Bosch Motronic AFM => Point - Counterpoint

[Land Shark]

At 07:01 AM 6/26/96 -0600, you wrote:
>At the risk of continuing this discussion into the new millennium, I present
>this response from Pat Braden, author of the book "The Alfa Romeo
>Owners' Bible", in response to our recent discussion.

 With all due respect to Mr. Braden he has not a CLUE how Motronic works!!

>Hate to disagree. Presuming that the spring has any effect on air/fuel
>ratio, loosening it (that is, allowing more air to enter) would _lean_ the
>mixture, not richen it. Conceivably, weakening the spring could allow the
>flapper valve to swing further for a given air velocity, indicating a
>higher volume of air, for which the control module would immediately
>provide a richer mixture. However, the mixture would immediately be
>corrected to the stoichiometric value of about 13.4:1 by the control
>module, acting on information from the Oxygen sensor.

 Really??

 Well, #1, stoich. for GASOLINE is arounf 14.68:1 (14.7:1 is the common number)

 Second, in for example .. a porsche 911 :),  the oxygen sensor is only active
 in the area necessary to pass the emissions test (FTP - Federal Test Procedure)

 This is done with a WINDOW of speed and load .. any speed (rpm) value greater
 than X .. or any load (Tl) value greater than Y for Z seconds will cause the
 system to run open loop .. as does ..

 1) WOT
 2) Accel
 3) Decel fuel cut off
 4) Warmup cycle

 So in ALL these cases the EGO feedback is OFF, and the car acts as if it had
 none, with the EGO correction in the DME defaulting to 0x8000 which is "one"

 Now on to the effect of messing with the spring .. if you loosen the spring
 you will INDEED richen the mixture .. the problem is that while the output
 function of the meter is a simple Log (base E) function of the air flow,
 that function is comprised of the COMPLEX functions of the individual 
 components in the DME ...

 Which are:

 1) The flap
 2) The spring
 3) The cailbrated pot trace 

 Loosening the spring DOES richen things up ALL over the scale, but also
 CHANGES the function of the curve.. i.e. it's SHAPE ...

 Since the average joe cannot then tweak the internal Motronic tables to 
 correct this, you start to generate a FALSE load signal in the DME..

 This is not good, because the relationship of the whole thing is mathematical
 and now you have broken it!

 Those relationships (for a flapper) are ..

 Q=f(Uv/Up)   (fcn is a log function scaled by 2 constants)

 Tl=Q/(n*Ki)

 Tl or LOAD is the base for the injection time .. 

 Thou shouldst not mess with T sub ell ...

 :)

>My understanding of the system is that the spring is there only to return
>the flapper and act as a damper on its actions. Weakening the spring
>reduces the damping effect, allowing a faster throttle response but not
>changing the A/F ratio.

 No, loosening the spring richens the mixture, on accell we add even MORE
 fuel and get better transitions .. we can do the SAME thing by bumping
 up the main (non idle, non WOT) fuel table a bit (usually 3-7%)..

 During EGO, there is no extra emissions, but upon accel, you get more
 fuel for better transitions ..

>With any Motronic system, presuming it is in working order, nothing you
>do to the mixture will change the actual air/fuel ratio because the Oxygen
>sensor signal will simply force the control module to correct the mixture
>strength to stoichiometery. 

 I wont even start talking about skewing the AFR from Lambda=1 by modifying
 the constants in the EGO feedpack PID control loop .. you can make a normal
 EGO sensor control things either a bit lean or rich of center this way!

>With the Oxygen sensor disabled, the easiest
>way to get a richer mixture is to fool the coolant temperature sensor  by
>putting resistance in line. 

 Actually on most DME's the EASIEST way is to simply adjust the FQS which
 is on most DME's from 88 and earlier years .. and 8 position (3 bit binary)
 encoder which changes fueling and spark base parameters ..

 Jim

 PS: Sorry to pontificate, but I've just had to UNDO too many messes that were
     prompted by "experts" ... most people have no IDEA how Motronic DME works

     They try to make it complicated when in fact it is a SIMPLE deal ..

     I can always show you commented source code if you are from Missouri ;)