Sensor Questions?

[SRavet at bangate.compaq.com]

Markus Strobl <eusmsrt at exu.ericsson.se> Wrote:
| 
| The Q of MAF vs speed density has come up in 2 mailing lists at once,
| so this goes out to both lists...
| 
| 
| > Interesting, Scot.  It doesn't seem that there is really a need for a 
MAF, 
| > then.  Is MAF just an aid to enable the CPU to learn better?  Or is 
there a 
| > MAF term in the fuel equation?
| > 
| > --steve
| 
| I like to think about it this way (it is a bit simplified, but it's
| basically true):
| 
| A speed density (SD) system will only know how much air is entering the
| engine as a percentage of the maximum flow. For instance, driving at half 

| throttle at a certain condition may tell the ECM the engine is using 50%
| of maximum flow. Dyno test run by the manufacturer indicate that max
| flow is 600 cfm (cubic feet per minute) of air. The ECM then knows how
| much fuel to inject for 300 cfm of air. 
| 
| This sounds like a perfect setup, until you start to think about it. 
| Production tolerances will affect max flow in an engine (casting flash,
| mismatched cylinderheads-to-manifold, etc). Also, max flow will change
| as the engine ages. And we haven't even started considering modifications
| such as ported heads, hotter cam etc.
| 
| So as soon mods are made to a SD engine, a new chip has to be burnt. 

But if an O2 sensor is available, the SD tables can be updated as the ECM 
learns lean or rich conditions.  That is the BLM (block learn mode) that 
Scott Sealander referred to.  BLM is basically a percentage that the pulse 
width is modified by.  IF you add headers, BLM increases pulse width across 
the board, and this is still used even in open loop mode.  So an SD system 
is capable of learning about modifications.



Steve Ravet
sravet at bangate.compaq.com
Baby you're a genius when it comes to cooking up some chili sauce...