(Michael) wrote in message . com...

Over the years I've been involved in general aviation, I've had
first-hand knowledge of quite a few accidents and incidents that
eventually wound up in the NTSB database (and some that should have
but did not). Reading the report after the fact, I find that
inaccuracies are the norm. In fact, some of them read like a work of
fiction.

I couldn't agree more. After reading the NTSB report on my
accident, I no longer read the reports and think "what was that guy
thinking?" afterwards. It's safer to assume that it didn't happen the
way the report said it did.

My investigation was actually done by airframe and powerplant reps
working for the NTSB (no FAA guys involved). They were extremely
thorough and went the extra mile during the investigation. While
initially, they were leaning towards the ubiquitous "failure to detect
carb ice", which they tend to do if they can't find other evidence for
an engine failure, they kept at it and eventually found that my carb
had actually come apart in flight. The FAA was fine with that and
took no action. When the report came out, it bore little resemblance
to anything that I had told the investigators, or anything that the
investigators had told me or the FAA.

Ultimately, my report was a victim of politics. At the time, the
NTSB was pushing the FAA to issue an AD on the two-piece venturi in
most Marvel-Schebler carbs. The FAA wasn't ready to do that. Someone
up the NTSB chain took my report and magically changed the cause of
the failure to support the proposed AD that the NTSB was pushing for.
I, the FAA rep, and the NTSB reps all know that there was no venturi
failure in my carb. When the bottom half of the carb detaches itself
from the top half of the carb, they type of venturi doesn't make much
difference.

While it is often touted that the NTSB is a totally impartial and
independent body, it's also true that they have their own agendas.
Independence alone doesn't make them impartial. They're still part of
a bureacracy.

John Galban=====N4BQ (PA28-180)