Base Freq. for water injection

Orin Eman orin at wolfenet.com
Wed Nov 26 08:35:51 GMT 1997


> >Has anyone ever done/thought of using a knock sensor to control boost ?
> >ie keep the boost rising untill it knocks then bring it back a bit.....this
> would take into 
> >account air temp, altitude etc and give you max boost for the conditions.......
> >
> It's what my SAAB Turbo does according to the user manual.  Real life
> experience shows that the grade of gasoline does indeed affect both power
> and knock.  ie: Knock happens sooner under har accel. with lower octane
> gasoline and the car is also not as zippy.  I imagine the waste gate opens
> sooner too but the boost gauge isn't precise enough to tell the difference.

>From decoding the Audi turbo ECU, it will try to control boost as a result
of knock, but only as a last resort.  It retards the timing first
and only after it has switched from a premium fuel ignition map to
a regular fuel map _and_ has maximum retard does it cut boost...
or try to.  The computer control is only +/- 0.1 bar or so.
Those who put stiffer springs in the wastegate don't get this
benefit!

I would think that the reason for retarding first is that you
can retard the _next_ cylinder, whereas it will take a _lot_
longer for any boost control to take effect... it isn't good
to wait for boost to go down while you are knocking!

Orin.



More information about the Diy_efi mailing list