l2r99gst

iTrader: (2)

Just rereading the first two posts in the advanced fuel thread by mrfred, there are several factors that could be the culprit here, or a number of them together, or something more that isn't defined.

It may be best to go through those couple posts and start eliminating the different variables to see if any one or combination of them can be identified.

For example, looking at the equations, something as simple as when the fuel trims switch from LTFT Lo to LTFT Mid can have an effect, if your LTFT Mid is much different than your LTFT Lo. But that's just one example...there are 10s of variables to go through.

03whitegsr

iTrader: (8)

It is just that and I apologize if it comes off like I'm attacking you.

It seems like every thread were people seem to have the symptoms of this issue, you get into the thread to help but it always comes out as "change your VE maps." I know you mean well, but from my testing, I feel it has nothing to do with it and your post seems to undermine people actually looking into the issue further. The fact that you haven't experienced this issue to me either means it is really setup dependent or it is just how you drive the car and you happen to drive the car in a manner that you never noticed it. I've just seen to many people either post about symptoms of it or work arounds they have done to cover it up, either knowingly or unknowingly to think it is my tuning. Originally, I thought it was my tuning though and it was very frustrating when everybody kept saying to retune.

It's finally to the point though were I'm not the only one noticing it or at least posting about it. If you look back to my posts the first day I used this, I reported this very issue. I reported the issue in the jitter fix thread to and just wrote it off as a tuning issue at the time. I have had this issue since day 1 though and I have literally retuned the car probably 10 different ways over the course of 30+ retuned to try and get rid of the issue. I can change the main fuel map and accel tables and get it to "go away" just like everybody else can, but it's just a cover up and if you look at the logs when it's doing it, you can tell there is something else stepping in that is messing with the main pulse width.

Originally, I was thinking it was just on the cold engine. I started paying attention though and started noticing the exact same thing happening when it warmed up, but as mentioned above, the car drives through it so it's not noticeable through "fell" alone and you have to watch AFRs.


The baro issue cause the closed loop stuff to be a little odd since you stay in closed loop even at atmospheric conditions that with corrected load would be a load of 100. It also seems to cause start up issues since it doesn't know the altitude is different. Driving wise though, baro seems to have little imapct once the car goes into open loop.

As a last note, when the car is doing this, the accel and decel MUT channels go CRAZY yet the load stays stable. It's like there is some other load that is getting switched in during part of the calcs and it causes the load based accel stuff to go crazy, but then the load we log never shows it. I was looking at the main IPW sub last night and there are a couple places where some interesting things happen with the MAF frequency switching between different variables during the accel and decel stuff and that is where I will be looking next.

03whitegsr

iTrader: (8)

FWIW, here is hw I've been able to make this issue EXTREMELY repeatable.

This will be most noticable when ECT is below about 30C because (in my car anyway) the car misfires and jerks like crazy. Even at operational temps though, the lean issue will still be there, the car just likely won't jerk a bunch because of it.

Get the car into a higher gear, 3rd or 4th, and let the revs drop all the way down to near idle conditions (less then 1100 RPM), try to accelerate now. 100% throttle, partial throttle... shouldn't make a difference. The car should go incredibly lean until past 1500 RPM. The lean issue will go away between about 1750-2250 RPM.

NOW, richen up the RPM and MAP VE in this area and see if it actually has an effect. For me, it wasn't proportional. I could add 20% fuel and it barely did anything to AFRs during the conditions that caused missfired...it made the car idle at 10:1 though...

l2r99gst

iTrader: (2)

Sorry if I come across that way...I DEFINITELY don't ever mean to undermine anyone. I simply want to make sure that we get a grasp on the true problem. This is the first time I have seen a post by crx about this, so I wanted to get all of the bases covered first. It's always nice to get the simple solutions covered first.

And believe me, I know what you are going through. Remember when I first brought up the jitter issue? No one believe me and everyone kept giving other suggestions and answers post after post. But, I backed up my claim with evidence and detailed logs and screenshots, along with an explanation of exactly how to reproduce the problem...sort of like your description above (which is a nice detailed description on how to reproduce it). Without logs or a good detailed description, then all we can do is guess.

If/when I ever do switch back to SD again, I will try to reproduce this and get some DMA logs from the various variables.

What have you personally tried so far as to fix it (besides the fuel map and accel tables, which maybe you can post to help crx in the meantime)? How many of the newly defined advanced fuel control tables have you tried from mrfred? If you see the decel and accel channels going nuts, have you tried messing with any of the related tables and settings? My first inclination would be to look at any tables that have different values based on RPM or load. I would like to see a log of yours to see the issue and when it finally corrects itself, I would like to see what the values of all pertinent data is right before and after (ie all loads, maf hz, rpm, etc, etc)...maybe we can correlate it to one of the settings in the newer tables.

I always like looking at screenshots, so maybe highlight the area of where the issue is and then where it corrects, so we can get a graphical view of what is changing (if anything) at that point. Maybe the change in delta of a certain channel (I think I remember some settings for change in load for sync accel stuff, for example).

03whitegsr

iTrader: (8)

Yeah, I've tried all kinds of changes and I've made posts about the changes in various places with logs, settings, graphs, all kinds of stuff. Probably the biggest problem really is that every time I try something else, it leads me somewhere else. I keep trying all kinds of things that are suggested and it ends up messing up my tune. After I try about 1 million things, I end up jacking up my base map and then have to completely retune. Or I'll think I found something that helped (but not fixed it) and then drive around on it long enough that I forget what was my original map. It got bad enough that I drove around on a tune for several weeks where I couldn't go into boost at all because I had just spent so much time on trying to fix this issue that the rest of my tune was uber junk. I got sick of it and finally just put the simplest tune on the car I could and started ignoring the issue. the car drove GREAT above 2000 RPM and just starting off from a stop was the only time the car drove like crap so I worked around it.


FWIW,
https://www.evolutionm.net/forums/ec...6531706-a.html

If you look, slo crx even posted in there suggesting my VE tables were incorrect and now he's posting about the same issue. It kind of shows how it is very easy to overlook until you see it for the first time, then it pisses you off every time it does it.

l2r99gst

iTrader: (2)

Yeah, I hear ya on making a million different changes that just leads to a million other changes.

I think this is where DMA/LiveMappingwould be invaluable. Just put like 10-20 advanced fuel related maps (or however many you could fit) into RAM and then make live changes one at a time to each map. Go to the extremes on each map and see how it affects the issue, to start to narrow down the culprit, if possible.

That way at least, a million changes takes very little time instead of continuously flashing/reflashing, stopping/restarting, etc, etc. You can see realtime affects, within seconds of each other and blast through all the tables in one short session.

Hopefully, in the end, you find the solution. If I ever go back, I'll see if I can reproduce and test as well.

l2r99gst

iTrader: (2)

03white,

For fun, I went through and reread your whole thread on your issue, that you linked to above.

I mentioned in one of the posts to adjust your the first 3 cells of your 'Ceiling Load vs RPM for Sync Load Accel/Decel Contributions' to a much higher number to see if the sync accel FPW didn't drop out.

You mentioned that you tried that (along with a couple other changes) and it made the IPW even jumpier, but I don't think you posted another log/screenshot. Did you happen to have a log of that by any chance? I'm just curious because after rereading that thread again, and looking at your second graph of your issue, it coincides exactly with the the sync accel fpw dropping out. I'm just curious if changing that table enabled the sync accel fpw to come back in the range where you had the lean issue.

Also, did you lower the values in the following tables?:
Min Uncomp Load Change at Steady Throttle for SyndLoadAccel/Decel Contribution
Min Uncomp Load Change at Changing Throttle for SyndLoadAccel/Decel Contribution

Slo_crx1

iTrader: (9)

I hadn't posted any logs due to the fact the size of most of my logs are pretty large (around 8-10MB or so) due to large amounts of driving and logging. I have one that's around 2MB I'll try and post if it will let me.

*Edit: I just tried to upload the file, but for some reason I can only see up to logs from this past january in the explorer even though I can open the logs from earlier today in Evoscan. I can't win.

The lean spool didn't really make any difference except for the areas I set it to, which only really leaned out my afr's heading into boost. I did bump up my latency voltage under 12v just to see if that would make any difference as well, and even though it made the driving smoother it didn't really have much effect on the lean spot. To me it seems to depend on the load...for example, on a flat road or slight downhill it will only lean out into the 15afr range which I can deal with. Anything up hill will drop into the 17-18 range due to the higher load placed on the engine, and in which case I have to let off the throttle slightly in order to let the o2 catch up and bring my STFT back up to a manageable afr.

For some reason I'm one of the few that doesn't nor never had the lean start up issue that I've seen on here quite a bit. During warm up I usually see low 13's until coming up to temp where it stabilizes at ~14.7afr. I never posted much up about the lean spot because I just figured it was a by-product of the patch and had to deal with it, but lately it's been getting on my nerves and I'm seeing the same issue with my friend's car as well. So far I see the 118 to 287Hz airflow ranges as being the worst on a consistent basis while under 2000rpm, while anything over 2000rpm in those ranges has a decent STFT correction.

Unfortunately posting the log I'm currently looking at would be useless since once again my trims are moving and I need to re-correct my RPM VE table again, even though as of now I'm still only +4% LTFT Low and -4%LTFT Mid. This is probably where a 3D VE table would really help out.

l2r99gst

iTrader: (2)

You don't need to post entire logs...ideally just a log, or even better yet, a screenshot of a log in a nice program like LogWorks would help a ton.

I wonder if you have same sync accel fpw issue that 03white saw in his thread. If you read his thread that he linked to, some screenshots are in the first post that shows this. I would be interested to see if you are getting something identical.

If so, maybe try some of the advanced fuel control tweaks that were mentioned, or at the very least use 03white's fixes to at least overcome the issue for the time being.

Slo_crx1

iTrader: (9)

It's very possible I might have the sync accel fpw issue, I'll have to go over the thread again since I've been off of here for a little while and missed quite a bit of information. I'll post up a screenshot at the least and read 03white's thread completely and see if anything from there will work. Thanks guys!

Slo_crx1

iTrader: (9)

I read through 03whitegsr's thread to see if I could get anything to work that he may have used, and no such luck. Nothing I've done so far has combated the lean out issue...nothing with the async tables, nothing with the sync tables, nothing in the main fuel tables, nothing with throttle crossover, nothing with the RPM and MAP VE tables, nothing with the injector latency or scaling, nothing with the MAF compensation or scaling tables. Frankly I'm running out of ideas here. I have a couple screen shots from a log yesterday afternoon to show what I mean:

This is right as I'm about to take off, the spiking up and down is from me tapping the throttle to bring the rpm up as I'm slipping off the clutch at an off ramp.

















Any suggestions? LTFT's were +2 and -2 during this log, eventually the STFT brings the fuel back after it jumps to 18:1 afr but it still makes things very interesting and HIGHLY DANGEROUS when pulling out into traffic.

Slo_crx1

iTrader: (9)

Odd thing I noticed today, I upgraded my Tial WG spring to a 18psi set (instead of the anemic 5psi one that was in it) and between the upgraded spring and the cooler air temps (in the low 60's today instead of the 80's we've been having) the lean out issue was less pronounced. I never really remember having that big of a problem with it when the temps were colder when I was tuning the car, I'll dig up some logs from back in January and see if it was there as well and if it was as pronounced.

*Edit: Just looked over a log from January, the leanest point during the log from a take-off was 14.9afr, but for some weird reason the IPW throughout all of them is always 2.5ms for that stretch of lean period.

elhalisf

iTrader: (171)

I have the very same issue with mine when i converted to SD last week.
now what does the Sync Load Accel Compensation vs RPM table do?
it is showing a factor of 1.0 from 1500 to 2500 rpm, the exact rpm i am having the issue at.

Also, the table Async Accel vs RPM Adder shows 0 units at the same rpm tables.

Has anyone changed those tables?

Slo_crx1

iTrader: (9)

I changed the Async Accel vs RPM Adder up to 10 although I don't know exactly how much it helped out. I also tried the Sync Load Accel Compensation for the trouble range and it did nothing except screw up my lower rpm cruise afr's causing them to go very rich and absolutely nothing for the lean issue during initial take off.

Today I didn't see this issue like I usually do, so I'm really starting to wonder if intake temp really does have any effect on this. Today driving home from work was still in the low 60's ambient, later in the week is supposed to be back up to the mid 80's so I can see if the issue returns. It's almost a shame there wasn't a way to force the ecu to go "Alpha-N" speed density for that trouble range and back to the MAP sensor after.

JohnBradley

iTrader: (30)

Lean issue is caused because of the adjustment at the 99kpa side (9653's for reference, poet 8). See how the righthand is at 92. This means it corrects lean, needs to go the other way like 102 or 105 or more. All in the MAPVE setup, Asynch Delta can be left stock and this can be correctly tuned for all RPM based on that table.