I'm running the 10-map set posted here for the SMART system. There exists a set of conditions (slow gear change, rev'ed over 7000 in 1st) where my car trips the lean-run protect or overboost map during transition from first to second gear. Basically I need help reading what is going on and suggesting a solution.

Attached is an on road log which shows MAF in, MAF out, AFR and knock. Reading from left to right is the first gear run up to ~7300 RPM, followed by the gear change to second and finally the start of the pull in second gear. The right hand section of the plot is where you can see that the Xede has zero'ed out the MAF output due to a lean-run condition. This occurs just after the minimum RPM point in the shift as RPMs are just increasing. Note that boost is ~0.4-0.5 bar at this point.

Is this condition is a transient area where the ECU has legitimately cut fuel or if something else is wrong?

(Also attached in a spreadsheet of the run.)

 

Here is a run with the same maps where the fuel cut didn't occur.

 

Attached is a segment of the log which shows a strange results from the Xede. In the first two highlighted rows the Xede zero'ed out the MAF output while the measured load was less than the smart fuel switch map (inactive below 41). Thus it would appear that the Xede shouldn't have cut fuel based upon the logged data.

In multiple runs when this occurs, a short time later the AFR logged by the Xede goes lean and the Xede shuts off fuel. Or is it the Xede shuts off the fuel and the AFR is then reported lean? I can't tell from the data.

 

I have the same problem like you it happens sometimes on the pull when I go wot and then change the gear , so I loaded the base 93 oct.stage1+ map converted to SMART template till I get custom tuned .

 

Does the problem go away once you turn off leanrun protection (Zero out the table)?

Shiv

 

From the data I didn't know if a real lean run condition occurred at a load of 40 to 45, or if the Xede had pulled fuel and the result showed up in the wideband. That question kept me from zero'ing out the lean run map, especially given the load. What is your read on real or not? I did zero the over boost map and confirmed that the condition was unchanged.

I just went back to commslog and it doesn't appear that I can log anything else that would help in resolving what is occurring (e.g. log one of the user output variables). Also I can't test anything else until Tuesday.

Is it possible to replicate this on one of the vehicles available to the shop? I think that one needs to be on the road, not the dyno, and use the following sequence of events,

- WOT in first gear to >7000 RPM
- Shift up to second with zero or near zero throttle during the shift
- WOT in second gear

My experience across numerous trials is that when the car builds about 0.4 to 0.5 bar boost in second gear the fuel cut occurs.

 

First off disabling the lean run protection in the RPM region encountered on a WOT up shift does in fact resolve the problem. See attached.

That said if you consult the posted log you can see that there exists reported values where the lean run protection SHOULD NOT have been activated. This suggests that there is a V3.2 firmware problem or high speed logging is not fully accurate on a sample-by-sample basis. I'm going to assume the latter for now.

This issue also suggests that the simplistic means of detecting a lean run condition is not sufficient. The outlined software only scheme defined in another thread doesn't seem to be significantly different as to eliminate the possibility of false positives. Or did I miss something?
Why do I feel like I'm out here all alone with the SMART system...?

 

I don't understand what you said.

BTW, are you using launch control/WOT shifting?

Shiv

 

What I mean by this statement is best explained by the sequence of events necessary to produce the behavior.

First gear w/ WOT -> clutch in & off throttle -> shift to second quickly -> clutch out & WOT

Given that the shift occurs at >7000 RPM in first gear, the resulting second gear RPMs upon clutch engagement is ~4900 RPM. Hence zero'ing the region around 5000 RPM inhibits the false positive detection of a lean run condition.

If you look at the spreadsheet attached above (sorry for the quality but I had to keep the file size down), there are three points with the Xede zero'ed out the fuel. In the first and second cases, the load that is employed by the Xede is less than the SMART fuel switch map (e.g. < 41), thus one would assume that the lean run map wouldn't have been tripped. But in fact it was so either the Xede did something wrong or the high speed logging reported the wrong value for the load where the fuel was zero'ed.

Referring back to the spreadsheet, the Xede is triggering the lean run protection on a region where the engine is transitioning between a no load zone (mid-shift) to high load due to the up shift w/ throttle lift. But in fact a lean run condition really didn't occur in my opinion. It is natural to slow down a shift when shifting from red line versus shifting at lower RPMs. I posted an example above where the lean run protection didn't trigger. The difference is that the shift was quicker (shift occurs < 7000 RPM) and AFR didn't go as lean as measured by the LC-1. Hence the lean run protection didn't trigger.

Now if you are asking me how to make this lean run protection thing more robust, I'm not sure without perhaps knowing the throttle position sensor's output. But I haven't had time to give it the thought it deserves.

I have the launch control map zero'ed out and in fact dont' have the Xede wired to the clutch switch.

 

The rest of us need an answer to this also!

 

Yes I need an answer to this also .

 

Can you try loading a map without lean run protect. There may be a chance you are catching it in certain conditions. There is one other thing you can try. If we assume that the car wont even combust at 20:1, which it wont, we can put zeros in the top row, and negative 100s only in the middle rows. (perhaps from about 14:1 up to around 16:1.) The question at that point will be if the LC1 passes through these areas, or simply shoots straight to 20:1.

I will try to do a little testing this weekend as well.

 

In my experience there are zero problems running a "standard" non-smart map (e.g. a map downloaded from your web site converted to version 3).

I would remove the lean-run protect map but that requires "rewiring" a smart template but as you are aware the user version of Xmap doesn't permit this.

You will notice from the logs the AFR values that cause this condition to trip are just above the y-axis on the lean run map. I would change the points on this axis to give a little more room, but again a "user" isn't allowed to edit the map.

[QUOTE=Dustin@Vishnu]Can you try loading a map without lean run protect. There may be a chance you are catching it in certain conditions. QUOTE]

Not a bad idea, but the range of AFRs you suggest is outside of the range defined by the lean run map's y-axis and as cited above we can't change it. For reference the maximum AFR value in the lean run map is 45% or an AFR of ~14.

Given the user map limitations I believe that we are relying on your work at this juncture. In my opinion Vishnu should reconsider limiting the full Xmap to "tuners". I'm not sure what I'm doing for the last few weeks, but I think that it's tuning.

 

I am having the same problems, but I think on the overboost map though not sure. I have the lean run map where it was when opened, which is pretty high I think. The overboost map is waaayyyy up there. On a third gear pull, and when downshifting from third to second I am hitting the cutoff and throwing some p0300 codes, which my car has only done ONE other time. I have subjected my father to many, many full throttle third gear pulls while he was looking at the overboost and lean run map, the highlighted boxes were not even close to the lean fuel cut or overboost cells, but it sure as hell was hitting the cut anyway. The car was running very well before I loaded the 10 window, I will probably go back to the six window unless Vishnu finds the cause.

Details:

1. Six window 3.2....car ran very well @ 21-21 psi hitting mid 10's AFR in the top end.
2. Loaded 10 window, duplicated the parameters in the 6 window (timing, boost, fuel, SMART parameters, ect)
3. Left the flat shift, lean cut, and methonol maps alone.
4. Have messed with the overboost map (was hitting it right off the bat) but it is very high now.
5. Car is hitting some sort of cut and is throwing 0300's.
6. The highlighted cells are not even getting close to the cells which would trigger a cut.
7. It seems to take a bit longer for the 10 map system to boot up, at leats 4 seconds. Otherwise no boost.
8. So far the 6 map system seems to be much, much smoother, although I would like to have the saftey of the 10 map system if the bugs get worked out.

 

Why dont both of you guys go back to the six window map until I can replicate what you are talking about.