weitek Posted July 9, 2009 Share Posted July 9, 2009 Thanks for the link. Are you still on Rev22a? Link to comment Share on other sites More sharing options...
Spec B Posted July 9, 2009 Share Posted July 9, 2009 Not really...I have made some minor tweaks since that version, but most of the tables are the same. Link to comment Share on other sites More sharing options...
weitek Posted July 9, 2009 Share Posted July 9, 2009 Would you mind posting your current version? (If that's okay w/ MickeyD or whomever). I'd like to check it out and if possible I'll add some tweaks and post it up for you to try. Link to comment Share on other sites More sharing options...
Spec B Posted July 9, 2009 Share Posted July 9, 2009 Okay..I will try to post tonight, I have it on my logging net book. Link to comment Share on other sites More sharing options...
Spec B Posted July 10, 2009 Share Posted July 10, 2009 Here is the latest version of the my map.. http://www.box.net/shared/cnnxxpvio8 Link to comment Share on other sites More sharing options...
weitek Posted July 10, 2009 Share Posted July 10, 2009 Thanks for posting. I will look it over later today and get back to you. Link to comment Share on other sites More sharing options...
Spec B Posted July 10, 2009 Share Posted July 10, 2009 Cool..I am always interested in learning new tuning tweaks. Link to comment Share on other sites More sharing options...
weitek Posted July 10, 2009 Share Posted July 10, 2009 Eek.. you are running 2 degrees of advance in cylinder D. I would zero that out asap if I were you.... Link to comment Share on other sites More sharing options...
Spec B Posted July 10, 2009 Share Posted July 10, 2009 I think that is cylinder 4 (the one closest to the knock sensor) and I am did it that way so that if any cylinder knocks it will be that one, the one closest to the sensor.. Link to comment Share on other sites More sharing options...
weitek Posted July 10, 2009 Share Posted July 10, 2009 Yeah but 2 degrees is quite a lot. I prefer to take a different approach - you'll see what I mean when I upload the map I made for you. Do what you feel comfortable with though, it's your car.  Also - do you have a wide band o2 sensor? Link to comment Share on other sites More sharing options...
Spec B Posted July 10, 2009 Share Posted July 10, 2009 Thanks. Just let me know what changes were made to the map..so I could keep track of them and hopefully learn a little bit. Â My mods are:Â typhoon intake - scaled with the wb02 1mm pill xo2 FMIC invidia catless DP xo2 mid and Y and borla mufflers LM1 WBo2Â Basically a stage 2 with FMIC. Link to comment Share on other sites More sharing options...
weitek Posted July 11, 2009 Share Posted July 11, 2009 Here it is: http://www.subyshop.com/Spec B Rev 26.hex Sorry it took me so long. I've been really busy and I also put a lot of work in to your map as I wanted to get it right on the first try. I changed 75%+ of the tables so I suggest you look it over and study it for a bit and then ask me some questions before you flash it. It's got a little less here, a little more there, and much smoother over all. I suggest viewing the tables in EcuFlash and use the 3d graph feature. Highlights are: Re-worked and smoothed Target Boost, WGDC, Open Loop Fueling, Base + Advance Timing, AVCS & DBW. This map should be 100% safe but please do not flash it to your car unless you fully understand what you are doing. Also you MUST log your first few runs to make sure it runs safely and does not have any un-expected issues. I expect we will be able to make some big improvements after I get some logs and see how it runs. For example you can definitely lean it out a bit more and it will be completely safe but I need to see how well your MAF is scaled before I would attempt that. Please log the following things when you log it: A/F Correction #1 (%) A/F Learning #1 (%) Engine Load (Calculated) (g/rev) Engine Speed (rpm) Feedback Knock Correction (degrees) Fine Learning Knock Correction (degrees) Ignition Timing Advance (degrees) Ignition Total Timing (degrees) Manifold Relative Pressure (Corrected) (psi) Mass Airflow Sensor Voltage (V) Primary Wastegate Duty Cycle (%) Throttle Opening Angle (%) Wide-Band O2 Sensor (AFR) Looking forward to your feedback. Link to comment Share on other sites More sharing options...
Spec B Posted July 11, 2009 Share Posted July 11, 2009 Thanks a lot. Took a quick look at your changes..and flashed it. Sorry the logs are not as exactly you asked for, I was trying to get some quick runs in before the rain. Here is a WOT 3rd gear run and a LV...I think the only thing I failed to log was maf V. Overall the car feels more responsive...power is about the the same if not a little lower than previouslyr logs..but that might be due to the weather (it was in the low 80s today). My idle fuel trims have never looked that bad (+9%), I will check my intake. They were at 2% earlier today, but I was fiddling by the intake right before I flashed your map. So lets ignore that, I am sure it will come down to less than 5%. romraiderlog_20090711_153316.csv Link to comment Share on other sites More sharing options...
weitek Posted July 12, 2009 Share Posted July 12, 2009 Your logs tell me the car isn't knocking and it's running really rich but that's about it.  Please do some logs again with the specific settings I listed. Also please don't include anything else, only the ones in the list. For example don't log 'knock sum', log feedback & fine learning. And we don't need turbo dynamics or MAF g/s - only MAF v << this is actually one of the most important values I need to see. I hope that doesn't sound rude - I just need those logs to be able to help you.  Edit: Please do at least 3 WOT pulls in 2nd gear from 2k to 7k. I need to see the entire range and I need multiple pulls so I get a good average. Also log the car sitting idle for at least 5-10 mins and also at least 10-15 mins worth of normal cruising around without boosting - stops and starts and some part throttle. Try to mix in some city and highway driving. Link to comment Share on other sites More sharing options...
NSFW Posted July 12, 2009 Share Posted July 12, 2009 I have to ask.. Why do you prefer MAFv to MAF g/s? Why do you prefer knock feedback and learning to knock sum? Link to comment Share on other sites More sharing options...
weitek Posted July 12, 2009 Share Posted July 12, 2009 I have to ask.. Why do you prefer MAFv to MAF g/s? Why do you prefer knock feedback and learning to knock sum? I use mafv when I tune the maf scaling. I take a whole bunch of logs and put them in excel as a scatter plot with the wide band as the Y axis and mafv as the X axis. It's just a lot easier to work with numbers from 1-4 instead of 1-300 or more. Maf g/s is more useful if you are trying to gauge how much power the car is making (assuming the scaling is accurate, which from my experience it almost never is unless I've personally tuned it). For example Spec B's car is dipping into the high 10's on those logs he did, and I even saw it go as low as 10.5 for a bit. It seems to be scalled pretty well in closed loop but the open loop fueling is all over the place. I've read about knock sum and it seems like a useful tool but I just prefer seeing fine learning vs feedback as it's how I learned and I'm more comfortable with it. I feel like I understand what is going on better when I'm able to see which it is. Just a personal preference though I guess. Link to comment Share on other sites More sharing options...
NSFW Posted July 12, 2009 Share Posted July 12, 2009 It's very common for FMIC'd Subarus to dip into the 10s when you hit the gas. There have been threads about it at Nasioc and RomRaider... when I saw my car doing it, I suspected tip-in, so I zeroed out tip-in and the problem was still there, so I'm certain that's not the cause. The prevailing theory so far is that there's a delay between when the MAF sees a rapid rise in airflow and when all that air actually makes it into the cylinders. Basically, it takes time to pressurize the FMIC and charge pipes, and during that time the MAF is reading more airflow than the cylinders are getting, so the mixture goes lean. I'm not convinced that's actually what causes it, but some people are. At some point I want to switch to blow-through MAF, to see if that really makes a difference. A second possible explanation is called Tau, or X-Tau, or "wall wetting." I've actually been meaning to bring this up on the RomRaider forums for a while now... http://www.megamanual.com/ms2/xtau.htm I have a couple ideas for investigating that possibility via creative logging, for example free-revving the engine just prior to the log pull to see if that blows out the fuel on the back of the valve / walls of the intake port.  But even if the "tau" theory is correct, it still leaves the question of why the rich dip is more pronounced with FMICs, so I'm guessing there's still more to the story. Link to comment Share on other sites More sharing options...
weitek Posted July 12, 2009 Share Posted July 12, 2009 It's very common for FMIC'd Subarus to dip into the 10s when you hit the gas. There have been threads about it at Nasioc and RomRaider... when I saw my car doing it, I suspected tip-in, so I zeroed out tip-in and the problem was still there, so I'm certain that's not the cause. The prevailing theory so far is that there's a delay between when the MAF sees a rapid rise in airflow and when all that air actually makes it into the cylinders. Basically, it takes time to pressurize the FMIC and charge pipes, and during that time the MAF is reading more airflow than the cylinders are getting, so the mixture goes lean. I'm not convinced that's actually what causes it, but some people are. At some point I want to switch to blow-through MAF, to see if that really makes a difference. A second possible explanation is called Tau. I've actually been meaning to bring this up on the RomRaider forums for a while now... http://www.megamanual.com/ms2/xtau.htm I have a couple ideas for investigating that possibility via creative logging, for example free-revving the engine just prior to the log pull to see if that blows out the fuel on the back of the valve / walls of the intake port.  But even if the "tau" theory is correct, it still leaves the question of why the rich dip is more pronounced with FMICs, so I'm guessing there's still more to the story. Hmmmm very interesting. From my experience it's very common for modded subaru's to dip rich when you hit the gas. On some cars it can be very frustrating to compensate for. It's almost always been due to bad maf scaling though when I've seen it - fmic or not. However I will admit my experience is still limited and I am still learning new tricks all the time.  I'll take a look at the tau thing tomorrow. I'm looking forward to those logs now. This should be fun. Link to comment Share on other sites More sharing options...
Spec B Posted July 12, 2009 Share Posted July 12, 2009 ^ Okay...I will log exactly what you need. Knock sum has always been my preference because it shows you exactly when and where I knock (instead of the ecu's reaction to it), so for my sanity, I am going to add that parameter to your list, so I could catch knock immediately while driving around. I will be a little slow getting the logs to you...I barely drive my car, but I will get them. Just to confirm, you want 2nd gear logs, not 3rd? From 3500 RPM up, the scaling has been historically pretty much dead on, I have been logging/tuning for a while now, and I spent a ton of time scaling the maf...I could never nail down the 2000-3500 range. Link to comment Share on other sites More sharing options...
weitek Posted July 12, 2009 Share Posted July 12, 2009 ^ Okay...I will log exactly what you need. Knock sum has always been my preference because it shows you exactly when and where I knock (instead of the ecu's reaction to it), so for my sanity, I am going to add that parameter to your list, so I could catch knock immediately while driving around. I will be a little slow getting the logs to you...I barely drive my car, but I will get them. Just to confirm, you want 2nd gear logs, not 3rd? From 3500 RPM up, the scaling has been historically pretty much dead on, I have been logging/tuning for a while now, and I spent a ton of time scaling the maf...I could never nail down the 2000-3500 range. Hmmm.. I thought KS was just a different way of displaying the same data.. i.e. if you were to log all 3 you'd see the KS change at the same time index as feedback/learning. I guess I will have to read up on it some more. 2nd gear logs work well as there is generally less fluctuation in boost than in higher gears (which will make the maf jump all over the place). Also it's usually easier for the driver since you don't have to go as fast so I've learned I'm more likely to get good data if I ask for 2nd instead of 3rd. Plus it's probably less likely to knock in a lower gear, so there's no point to put the engine at greater risk until it's all tuned better.  As for the scaling - it's definitely not the worst I've seen and I don't mean any offense by my comments - but it's high by ~.4 from 3200-3800, and then low by ~.2 from 4000-4400, and again after 5900. Plus I'm still hoping I can fix that rich spot at the beginning as that's got to be killing your performance down low. Link to comment Share on other sites More sharing options...
Spec B Posted July 12, 2009 Share Posted July 12, 2009 No offense at all..I just notice that the high end WB02 readings are always + or- .01-.02 off depending on the weather. I don't know if it could be scaled any better up top...but if it can, lets do it! Â It is kind of redundant, Knock sum is a counter that counts actual knock events...so if there is any knock, the counter will increment by one. When you log feedback knock correction and fine learning knock correction, it will show timing pulled immediately after the knock event and the pulled timing will generally stay there until the ecu is happy again (no knock). Obviously, this all happens really fast. Remember, the ECU can only use on method at a time to pull timing so it is either Feedback, Fine learning, or Rough (IAM). Â The nice thing about KS is that right before a log run I look over and see that my KS is say 15..I do my pull and I could quickly look over again and see where the KS is. If it is still at 15, I had no knock and there is no need to stop to review my log before going on to the next run. If after the run KS went up to 18, Then I know there was 3 knock events during the pull and I need to check what is up. Link to comment Share on other sites More sharing options...
rtbrjason Posted July 12, 2009 Share Posted July 12, 2009 I notice IAM initial is set to 1. I prefer to have it at .5 so it can raise to 1 if all is well.The NSFW strategy was undone also. I like this approach myself and would put that back.. That is the retard value being -1.41 vs -1.4, and the advance value being .2 vs .35 for counting knock events with the hundredths column as well as easily seeing how many corrections have occurred. I haven't gone through the whole thing and compared, but the boost curve looks a little smoother on the new map as per excel charting. I'm not sure why you need wgdc to hit 98 though. Does it seem to spool any better or hold boost higher than previously? Link to comment Share on other sites More sharing options...
Spec B Posted July 12, 2009 Share Posted July 12, 2009 I was very happy with my previous tune..but I am curious to see how other people tune. I agree with what you said regarding the initial IAM and the NSFW strategy, I will let weitek do his thing and then tweak the map to my preferences later. Spool is a little slower..but boost is holding better on the top end. What most impressed me so far was the throttle response, there is definately less lag. It also seems to pull a bit harder on the high end..I think that is a result of the AVCS changes. Since I am on the stock turbo, I am pretty sure my max power has peaked, but if drivabilty and high end power improves a bit..that would be great! Link to comment Share on other sites More sharing options...
rtbrjason Posted July 12, 2009 Share Posted July 12, 2009 What most impressed me so far was the throttle response, there is definately less lag. Â Did you notice same when using rev 25 with sti throttle mapping or was it the rev 26 that weitek did where that became more apparent? Link to comment Share on other sites More sharing options...
Spec B Posted July 12, 2009 Share Posted July 12, 2009 I noticed a slight change when I changed the DBW map (rev 25)...but I notice it more on this latest rev. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.