When sorting out some increased knock activity that started to appear during these hot summer temps - I came across folloving behaviour. It has been already discussed that when doing wot pulls - the first run log is usually garbage. I have a perfect example of this. Did 4 runs (3rd gear) immediatelly after each other (all four done within 1 minute).
From ugly 7 knocks (1st run) to zero (in 3rd run) was a satisfying result. Later when analysing the logs in more details - I've noticed that the logged AFRMAP value (as well as the real Wideband AFR) has been not consistently the same for the same RPM,Load etc.
The only factor that was showing some variance between logs was the average coolant temperature (80 degC to 83 degC - this is the normal operating temp).
Is this small variance such important that the AFRMAP value dropped almost by 0.9 point (similarly richer was the real AFR measured with WB)?
Otherwise I do not find a reasonable clarification. Do you see similar behaviour related to coolant temp - or am I missing some other variable which affects this?

I have attached a table which sort of shows it nicelly - each run next to each other.

The car is EDM Evo 8 with TBE, Cone filter and Ecuflashed (all the temp compensation tables are stock) - ROM 96530006

 

Only had a quick look, but you may want to repeat it with lean spool disabled (set enable and disable RPM to say 1500 or 2000 RPM) and the fuel map suitably richened. I've not done this personally, but you may find that the first run's AFRMAP and WBAFR are richer and it doesn't knock?

 

thanks for the tip - I can try but as all four runs were done in the 3rd gear - I do not expect that lean spool was the culprit. I rather expect some coolant temp compensation table that is not discovered yet

 

There are other factors that affect the tune on a car. AIT and Coolant temp will affect the tune on a fuel injected car. The ECU will adjust fuel and timing slightly. The best way to tune the car is always tune at the same coolant temp.

 

3c change in coolant temp will make no difference, at least from what I have seen.

When I logged our time attack car in Utah, there was huge coolant (180-220F) and intake temp (100 to 120F) swings and the AFR and timing was pretty constant.

Lean spool was off.

Barometric compensation did kick in as the car was tuned at almost sea level and Miller Motorsports Park in Utah was 4300 ft. AFR leaned out .3 from the differences, but once I brought it back to how I wanted it it was constant for the weekend.

Here is a log from that weekend. Notice the large difference in AIT and Coolant temp from the start of the log to the end. Now look at AFR which is the red plot towards the bottom. You could put a ruler up to it.

 

It does look flat but to be fair, the scaling is pretty compressed. If you open up the y-axis a bit and re-plot, it may not be so flat. Or it might still be Nebraska-flat.

 

Here it is zoomed in:



Still constant WOT to WOT.

 

Nice. Still Nebraska. Looking good.

 

On my logs (and summary table) - boost, load, AIT were almost the same. Ignition was more advanced in the runs without knock obviously - the only parameters with observed changes were Coolant temp (diff 3 degC), AFRMAP and Wideband AFR.

Will try with lean spool off - but I still think the explanation why the car went richer within on the same spots - within 1 minute between pull 1 andpull 4 - must be somewhere else.

Thanks for the log - seems that I have to figure out how to have exactly the same behaviour everytime I kick the throttle.

 

I would look at your cone filter and inlet hose. Make sure its secure and not hitting anything making the maf go crazy.

 

evo828 is sorta right! I was looking again this evening at the purported "accel enrichment" subroutine and realized that there is a coolant temperature compensation to IPW. There are two tables:



It can be seen though that there is no compensation after 77C, so its not the cause of the richer AFR that he's seeing.

Here are the 88590015 Evo 9 definitions if anyone want to tinker:

 

Yeah - I did - no apparent issues here. I have another series of runs (wot, 10 secs, wot, 10secs, wot - all in the same gear) where the story is the same - 1st run - leaner with knock and lower Coolant temp, 2nd and 3rd run richer (AFRMAP, WBAFR) and 2-3 degC higher coolant.

mrfred - thanks for the investigation - I will try to find these in my ROM - may be they are set differently. The question here is also, whether the coolant temp that I am logging (MUT10) is really the one referred in the comp table (you know we have multiple load variables - so we can have more temperature variables )

 

Lean spool seems to work on timers, not had time to work out the units properly, but regardless of whether the same or different gears, there does seem to be a leaner run the first time.

 

John - thank you - thats an interesting point - I will report the results soon.

 

MUT10 is definitely the correct coolant temp variable to log. In all the ROMs I've looked at (7s, 8s, 9s), there's only one temperature variable that's used for controlling engine functions. Per jcsbanks, I suggest making sure that lean spool is disabled before spending too much time trying to find those tables.