I posted this question on the 996 forum, but didn't get very clear responses. I hope I'm not violating site rules by reposting here.

I built ECUs from scratch in a prior lifetime, but am totally unfamiliar with the 996 ECU control strategy. Can someone tell me the limits of the factory computer with respect to boost control? Can boost be made to vary with RPM? If not, do any aftermarket tuners do it?

The reason I ask is that I've always felt that constant boost left some power curve on the table, and tuning for constant cylinder charge, subject to detonation limits, was the way to go. So, for example, if the volumetric efficiency without the turbo peaked at 4000 at 90%, and was 80% at 3000 rpm, why couldn't you run 90/80 = 1.125 times the manifold pressure at 3000 that you do at 4000 (subject to adjustments for added temperature rise, etc.). To be overly clear. If you can run 1.3 bar at the intake at 4000 rpm in the scenario above, why is it any more stress (other than compression heating and intercooler inefficiency) to run 1.125*1.3= 1.4625 bar at 3000 rpm? So, basically, the strategy is to run max constant cylinder charge unless you run against the detonation limit, or the turbo runs out of ability to reach that pressure. Basically, this should show up as constant torque curve over that range.

Since these are drive by wire cars, I guess another option would be to vary throttle position to map intake manifold pressure versus RPM to generate constant cylinder charge/torque.

Is this common practice? Why or why not? Thoughts?

 

The way it was explained to me is the Turbo's DME attempts to satisfy the torque demand of the driver as signaled by the e-Gas and thus the DME will allow as much boost as necessary, even going over the nominal 0.7 bar threshold (for the standard Turbo), while staying within whatever operating limits or constraints there are. These constraints probably include a maximum boost to limit the maximum of torque the drive train is subjected to. The clutch is probably the weakest link and while quite able to withstand the output from a stock Turbo if this suddenly went up the clutch would not withstand this as well.


Based on my observation boost can obtain 0.7 bar (or 0.8 bar even 0.9 bar) at its peak, and this level of boost is sustained long enough to know it is not a "glitch", but boost invariably falls off as RPMs climb.


I do not have any data logs of this nor video clips so I can't provide you with any real numbers or curves but at some point as RPMs climb over 4K and go higher whatever max. boost was the boost has now dropped to 0.4 bar or 0.3 bar. As best I can remember the boost doesn't drop off to 0.0 bar as RPMs approach redline but by this time the car has reached a velocity that requires my full attention in aiming the thing. In lower gears RPMs climb so quickly there's no time to observe and contemplate it all.


As boost level can be, pardon the expression, boosted over nominal it can also be restricted to less than nominal if for instance the DME detects detonation via the knock sensors or there is some other input that has the DME backing off on boost.


You might pick up some interesting info by seeking out posts by Turbo owners who have posted charts/graphs of their engines' telemetry with aftermarket DME firmware installed. Generally all I recall seeing is hp/torque vs. RPM though so you might search in vain.


But you also might visit some tuner sites and see if any have posted charts/graphs with more than just hp/torque vs. RPM. I'm thinking if you can see boost too this would perhaps shed some light on the subject.