general info.

[Todd Knighton]

tom cloud wrote:
> 
> Now, it seems to me that a simple EFI system would start with a fixed
> injector pulse width.  Then increased RPM would give a corresponding
> increase in fuel.  RPM and TPS would also (each) give a relative
> indication of the expected operating needs of the engine, a "typical"
> correlation between the two being empirically determined.  Then
> a large disparity between RPM and TPS could assume a need for more
> fuel (TPS > RPM) or for less than the "normal" amount.  Engine
> temperature could be used to increase the pulse width at lower RPM
> for cold operation.
> 

YUP.

> Now comes feedback.  EGO supposedly tells when conditions are
> optimum for the most complete burning of the fuel (stoichiometric,
> or A/F of 14.7:1).

>  But, economy is achieved at about 18:1

Maybe in a diesel application, most 4 stroke gas motors won't tolerate
much over 15:1 without lean misfire, unless your HONDA with a CVCC.

> while maximum power is found at about 11:1.

We don't even get there at 20 psi on an air cooled Porsche Motor, Max
power is still at 11.6:1 or so.  Normally Aspirated Max power can
usually be found from 12.8:1 to 13.3:1

>  From recent discussions in this group, it appears that the EGO 
> sensor is useless at determining anything but 14:7 -- is this right?  > And even then how, specifically, is it's output used?

They do work typically from 23.5 to 11:1, at least a Bosch 4 wire sensor
does, typical 1 or 3 wire sensors are not very good beyond 12 to 17:1

>  I have an O2 sensor connected to a DPM and use
> it to tune my EFI system.  The writings that I have seen say that
> its output swings back and forth between a high level and a low one,
> with an average of about .5 volt being sought.  My DPM shows no
> pulsating.

This in on cars that are being controlled, not very well, by a
Microcontroller.

>  Can the EGO sensor's output be filtered to average or
> integrate the pulses and provide a steady control signal?

Yup.

> 
> Other sensors that I know of are the MAP, BAP and MAF.  I can see how
> a reduced manifold pressure would imply an increased need for fuel and
> would modify the previous RPM / TPS determined pulse width, but how
> does BAP (barometric atmos. press.) modify that, and is it necessary?

To maintain specific air/fuels at specific gauge pressures typically on
emmissions control type stuff.
> 
> Apparently BAP is not needed if MAF is used.  I can see that, with
> MAF, one could actually determine the amount of air entering the
> engine and therefore more correctly ascertain the amount of fuel to
> add.  How hard is it to add?

Fairly easy to compute with a MCU

> How could it be added to a TBI system?

Depends on the system

> Why isn't manifold pressure and throttle position an adequate 
> indication of air intake volume?

It is until there is a change over time.  Pressure lags behind air flow
and thus has to be compensated for.  But with throttle position that can
be calc'd.

> In fact, since MAF sensors seem to be a pain,
> it would seem to me that air volume could be determined quite well
> from knowing BAP (the available air density), TPS (the size of the
> opening the air is trying to get through, and MAP (the amount of
> pressure on the opening).

YUP.

> 
> Then there's my question about EGR.  This is not a sensor, but it
> has a marked effect on engine operation.  It seems to give a higher
> effective octane thereby allowing the timing curve to be more
> advanced than without it.  What that means is that anyone monkeying
> with the EGR will end up with pinging at rpm under load and a loss
> in power.  This is a particularly onerous problem for anyone
> trying to increase engine performance with aftermarket products --
> specifically intake manifolds or, as in my case, a carburetor-to-
> fuel injection upgrade.

Unfortunately EGR also decreases the effective oxygen in the combustion
chamber, thus decreasing power output.  IF you removed the EGR and
recalibrated the timing, an overal net increase in power would be the
result.

> 
> I would like to know if anyone has any control algorithms (digital
> or analog) to share, as studying what you've discovered will help
> me.  I would like to know any correlations between sensor readings
> and fuel control and I need sources of engineering specs. on the various
> sensors.
> 
Pressure Sensing: PV=nRT basically sums it up
MAF: Pain to calc.  Need to air flow by rpms to generate a load value. 
It almost reverts back to pressure sensing.  Low rpms have poor
resolution on large MAF sensors due to relative low change in air flow
from small to large throttle openings.
AFM: Just like MAF though measures CFM not Mass, needs barometric and
temperature compensation.
TPS: Crude but effective, recommended only when other methods won't work
well, i.e. large cams, independent throttle bodies.

Motorola is a good source for specs and pressure sensors.
Siemens has some good temp stuff.
Bosch makes good TPS sensors


> That pretty well sums my rather large ignorance of the subject.
> 
> Can anyone help?
> 
> Thanks
> Tom Cloud