[s-cars] Fuzz logic vs adaptive learning

[Mihnea Cotet]

Yup, I fully agree with you from this standpoint Feico...

Programming the Motronic to allow 90% of full boost at 30-40% TPS and/or 
90% WGFV duty cycle works, and a stiffer WG spring as well actually, I've 
BTDT several times.

What you don't want however, is too aggressive timing settings/too lean 
fuel settings at part throttle, exactly where boost wasn't present before, 
because otherwise it might end up not running properly.

Another thing to consider when doing so is the factory cruise control 
system, I've also BTDT in some cases where it couldn't keep a constant 
speed anymore, because too much boost was generated at too low throttle 
openings, but if care is taken in programming the WGFV map at those 
particular points where the cruise control is active, it will work ok IMO/IME.


HTH,

Mihnea

At 12:03 26/01/2004 -0500, FvAMI at aol.com wrote:
>You would need to have a Normally Open solenoid to do this.
>
>The other option I once thought of was to use the stock solenoid invert 
>the signal to the solenoid and/or connect the V+ to GND and confirm the 
>Motronics had the ability to continuous source the current to the solenoid 
>(It sinks the current) ... the only problem is the solenoid would be 
>active for 99% of the time and you would substantially decrease the life 
>of the solenoid ... in addition the Motronics unit would be sourcing in 
>excess of 1 amp to the solenoid almost continuously.
>
>The older Audis may have a normally open solenoid. Personally I felt it 
>would be better to do an analysis first on exactly the response time of 
>the standard system. Some more aggressive boost programming may be able to 
>compensate for the response time ... i.e. program the Motronics unit to be 
>at 80% - 90% maximum boost at lower throttle settings.
>
>Best regards,
>Feico van der Laan