>     Subject: Alternator Voltage Spikes
>       I have been noticing all the comments about voltage spikes during
>     starting/shutdown. One poster mentioned that a battery couldn't put
>     out 200
>     volts, so I thought I would toss my 2 cents in. I worked for several
>     years
>     as an IC design manager at Texas Instruments in the automotive group
>     where
>     we designed ICs for automotive regulators. I will assure you that
>     charging
>     systems for cars (and airplanes) do indeed put out spikes well over > 200
>     volts when the alternator is rotating very slowly. This voltage can
>     easily
>     cause oxide failure, known as 'punch-through', which shorts VCC and
>     Ground
>     inside the IC. Once that happens, heavy current flows through the IC
>     causing
>     the smoke stored in the IC to escape. ICs don't work once you let
>     the smoke
>     out! :-) To understand how a 12 volt battery/alternator can put out
>     a high
>     voltage, you have to understand that in an inductor (i.e. the
>     windings of
>     the alternator), the voltage across the inductor is equal to
>     L*di/dt, or the
>     Inductance of the winding times the rate of change of the current
>     through
>     the inductor. If you attempt to instantly stop or start the current
>     through
>     an inductor, di/dt will become very large, and the voltage will
>     increase to
>     whatever level necessary to collapse or create the magnetic field
>     around the
>     inductor. When an alternator is turned very slowly, there are points
>     where
>     the windings are open circuited (or routed through high resistance
>     paths)
>     which causes voltages to rise to the level that breaks down the > primary
>     protection - around 200 volts. In an automobile, there is a highly
>     reliable
>     circuit that disconnects the radio bus from the charging bus during
>     start
>     and shut-down to protect equipment from these spikes. This is what
>     Cessna
>     attempted to do with some aircraft, but the circuit they used is
>     somewhat
>     crude and potentially unreliable, so people disconnected its output
>     that was
>     supposed to control the avionics power relay, and rewired this relay
>     to be
>     controlled by a simple switch on the panel.
>     Marc Wiese
>
>
>
> The energy stored in such a 200V pulse is actually quite difficult to
> transmit to components beyond the charging system.  The energy is low -
> the source is not tremendously inductive, and the peak current at low
> RPM is.. low.  Plus, because the pulse is so narrow (fast rise, fast
> fall), it has high frequency content.  Transmission of high frequency
> signal is difficult within an impedance controlled environment (coax
> cable).  Transmission of said energy through such lossy/dispersive
> environment as battery cable is inefficient.  So, while the pulse may
> pose a threat to internal regulator components (because of their
> proximity to the pulse source), it's much less of a concern to remotely
> mounted components.  Especially when the remotely mounted components
> already must include protection from human body model ESD events
> (several kV).
>
> Cessna had some problems with early transistor avionics (Ge based, as I
> recall) which were inadequately protected from ESD events.  I'd bet as
> many radios were damaged by techs not taking adequate precautions as by
> starting/alternator "spikes"..
>
>
> Regards,
>
> Matt-


Easily fixed.
Just connect a VDR (voltage dependant resistor) across the 12volt feed to
any electronics. These devices are nomally open circuit. When the voltage
across them exceeds their rating they are short circuit.

Ian. RV7 Renesis, bolting the wings on.

AeroElectric seems to have a neat solution for an over voltage scenario.  I
purchased an Over Voltage Protection module (crowbar), for $35, from
bandcspecialty.com.  It seems like an inexpensive way of protecting internal
electronics from surges.  I'm not an electrical engineer, but the logic seems
right when I read the documentation -- but I do give a little blind trust in the
merchant, as like I say, I'm not an engineer.  You can also make your own from
the attached PDF.  I downloaded it from this AeroElectric website
http://www.aeroelectric.com/articles.html

Cheers,

Burck



The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.