Controlling Turbo boost

[rdh@sli.com (Robert D. Houk)]

Why is nothing ever simple?

I've been trying to get a handle on how to control/modify the boost from
the turbo. "Stock" (if you can call an S4 turbo on an UrQ stock; well,
the wastegate is stock), the boost is nominally 10 to 11 PSI, somewhat
adjustable at the wastegate (on my '83, anyways; I gather some later cars
are not adjustable at all). With it nominally set to 10 PSI (less than
it can comfortably make, but necessary to keep it from building up too
fast for the stock ECU's peace of mind! [Now fixed]), I see varying
(from day to day) results - sometimes at 4000 to 5000 RPM I see 10-11
PSI, othertimes, 12-13 PSI, fairly consistently (i.e., when it wants to
run 10-11PSI, it does it time after time, but a day or three later, it
may be running 12-13 PSI). It seems to gradually build up after 3000
rpm. I've come to the conclusion that the wastegate might be "sticking"
somehow, and not fully opening, allowing gradual boost buildup at the
higher RPM levels. Maybe 115,000 miles of carbon buildup somewhere...
does this sound reasonable? [Time to pull the wastegate apart and check
it out, I guess].

Bleeding manifold boost to the wastegate to shift its opening point to
a higher point works quite nicely, within the constraints that, as per
above, in order to keep below 16 PSI at higher RPM levels, I can't use
over 3 PSI "adjustment", which translates to only 10-12 PSI at lower
RPM levels (like 2500). I know that the S4 turbo is capable much more
than this. By bleeding manifold directly to wastegate -- essentially
locking the wastegate closed -- the turbo quickly exceeds the 16PSI
cutoff point of my ECU, so I am loosing 4-5PSI at the lower RPM levels
due to mechanical recalcitrancy, so to speak.  Further, carefully
monitoring the air pressure regulator(s), I see other aberrant behavior.
One is thermally sensitive, varying its regulation point by a coupla
PSI from cool morning to hot-afternoon-sun-heatup; the other (that I
have tried) is much better behaved, thermally, but it still has a lit-
tle weird behaviour. There is a settling period as the manifold boost
builds before the air pressure regulator has enough pressure "excess"
that it can reliably control the output. This only takes a "second" or
so, but I can clearly see the air pressure output pretty much track 
the input pressure then back off as the regulator asserts control. Set
at "4" PSI output, the pressure initially rises to about 5 then backs
off to 4 again as the manifold pressure rises to 10-15 PSI. In higher
gears (3 or higher) this is not a problem, but in 1st and 2nd gears,
the RPM/Boost build too fast and can cause the ECU to overboost-cutout
before the air pressure regulator as settled and stabilized at it's
lower-than-peak holding point.

I have concluded that I will have to go to some sort of electronic
boost controller. So . . . anyone out there have any experiences one
way or the other with any of the aftermarket electronic boost control-
lers? From calling around, I have found two primary models, the "EVC"
by HKS, and one called "Speed Graphic" by <don't remember>. The Speed
Grpahic is apparently a "static" (non-reactive, you set it and it
maintains that setting statically, like the purely mechanical air pres-
sure regulators. It's claim to fame however is that it has not one but
a bunch (dunno how many; 4 to 6 I think) of setpoints, indexed by en-
gine speed. This allows one to run "higher" boost at low RPM, backing
off as engine speed increases. But it doesn't allow at all for varying
factors (such as altitude, temperature, sticking wastegates). The EVC
unit on the other hand has one setpoint, but it is a dynamic unit that
maintains the nominal absolute manifold pressure adjusting for altitude,
etc. Both units therefore should allow full Boost at lowest possible
RPM via applying full manifold pressure to the wastegate and not backing
off until they need to.

Anyone know of any others? (I still haven't gotten replies from a couple
of the phone calls I've made)

Anyone have any other suggestions on Boost control?

					-RDH