Re: MB engine overboost.

[Robert Houk - SMCC Bos Desktop Hardware <Robert.Houk@East.Sun.COM>]

   >  All the symptoms of a sticking waste gate? Very high boost followed by
   >  fuel pump shut down and then everything is back to normal.

   No, we don't think so.  As far as I know, the only professional who's
   ever experienced one 'live' was Martin at BR Motorsport, who said he'd
   never felt anything like it.  If it was something like a sticky
   wastegate, we'd see it on other models - especially MC-2/1Bs.

   The problem running round in my mind is: Where is all the fuel coming
   from?  Other experiments are starting to suggest that an MB is close
   to the fuel system's limits when only gently tweaked - how is the
   fuel for this massive surge being supplied?

   The only idea that seems to make any sense so far is that the cold start
   valve is being triggered excessively.  The MAC12 logic (like its close
   relatives) provides for triggering the cold start valve if the engine is
   accelerated hard when cold, where 'cold' is indeterminate.  We haven't
   investigated this condition much, because none of us likes thrashing a
   cold engine.

   The reason the cold start valve makes so much sense is that it's
   independently fed by the fuel pump - it's attached to the main feed
   line at the fuel metering head, but it isn't in any way _metered_
   by the head, and so it represents additional fuel throughput.  The MB
   has a tendency to run lean at full throttle (especially with aged
   fuel system components) so dumping fuel into the inlet manifold would
   have a proportionately greater effect than one might expect.

I'll contribute a few random thoughts, since I'm not getting any real
work done anyways.

It "can't" be the cold-start injector. All that can do is to totally
ruin your mixture (if you're runing way lean, it may "fix" it...) by
going over-rich. Assuming you're anywhere near "normal" (as one other
lister mentioned "2000 rpm part-throttle"), your mixture will go any-
where from pleasantly rich to drowning the engine, depending on how
much air you have, and how much fuel is being dumped into it. It might
be worth a few-hundred RPM on the top-end; remember the CSI is designed
to run "wide open" at starting-speeds [less-than-idle RPM!] without
overtly flooding the engine (although it can easily flood a warm engine).
It just cannot contribute "100+" horsepower. (disclaimer: experience
based on older I-5 engines...).

If you are truly getting "overboost" (has anyone actually looked at
a manifold pressure guage, and gotten a reading in that brief instant
between the dreaded unintended accelleration and head slaming into
the windshield on fuel-cutoff?), there is only *one* mechanism that can
account for this -- the wastegate is not properly bypassing the turbo.
[I tacitly assume you don't have a NOS-injection system that might be
spontaneously triggering itself!]

The wastegate is a purely mechanical beastie, subject to getting dirty
and occasionally "hanging up on the crud", so it's easy to conceive of
the sporadic overboost from this mechanism.

In addition, if you have a computer-controlled wastegate, then of
course you have a whole 'nuther kettle of worms. The computer is free
to do all sorts of interesting things, like overboost and not care, then
quickly change its mind... one "obvious" scenario that pops into this
cynical mind is that the ECU has selected transient overboost (you've
just stomped on it, and it's going to give you 10 seconds of a few
extra PSIies, as well as running "rich" to help, you know, cool the
exhaust valves), then gets confused for a moment [anyone ever see their
PC/workstation "just freeze up" for a moment or two, then recover and
continue as if nothing had happened?] as the engine "comes on the cam"
[starts developing some real power...], then regains its sanity and
notices that the manifold pressure is now approaching 3 atmospheres
absolute and shuts things off in a panic.

If you have (and I think the car(s) you're all talking about do...)
ECU-controlled overboost, and you have a reproducible case (which it
also sounds like), then the "obvious" experiment is to simply discon-
nect the ECU from the wastegate, leaving it purely mechanical (this
may involve disconnecting and/or blocking various hoses and/or un-
plugging other little goodies; left as an exercise for the reader).
If the problem goes away, it's the ECU or the valve used to control
the wastegate. Reconnect, and monitor the duty cycle (or whatever
mechanism the ECU uses to control the wastegate control valve), and
see what it does as you overboost.

I'm inclined to think it *must* be the ECU, since normally the ECU
will shut off the engine *long* before it builds up that dramatic
of an overboost (i.e., the ECU must be distracted from its normal
boost-monitoring operation in order to allow a build up of "300 hp"
of overboost -- it just can't happen any other "single point of
failure" way; now if its one of those dreaded "multiple points of
failure" problems, good luck!). (I'm also assuming more-or-less
stock ECUs, no "shutoff disabled" or otherwise hacked-up code...)

Some "real numbers" would be good to see (actual manifold pressure
readings; wastegate control duty cycle values; O2-sensor readings;
etc., and so forth)...

Just "thought experiments" on my part; the only overboost my car
has ever seen has been my fault, more or less.

					-RDH