|
Al,
Thanks for your reply. I tried a few more
things since my earlier post. I put a relay in the alternator output line
(controlled in parallel with the field power, so I could start with the filter
out of circuit. Engine started fine, turned alt on (now with filter in
circuit) and the engine continued running, but same results as before, missing
ignition pulses. I am still perplexed and no plausible explanation why a
filter (read capacitor) would prevent starting. Since the filter did not
cure the miss, I moved on to try again to figure out how the spikes from the
alternator (normal or not, will scope my cars alt tomorrow to verify what I am
seeing is normal, I do not believe I have any blown diodes.)
SO what is the underlying cause. Looking at
this the last couple of weeks (very limited time due to work and cold temps -
20F in SC is really cold, thought that was why I left the north.) I was
suspecting that this alternator noise (spike) was inducing a current in the
injector lines that was back feeding into the EC2 and if the spike occurred at
the same time the EC2 was triggering, the ignition pulse would be
inhibited. I had verified, with a scope, at this point that the miss was
in fact a missing ignition pulse.
So working under the assumption that the noise was
back feeding into the injector driver, I was considering construction an opto
isolator to test this theory. Then you comments prompted me to take
another look at my wiring. The injector lines are all twisted pair,
shielded, shield grounded one end, cross any power lines at a 90, no parallel
runs closer than 4". It all looked good. Then a light came on.
The alternator wiring ran along and was attached to the aluminum mounting
fixture that hold the secondary fuel rail. The steel (read
transformer core material) secondary fuel rail that is connected to the
steel braided fuel lines. One of the steel braided fuel lines that runs
along the same path as the fuel injector lines. So the enhanced theory is
that the fuel rail/lines were acting like a core in a transformer.
The test was to disconnect the existing alternator
line between the connection at the starter and the alternator, and run a new
line that stayed well away from and steel fuel components. Jumping
back into the pilot seat which has never been more than 3' above the ground (3
1/2 when on jacks), I flipped all the switches I have so may times before,
waited for the Pentium 4 to boot up. After the P4 booted, I started the
engine monitor, started comms with the EC2, primed 2 shots and hit the start
button. As usual (without filter) she (yes, she is short for the Glass
Mistress, don't tell my wife) fired right up. Now the moment of truth,
with fingers crossed I flipped on the alternator, unintentionally holding my
breath I waited for the first miss. The first miss came but later than
normal, at which time I realized the mixture was a bit rich, leaned it a bit and
continued to wait for that oh so familiar miss, and continued to wait. Can
you say purrrrrr.
Still perplexed by the no start with filter
installed, but that can wait for another day. Thanks for your random
thoughs, they encouraged me go go back out into the cold to try a few more
random thoughs, one of which discovered the underlying problem. I am
anxious to rerun in the AM to make sure my results were not a fluke.
Then I can return the Fluke Scopemeter to the office.
Sincerely,
Joe
----- Original Message -----
Sent: Friday, January 15, 2010 7:33
PM
Subject: [FlyRotary] Re: Alternator
Filter
Joe;
Gees, that’s a
weird one.
I assume from your
note that problem is no ignition trigger pulse, and that the injectors are
firing. Have you determined where the signal is being interrupted?
Is the loss of signal from the CAS, or out of the
EC2? The capacitor grounded to the engine will be inducing some sort of
opposite signal to ground – maybe affecting the CAS? Just some
WAG.
I have no filter on
the alternator output – never saw more than the normal ripple there. Could a
failed diode in the output converter be making big spikes? Wait – even with
alt off, no start? What’s in the ‘alt filter’ besides a capacitor? Is the
filter in the circuit when you run the external wire?
Just a few random
thoughts – I expect you know more than I about how this stuff
works.
Al
G
-----Original
Message-----
From: Rotary
motors in aircraft [mailto:flyrotary@lancaironline.net] On Behalf Of Joe Ewen
Sent: Friday, January 15, 2010 3:13
PM
To: Rotary motors in
aircraft
Subject: [FlyRotary]
Alternator Filter
I have been having problems with
my engine installation. Starts, runs great - turn on the alternator and
the engine starts to miss. Put a scope on R1 ignition trigger lead and
sure enough missing trigger pulses. Turn alternator off no more
missing pulses. Looked at the alternator output with the scope and as
expected saw some spikes. Bought an aircraft spruce alternator
filter. Ran the engine , worked fine until it would not prime or run any
more. That turned out to be a defective store switch. Replaced the
switch and back to normal. Here is the part I do not yet
understand.
Engine runs fine, added alternator
filter and can not even get a pop. Take the filter off, starts
fine. Filter back on (alt off), no joy. Filter on (alt on), no
joy. Filter off, runs fine. I can not imagine how a filter
capacitor is interfering with the ignition of the controller. If anyone
has any insight please feel to speak up.
I have 2 wiring ducts: 1. quiet -
signals such as to the PCM, ignition firing leads, low voltage sensor, analog
signals. 2. Noisy side, battery cables, ignition power, injector power
and signal (twisted pair shielded (shield properly
grounded.)
Alternator connects to battery
cable at the starter. If I run an external wire between the alternator
and the battery (outside the airframe, long wire - engine in back, battery in
front (canard)) everything runs fine.) Aesthetically this might not be a
good approach, not to mention the aerodynamic impact.
This problem has kicked my butt
this week working on it in the cool 20F evening air. Any insight is
appreciated, especially why a capacitor (which does not conduct DC) would
render the EC2 controller inoperative would be
appreciated.
|