EFI

[Land Shark]

FORGIVE ME ALL .. I just HAD to answer this one .. a bit late due to the
holiday, but REPLY I MUST .. 

At 16:28 7/6/96 -0700, you wrote:
>Feel free to share with the EFI DIY list. Herewith the opinion of Don
>(ex-Alfa, now Fiat R&D):

 Ok .. well .. IMHO .. It's time for Fiat to hire me then..

>The EFI discussion is difficult: one must first know if the system is
>analog or digital and to which degree of autoadaptivity is provided.
>Example: Fiat uses/has used _many_ executions of Bosch Motronic alone,
>many L-Jet, Monotronic, Weber-Marelli, Rochester, Hitachi....

 We are talking about Bosch Motronic.. Digital Motor Electronics utilizing
 Multi-Point Fuel Injectors .. I have worked on Alfa, Ferrari, BMW, Porsche
 and VW systems .. my expertise is with mostly BMW .. others like Todd
 Knighton are the gurus at things like Porsches ..

>1. Forget that the closed loop O2 sensor function will correct, in the
>end, any minor changes upstream. It will, but that does not seem to be the
>issue. We talk open loop.

 Fine ..

>2. Spring tension makes a difference only during _transient_ movement and
>will enrichen until steady state, where it has no effect. There is no
>effect at WOT as you are already using the longest pulse found in the
>look-up table (MAP) for that set of parameters.

 Umm .. Motronic does NOT look up PULSE WIDTHS .. 

 It MEASURES air .. calculates LOAD from AIR and RPM and a constant (based
 upon injector size) then corrects for operation conditions (maps)
 to yeild injector PW ..

 Changing the "air" measurement DOES indeed change the AFR ..

 Not to argue this to death, but I've HAD cars on the 5 gas and SEEN it 
 happen when I had to UNDO the "tweaking" that someone had done to the
 "spring" in the AFM ..

>3. Remember, electrons are faster than air!

 And ..

>4. The base calibration is not always based on vane position: depends on
>system.

 On EVERY Motronic I have disassembled .. the base cal is directly
 correlated to the AFM or AirMass sensor Voltage output ..

>5. Even open-loop, later generations use autoadaptivity, which further
>permits default corrections for components operating outside their spec.
>ranges, within the ROM map.

 I'd like to see this .. I have seen no adaption outside of closed loop
 functionality .. I'm ready to sign the Fiat NDA ..

>6. Aside from AFM vs inputs to controller, this signal is further
>corrected by rpm, load, temps, etc., always cross-checking table for
>correct fuel delivery before the O2 sensor.

 Yup .. these are the "operating conditions" I spoke of above ..

 Although LOAD is in fact the BASIS for the fuel injection and not
 a "correction" to it ..

>7. To see the effects, get a J.C. Whitney A/F checker.

 NO WAY! .. all that will tell you is ..

 RICH or LEAN .. and that is that ... U need either a UEGO setup
 or a BOSCH LA-2 system

>8. In systems with separate ignition controllers, more benefits can be had
>by changing the spark map.

 Why not in systems with integrated spark also .. maps are easy to tweak!

>9. Significant changes cannot be made by DIY, nor without a dyno.

 UNTRUE .. I have brought BMW M3's from 240 to 278HP with the Torque
 raised from 225 to 252 lb/ft at peak .. this is VERY significant
 and only a bit of that tuning was EVER done on a dyno ..

>10. Perceptions and stop-watch times do not always concur.

 No, but a state of the art DynoJet dyno never lies!

>11. There are other ways of increasing fuel delivery such as raising rail
>pressure, larger nozzles, with larger AFM. But the intake runners and
>valves have to be sized accordingly.

 All KLUDGES, it is much easier to apply a complex mathematical function
 to the existing fuel maps ..

>12. Our experiences with DIY modified EFI systems do not yield measurable
>differences.

 Well, Fiat .. loan me a car, and I'll tune it tighter than you've ever
 seen one tuned .. 

 Jim Conforti