On 12/03/2012 11:06 PM, Rowland Carson wrote:
> But it seems the diagram I attached to my earlier posting did not
> make it to the list (for some unknown reason I didn't get an echo of
> the message myself so can't be sure), and it clearly showed earth
> connexions to the 2 wire screens and the 2 switch cases as well as to
> the 2 normally-open poles of the A & B switches.
Yes, but it is separate from any other ground connection in the
instrument panel... That means, if you have a ground bus bar in your
instrument panel, you should take care that it doesn't somehow connect
to the screen as well.
This would for instance happen when you have a metal facias on your
panel. Your ignition-switch touches it, and some other switch from,
let's say your landing light, touches it as well. This other switch has
a ground connection to your ground bus. Without realizing it you have
just created a ground loop.
Now suppose something happens to your landing light ground connection.
The electrons now find a path via the switch case to the metal facias,
to your ignition switch, and then via the ignition screen, to ground.
Your landing light will still work, so you won't notice that something
isn't as it should be. Maybe you will only detect a funny smell, when
the 3 amps of your landing light travel via the facias and screen of
your ignition cable, a cable not designed to run the 3 amps via the screen.
Ok, you make good connections you say, your ground cables won't fail. I
take that for granted. But here is another problem, and that is
electrical resistance. Your perfect ground cable has still resistance.
And when your starter cranks it will pull a few hundred amps through the
cable, and due to the resistance some electrons find an easier route via
the screen of your ignition cable. Repeat this a dozen times, and for
some "unknown reason" suddenly your ignition cable fails.
Some day you have a short cut somewhere in your electrical system. No
worries, you have a circuit breaker. But just at the instant between the
short cut occurence and the opening of the circuit breaker, you have an
amp spike on your system, possibly a few hundred amps. It doesn't cause
a fire because it only lasts a microsecond, but at the moment your
ground bus is overloaded some electrons will choose the screen as an
easier route. The amp spike on the screen causes an induction in the
core wire, and that core wire is attached to your precious ignition
coil... Just cross your fingers and hope it doesn't do anything bad.
I can give you many more examples. The bottomline is: Ground Loops Are
Bad (Pun intended). Always make sure there is only ONE way to ground.
This screen of the ignition wire makes your system vulnerable to ground
loops, hard to trouble shoot because you think "ground is ground" and
undetectable for simple measurements because you can't measure which
path the electrons have taken.
A much safer way is to connect the ignition switch to your ground bus in
the instrument panel, and not connect it to the screen. The result is
exactly the same, but the dangerous ground loop can never develop.
> I understand your concern - IF the normal earth should fail.
Not only when it fails, but also when it has resistance... and it has.
Of course it is your airplane, but think about it for a moment, why you
should prefer an error-prone solution over a safer solution without
sacrificing anything in the process.
> I plan
> to use a B&C common earth system which has on each side of the
> firewall a forest of male blade connectors soldered to a brass plate.
> A brass bolt about 8mm diameter passes through the firewall to
> connect the 2 plates, and provide an anchorage for a ring termination
> on a 10AWG wire from the battery.
When you are done installing this, switch on all appliances in your
airplane, and take a multi-meter and measure the voltage drop from your
battery to the ground bus in the cockpit. You will be surprised!
Frans
|