WOT open loop/closed loop

[Eric Bryant]

> From: Ron Schroeder [mailto:rjs at bnl.gov]
> Subject: RE: WOT open loop/closed loop
> 
> So with a WB O2 sensor, you could go closed loop.
>

Well, it'd require some pretty significant code changes, I think.  That, and
the consequences of some sort of hardware failure at WOT would be a bit
severe.
 
> I was going to ask about the BLMs in WOT.  So maybe the ECM 
> looks at the O2
> only to be a fail safe to prevent running leaner than 14.7:1.
> 

The LT1 PCM isn't looking at the O2 sensor at all.  What it's doing is using
the closed-loop BLMs from the last cell used before power-enrichment mode is
entered.  But, it only uses those BLMs if they're above 128.  If the last
closed-loop BLM was under 128, the PCM automatically uses a BLM of 128 to
ensure that it always errs on the rich side (this way, some freak event that
cause a false low BLM at part-throttle won't cause lean operation at WOT). 

> So since part throttle uses BLMs (and the integrator), tuning 
> in that part
> of the maps will do very little unless the BLM values are WAY out but
> tuning the WOT part of the maps could bring out lots of 
> improvement.  Or am
> I off base?
>

For some reason, the LT1 PCM uses a VE table, despite the fact that it has a
MAF sensor.  Tuning the VE tables to keep closed-loop BLMs at or very near
to 128 seems very important with this motor.  If the closed-loop BLMs start
to creep away from 128, WOT fueling becomes difficult to predict (see my
above explaination), and the part-throttle drivibility and fuel economy also
seems to suffer.  I know that doesn't sound consistant with what closed-loop
operation is supposed to provide, but it's certainly the case with this
motor. 
 
> Does the GM ECM do any "learning" at WOT or interpolation 
> from near WOT BLM
> cells?

I hope I answered this question above, at least with regard to the LT1 PCM.
I'm not sure how other GM ECMs/PCMs deal with the BLM issue at WOT.

Eric Bryant
mailto:bryante at ghsp.com
http://www.bryantperformance.com  
----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org