PDA

View Full Version : Burst AFR ?



DrX
October 14th, 2006, 04:45 PM
Is there a way to reduce the lag(.5-1.0sec according to frame counts) between Commanded and Actual AFR at the onset of WOT. Boost comes on quickly and AFR actually goes leaner prior to dropping. I tried switching back to PE to enable earlier enrichment at WOT rather than just using B3647(custom OS) but it didn't seem to make a diffference. Doesn't matter what I do with the transition portions of the VE tables. BENs are still always lean in this area.


http://i114.photobucket.com/albums/n241/TDOC2004/DashLog.jpg

dc_justin
October 16th, 2006, 03:30 AM
Are you running in speed density? If so, I have the same issue with very rapid throttle application. I figure it just has to do with the way SD fueling works.
SD goes lean at initial stab of the throttle and rich when you quickly pop off of it. MAF-only mode goes rich at initial stab and lean when you pop off...

DrX
October 16th, 2006, 06:31 AM
Yes, SD. Looks like filtering the rapid MAP rise might help clean up the AutoVE process.

DrX
November 1st, 2006, 11:35 AM
Got the lean blip/bump down to .2-.5 sec on the logs now. Doesn't really seem to effect anything anyway. Never any KR at that point.

Filtering the rapid MAP increase also helped get more sane AutoVE results than just filtering throttle transitions. Initial boost onset is very quick with the Radix followed by a more gradual rise that is not filterd out.

joecar
November 1st, 2006, 12:15 PM
You have the PE delay(s) zeroed...?

DrX
November 1st, 2006, 02:59 PM
B3608 is zeroed. PE MAP threshold(B3613) was set to 65kPA which sometimes isn't hit until .05 sec after the throttle starts to open(i.e. somewhere within that first .1 sec frame) so I bumped that down a little more.

edit: Mind you, the same goes for TP as an enabler. It occurs somewhere in that .1 sec after the throttle starts to open. So I don't know if the delay in actual AFR catching up to commanded AFR can be reduced to any less than .5 sec. Also not sure what the lag is between commanded AFR and when the wideband sees the change. That little residual BEN bump could just be an insignificant artifact. It doesn't cause any KR, so who knows.

dc_justin
November 19th, 2006, 06:28 PM
edit: Mind you, the same goes for TP as an enabler. It occurs somewhere in that .1 sec after the throttle starts to open. So I don't know if the delay in actual AFR catching up to commanded AFR can be reduced to any less than .5 sec. Also not sure what the lag is between commanded AFR and when the wideband sees the change. That little residual BEN bump could just be an insignificant artifact. It doesn't cause any KR, so who knows.

Just bringing this back from the dead.

I'm toying with the idea that it could possibly be that the MAP sensor doesn't register the change in pressure nearly as quickly as it actually is occuring in the manifold. You're fortunate in that you don't get any KR in these situations. A 30-80% stab at 50mph in my truck will result in a spike up above 16:1 when commanding 11.9ish and have BIG KR.

A 40% throttle stab from a dead stop resulted in a 40* timing drop (Burst Knock Retard), and STILL managed to trigger a 6* KR spike. With this 91 octance CA gas, this is something that is always near and dear to my heart.

joecar
November 20th, 2006, 06:14 AM
...With this 91 octance CA gas, this is something that is always near and dear to my heart.I hear you mate.

DrX
November 21st, 2006, 05:46 PM
91 is the best around here too unless you go to a 10% ethanol blend that is only sold by one retail chain. Then premium goes up to 94. Soon all of our fuel will be required by law to contain an "average" of 10% ethanol.

I have Burst KR disabled.

Seems that there are several potential lag issues. IAT, MAP, WBO2 sensor, etc.

I'll follow up on this once I get around to testing the modified LS2 90mm TB.