rather, I believe that the
>problem is that the Carling switches are unusually sensitive to vibration.
Perhaps 'error' was the wrong term. Lack of understanding
leads to lack of attention to the sensitivities you've
cited . . . and without a doubt, tabs riveted to plastic
housings have a weakness of design where vibration and
temperature cycles are strong.
>
>In my next project, I am planning to use different switches, plus
>decoupling loops. In addition, strobe and landing light loads will
>be isolated with automotive relays which are generally more robust.
It will be interesting to learn of your experience
with this. Certainly addition of the relays will
eliminate all potential for heating effects.
It occurs to me that a little modification of
the Carling (or similar) switches would add some
degree of robustness. There's a product called
JB Weld that offers a process for beefing up
the riveted joints before the switches are installed.
As soon as I get home, I'll do the treatment on
a few switches and publish some photos.
Bob . . .
