|
Alternator Load Dumping and Voltage
Spike
I recently experienced a sudden failure of my PS Engineering
intercom during the take off roll, resulting in high volume static
engine noise. At the same time, but not fully appreciated, my
headset microphone was "fried." The intercom was sent
to PSE and the replaced it. The headset was sent to the company
and they replaced it. Both companies replaced their
components without question, but the engineering analysis was
pending.
I installed the intercom and went flying (with a different
headset), and in level flight, once again the same event occurred,
with tach noise and a fried microphone. Yes, I spent time (once
on the ground) searching through all CBs, both independent 14V busses,
all components on/off, all very interesting but not helpful.
I opened the panel and checked nearly everything, every
component, every continuity I could find. Nothing.
Subsequent discussions with the headset engineer comfirmed the
headset controller circuitry was not just fired, but is was scorched,
with burnt innards. High voltage, high amps...toast. PS
Engineering confirmed the same...their microphone and other components
were literally fried inside the box. "What could cause
this?" Again, "high voltage (>28V) and lots of
amps," they said.
The only thing in the plane that can cause that level of volts
and amps is the alternator. Further research revealed a
condition called "Alternator Load Dumping" that occurs when
the battery is suddently removed from the alternator circuit, or the
load on the alternator is suddenly cut off, and the alternator
experiences high voltage & high current output for 2-20
milliseconds (or more depending on decay rate), too fast to be
controllable by the B&C voltage regulator (L-3). This output
can fry electronics on the buss. I then started at the Buss's
battery and worked to the alternator, to the Master Buss solenoid and
to the Avionics Buss solenoid, looking for a loose wire that might
have interupted the battery connection.
Sure enough, I found a slightly loose connection (missing a lock
washer) on the excitation post of the Master Buss solenoid, which
presumably rattled and VERY BRIEFLY interrupted the load demand on the
alternator and battery, separating the two for a couple milliseconds.
This caused the Alternator output to spike and destroy the components
(presumably).
I thought I'd share this observation. I'll let you know if
I'm wrong and another PSE intercom gets fried.
As a footnote, I have two independent busses, each with their own
avionics buss, each with their own set of redundant equipment.
Only the components on the one buss were effected.
Jeff Liegner
LIVP
http://en.wikipedia.org/wiki/Load_dump
If the current load
is reduced to zero very quickly, the alternator's output voltage
becomes its open circuit voltage, which then decays as the field
current decays. If the battery is fitted, it can take charging current
surges from the alternator which are far smaller than the discharge
current surges that the starter motor takes. If the battery takes the
current surges, the alternator current does not fall to zero quickly,
and no voltage surge occurs. If the battery is missing, there is
nothing that can absorb the current and the full open-circuit voltage
appears on the terminals if there is a rapid change of load. The
action of disconnecting the battery can cause the change of load, so
it is particularly likely to cause damage. Load dump can be as high as
120 V and take as long as 400 ms to decay, but it depends on the
magnitude of the load change and the characteristics of what loads
remain.
http://www.sto-p.com/pfp/pfp-transients.htm
Destructive 'LOAD DUMP' transients occur when a
battery is disconnected from the charging system during moderate or
high charging rates. Load dump transients typically reach peak
voltages of 60 to 125 volts in 12 volt systems with relatively slow
rise times. Their duration usually exceeds several hundred ms and can
extend out to 1 second or more depending on the characteristics of the
charging system. Load
Dump transients also occur when heavy loads are switched off although
their magnitude and duration will be lower. These transients are
capable of destroying semiconductors on the first 'fault
event'.
|