Spec B Posted May 2, 2009 Share Posted May 2, 2009 Weird thing happened today while logging...IAM dropped without a knock event. This goes against everything I have read...is this just a quirk in RR? Weirdness starts at time marker 75031. romraiderlog_20090502_091614 (Dropping IAM).csv Link to comment Share on other sites More sharing options...
merchgod Posted May 3, 2009 Share Posted May 3, 2009 That shouldn't be the case, unless the knock signal was set the entire time since the knock sum was incremented to 1 (which was before you log started), since the knock sum is not incremented during consecutive knock events. Are you having an issue with low RPM knock at all? Link to comment Share on other sites More sharing options...
NSFW Posted May 3, 2009 Share Posted May 3, 2009 Doesn't make any sense to me either. I'm curious though... Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 That shouldn't be the case, unless the knock signal was set the entire time since the knock sum was incremented to 1 (which was before you log started), since the knock sum is not incremented during consecutive knock events. Are you having an issue with low RPM knock at all? I get some/plenty of low RPM low load knock, I have just written it off as noise (FMIC piping rubbing in the engine bay). The KS usually increments about 20-25 times on 40 minute highway cruise..all of it below 3000 RPM and under light throttle. I get zero knock at WOT. I have done a ton of logging while cruising, but I have never caught my IAM dropping during a log. If you look further down the attached log, you see KS incrementing to 3. Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 Okay...so I went back to see the original log (the attached csv is edited because of file size limits). I remember that there was a ton of cruise (with KS at 1) before the drop. Like a dummy, I deleted the minutes before the log while trying to reduce the size of the file. I know KS was 1 when I started the log...not 0. Link to comment Share on other sites More sharing options...
merchgod Posted May 3, 2009 Share Posted May 3, 2009 Well, that is completely odd. There should always be an increment to the knock sum whenever the IAM drops, unless you have consecutive knock event(s). What rom id are you running? What is the nature of the tune - OS, Cobb? Do you know what the initial IAM is set to? Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 I am running OS, MickeyD FMIC tune. I have upped the boost slightly and changed fueling since I got my WB02...and a few other minor tweaks (reduced the Per cylinder timing compensation, ect.) but it is 99% unchanged. I am very surprised, I have done a lot of logging and I have never seen this. ECU ID:4512783106, Initial IAM is set to .685 The log above was after 40 minutes of driving on a new map in which I raised the boost WGDC slightly. Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 Well, that is completely odd. There should always be an increment to the knock sum whenever the IAM drops, unless you have consecutive knock event(s). What do you mean by consecutive knock events? Wouldn't knock sum catch that (eg, knock sum goes from 3 to 6)? Link to comment Share on other sites More sharing options...
merchgod Posted May 3, 2009 Share Posted May 3, 2009 What do you mean by consecutive knock events? Wouldn't knock sum catch that (eg, knock sum goes from 3 to 6)? No, it would not. Knock sum only increments with a single isolated knock event or the first knock event in a series of consecutive knock events. The tune wouldn't impact the IAM behavior. I really see no other possibility other than the above, however, I will look at your specific ECU to see if there's something unusual that I overlooked. Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 Interesting..thanks for the info. Just a thought....could the ECU have learned to pull timing in that region so much that it caused the IAM to drop even without an actual knock event? Looks like it drops right at about 2000 RPM, where I get plenty of knock. I am sure there had to be some learned KC there before the IAM drops. Link to comment Share on other sites More sharing options...
merchgod Posted May 3, 2009 Share Posted May 3, 2009 Ok, I had a brain fart and didn't look at your log very closely. A decrease in the IAM is always due to a knock event after entering rough correction mode. What happened in your case, however, is that you entered rough correction mode when all the conditions were met and the IAM was reset to the initial value of 0.6875 (conditions = FLKC < - 3.9, advance max > 4, in rough corr load/rpm range, FLKC change made). FLKC was < -3.9 before you started the log, however, when the IAM was reset, there was a positive change to FLKC (but final value still < -3.9), which met ALL the conditions to reset the IAM to the initial value. When your IAM was reset, there was no knock event then and over the "delay" period for increasing the IAM, the IAM increased by the step value of 0.25, and then eventually back to max 1.0. So, in your particular case, you had a somewhat unusual set of conditions (FLKC < - 3.9 and positive FLKC change) that caused your IAM to reset to the initial value but not decrease. Usually, it would be FLKC < - 3.9 and a decrease in FLKC (along with the other conditions), all of which would occur rapidly and you wouldn't even catch the initial reset of the IAM in the log, only a decrease in the IAM over the initial IAM value (there's no delay when dealing with a knock event). Your issue is that you are getting a ton of false knock (i assume false knock) at low rpm. The solution would be to raise the FLKC rpm low threhsold. Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 Excellent..thank you. I am pretty sure that the low rpm knock is just noise since it is in the stock portion of the map and I have seen Knock Sum increment with the hood up and me tinkering with the FMIC pipes. The biggest issue I am having is that quick drop in IAM..I suspected it was happening because my LV's looked clean on the low load/low rpm (even though I know I get a knock in that range)..I assumed (maybe incorrectly, I am not sure) that IAM was dropping below 1 and resetting my learned KC. This is likely happening very often. Link to comment Share on other sites More sharing options...
merchgod Posted May 3, 2009 Share Posted May 3, 2009 The biggest issue I am having is that quick drop in IAM..I suspected it was happening because my LV's looked clean on the low load/low rpm (even though I know I get a knock in that range)..I assumed (maybe incorrectly, I am not sure) that IAM was dropping below 1 and resetting my learned KC. This is likely happening very often. Yes, anytime that the IAM changes, the FLKC table is cleared. This particular case is an odd one because you had a large learned negative FLKC but when all the conditions were right for entering rough correction mode, you happened to have some advance in the current FLKC but not enough to go above the FLKC threshold. So, your IAM didn't drop due to a knock event, it was reset due to entering rough correction mode and only increased when being re-evaluated. If, for example, your initial IAM was set to 1.0, nothing would have appeared to happen in your logs in this particular instance. If everything else was the same except that the FLKC change was in a negative direction in that instance, then the IAM would have dropped. Changing the FLKC min RPM, which is the minimum RPM in which the current FLKC can be changed (changed not applied), above the RPM in which you have false knock would prevent entering rough correction mode based on this false knock because FLKC < - 3.9 (or > 3.9) is a requirement. Usually this occurs very rapidly as usually all the conditions come together at the same time, but not in the case. Link to comment Share on other sites More sharing options...
Spec B Posted May 3, 2009 Author Share Posted May 3, 2009 Okay..I have raised flkc min rpm to 2500 RPM and rescaled the fklc rpm ranges (the rows on the LV)...hopefully this will cure my IAM dropping issues. Link to comment Share on other sites More sharing options...
NSFW Posted May 3, 2009 Share Posted May 3, 2009 Merchgod, how feasible would it be to create a patch, to the DTC patches, that would keep the ECU in fine correction mode at all times? I'd like to think that it would only require changing a single branch instruction to a jump or no-op so that one of the conditions for entering rough mode would never be true. I wouldn't want anyone to run that way after a tune is done, but I could see this being useful when we're checking learning view frequently. It would prevent confusion like this, and prevent us from losing knock history due to wipes of the FLKC table. Link to comment Share on other sites More sharing options...
merchgod Posted May 3, 2009 Share Posted May 3, 2009 That would be a bad idea. Imagine if you got a bad tank of gas and the ECU has to learn large negative corrections in every cell - bye-bye motor. FLKC has nothing to do with knock handled by FBKC, so looking at FLKC is not the whole picture - you need to log anyway. Knock sum is the easiest means for tracking knock events. Link to comment Share on other sites More sharing options...
Spec B Posted May 5, 2009 Author Share Posted May 5, 2009 Is it correct to assume that if you get very little or no FLKC, then your IAM will be remain at 1 regardless of what FBKC is doing? What protects the engine if I only get WOT knock? Is it only FBKC? Based on what I have read, FBKC is mainly active during big changes in RPM or load (eg WOT runs) and IAM is active only after FLKC has done it's thing (eg during cruising). Link to comment Share on other sites More sharing options...
merchgod Posted May 5, 2009 Share Posted May 5, 2009 FBKC is always the default and can be used at any time within the FBKC rpm range and if min load threshold is met. There are conditions which ALL must be met in order to disable FBKC which then would allow either a change to FLKC or a change to the IAM (with other conditions that must be met). One of the conditons to disable FBKC is that the change in load is moderate/low. Link to comment Share on other sites More sharing options...
Spec B Posted May 5, 2009 Author Share Posted May 5, 2009 Thanks. So the ECU can transition from FBKC mode to rough correction mode? Link to comment Share on other sites More sharing options...
merchgod Posted May 5, 2009 Share Posted May 5, 2009 It is best not think of an FBKC mode. Read the knock control sticky on romraider. it is all outlined there in detail. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.