Re: CPU controlled boost, 5000CS TQ boost

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

   It would appear from these tests that the CPU controlled WG solenoid can =
   help the boost rise a
   little quicker=20

   I do not understand how this could be true. So long as a 100% mechanical =
   WG's spring is holding the WG 100% closed, how could the CPU-controlled =
   WG 'help the boost rise a little quicker'?

Two ways (actually, the "two" I was thinking of are variations on the same
theme...). The spring is a [reasonably] linear device. As such, assuming a
"10 PSI" spring, it does not stay closed 'till 10PSI, then suddenly open
fully; rather at 5 PSI it is already partially open (maybe 25%??? wild
guess on my part), thus already dampening the engine's enthusiam for making
more boost. With a "computer" controlling the wastegate (e.g., through 
either a "freq valve" feeding positive manifold pressure to open the
waste gate -- Audi factory control methodology, as least on the one S4 I've
seen, and the "5000 Turbo" documentation I've read -- or feeding same
pressure to back-bias the waste gate and hold it closed -- the typical
SchrappelKnobben as discussed on this list, and the way I've currently
got my BoostGraphic controlling the waste gate on my UrQ [which, by
the by, is "backwards" from how SuperChips says to hook it up, but it
won't flow enough gas to control it the "stock" way]), the computer
can "lock the waste gate closed" until nearly 100% of boost has been
achieved (i.e., at 5/6/7/8 PSI, the waste will be held fully closed),
before "switching on" and activating the mechanical spring to open
the waste gate. This will "defer" the wastegate actuation substantially,
and "speed up" the boost rise correspondingly, getting you up to speed
that much sooner.

One [dangerous] way to see this in action would be to simply disconnect
your waste gate (mechanically, or disconnect/block the manifold pressure
fitting, or feed full manifold pressure into the atmosperic reference
port [which I suspect is the easiest]) and watch your boost performance.

On my UrQ with the S4 turbo, the engine will eventually build 6-7PSI at
2000RPM, and holds around 14-15 at 2500. By directly connecting the
manifold pressure to the atmosperic reference port, effectively locking
the wastegate fully closed always, the engine slams into the [former]
16PSI overboost shutdown around 2250 or so RPM. 

   Also, if the CPU is dialing back the boost after 3-4 seconds and/or as a =
   function of RPM then this would indicate that a 100% mechanical WG =
   control would be better for performance Vs CPU WG control.

Not necessarily. There are two reasons why dialing back (or providing a
transient "overboost") is a win. One is if your stock cooling system can
not handle the heat dissipation of "full" boost on a steady state basis,
but can handle a "momentary" surge. E.g., your "system" is rated such
that it can handle 10PSI full throttle full speed "indefinitely". Odds
are real good that the same "system" can "safely" handle 15PSI for 10
seconds or so before getting "too hot". So a "dial-back" boost control
is a big win here (gets you up to speed faster).

The second is the RPM-sensitive dial-back. Again, from a heat-dissipa-
tion perspective, at 6000RPM you have to shed a lot more heat than at
3000RPM (I'm gonna guess at 60-70% more heat -- not 100% since the hot
gases are more quickly expelled and don't sink as much of their thermal
energy into the block but rather spit it out the exhaust). Therefore if
your "system" can handle 10PSI at 6000RPM "indefinitely", you can probably
handle substantially more boost at 3000RPM (solely wrt heat load). So
again a dial-back boost control will allow more power (more boost) at
lower RPM, getting you "up to speed" faster than a strict mechanical
system will allow. The other "variation" on the RPM-sensitive control
(and the espoused raison d'etre of the SuperChips 'Boost Graphic' box)
is that you have a lot more fuel available at the lower RPMs to take
advantage of more boost at lower RPMs. E.g,. if your "system" can safely
supply enough fuel (total gasoline flow to not lean out) for 10PSI (but
no more) at 6000RPM, then you clearly have a "surplus" of fuel available
at 3000RPM, so you can "safely" [wrt fuel supply and *not* leaning out
and burning exhaust valves, etc.] run substantially more boost at the
lower RPM levels, and can get up to speed that much faster.

[Note: in the above paragraphs, by "safely run more boost", I assume
that all the other limiting factors are still accounted for, such as
compression and octane vs detonation/knock, strength of the head gasket
and bolts, melting point of pistons, etc., and so forth!!! Obviousy,
if your pistons begin to deform under the combustion pressure generated
by 11PSI of air/fuel, then 10PSI is your limit, and running 15PSI can-
not be safely done, even if your fuel pump could pump that much gas and
your radiator can shed that much heat. Caveat Emptor and all that!]

   I certainly can see that the CPU does provide protection for a bad tank =
   of fool or possible mechanical failure and would not be for everyone.=20

   Would not the OEM 'knock' light provide some level of warning about bad =
   fool and excessive knock?

The whole point is that lighting the "bad fool detected" light doesn't
prevent the detected bad fool from keeping his damnfool foot to the floor
and destroying the engine. In any event, the computer [hardware/software]
can(*) "modulate" the engine and keep it consistently right "on the hairy
edge" of max performance *much* better than any mere wetware system can
hope to match.

(*) Note I don't say "will", but the potential is there.

					-RDH