ECMlink Trying to get the MAF dialed, AFRs are unstable

[UofACATS]

edit: This is unresolved if you're reading this from a search, and have a similar problem. I threw in the towel and went SD.

This log is crazy. The LM-1 is all over the place.. I was having major issues with AFRs previously, it turned out to be a faulty fuel pump o-ring. The car was tuned very well after replacing it (and installing a Walbro 255). Trims and everything spot-on. Then a couple days after the CombinedFT shot up to 20%+ and I figured "huge boost leak," but I tested it yesterday, it holds 25psi just fine. I double checked every connection on the intake for cracks or leaks and found none. Then I figured the 255 was the culprit so I reinstalled the evoIX pump that was in there previously. I also climed under the car and double checked the front o2 and wideband sensors which were secure. The downpipe to o2 housing was a bit loose, so I tightened that up and also verified 43.5psi base fuel pressure. I cleared everything and reset the fuel trims and started over. (I'm running E78 fuel).

I'm attempting to dial my MAF in but I fear I have a tuning error or mechanical problem which is severely skewing my results.

I altered the OpenLoopThreshold and ran the MAFcomp tool. It showed huge compensation but my MAF has always been like that. I ran the tool again on the way home tonight and it shows another large jump in additional fuel. (I don't know it it's accurate or accounting for a mechanical or tuning error)

At some (inappropriate) points in the log the IdleSw activates. I tried selecting "simulate idle switch from TPS in this log to see if the weird IdleSw problem would stop.

At 67.75 seconds the FuelFlow and InjOn drop to 0.00 and the car jumps to OpenLoop.

 

[gofer]

Has this just recently started doing this or has it been an ongoing problem? What type of air filter and intake do you have on the car? I looked in your profile but couldn't find the info.

 

[UofACATS]

Has this just recently started doing this or has it been an ongoing problem?

You mean the roller coaster AFR ride? That just started. A while back I was having diverging AFR/AFRest at WOT past 10psi and thought it was the pump. I pulled the in-tank evoIX pump, and installed the new (used) 255. At that time I discovered the cracked O-ring sealing the pump to the sender. I replaced the o-ring and went ahead and tossed in the 255 anyway. I did a WOT pull and the problem was eliminated. I've done the idle tune so many times it's like clockwork I got the tune spot-on, trims nearly perfect, and had moved on to WOT tuning.

Then a couple days later I feel the car stumbling at idle and see the streaming log showing +20 on the CombinedFT's. It's such a massive correction it has to be mechanical. (and it wouldn't stay at 20+, it would float around from 10-22%+ and when I get on the throttle a bit the AFRs look much more reasonable. That's why I assumed massive boost (vacuum) leak and/or a torn intake allowing unmetered air. I also thought exhaust leak because of the floating (and obscene) nature of the tune. I found the sensors to be secure, and I did find a loose connection between the o2 housing and the downpipe, and have remedied that previous to the log (very little if any noticeable change).

What type of air filter and intake do you have on the car? I looked in your profile but couldn't find the info.

I don't list an intake because it's stock, but I do have a 4" air filter adapter on the MAF, and I think it's a Vibrant air filter. That filter has been there a while, the stock intake I re-installed after switching from the 3065 to the 16G.


Thank you Gofer.

 

[gofer]

I know you're running a 16g so you might care less but I thought you might be interested in this thread, if you haven't already seen it? HTA88 Killed My Filter

Does your stock intake pipe still have that long ass tube inside of it coming in from the bov recirc fitting? Rather than posting a driving log post a idle log with the car completely warmed up and it jumping around like that.

 

[UofACATS]

Checked the link, I just died a little inside


And I do care of course (is the Vibrant filter the same thing?), however the reasoning behind my MAF dial-in is to use SD VE Adjust (SDRatio) to dial in my Speed Density tune, for which I am otherwise ready to run. With a SD setup, I plan to run a 2.25" intake to match the compressor inlet. I'll likely have to cone-out to 2.5" to fit a filter I won't have to special order. At that time I'll get a new filter. I haven't done research on air filters in 5 years, but I'll do that before purchasing. That FP88 carnage is terrifying.


Yes, my stock intake still has the dump tube. It's horrifying in design, but it's a temporary piece until I can pull it off and put it where it belongs, on the shelf not being used.

I'll post an idle log, thank you for the input.

 

[gofer]

PULL THAT RECIRC TUBE INSIDE THE INTAKE! My theory is that damn thing is creating turbulence and jacking with the airflow.

 

[UofACATS]

really? the oem piece?

I'm just thinking out loud here.. I suppose it's not designed to work with a cone filter but it is designed to pair with the oem MAF. Additionally, it's after the airflow gets metered, not before..

Curious what you think, especially because it's not at higher RPM where you figure turbulence would rear its ugly head. But don't get me wrong either, I'll go cut that nasty pipe out right now especially considering it can't hurt.

 

[gofer]

It's super glued in there from the factory I think, no cutting required. Its just an idea/theory of something that could be causing the issue. Granted, it is after the MAF but sits close enough to it that somehow it could be creating turbulence up in the MAF itself... Pull it and datalog the car idling fully warm. Odds are its something else, like on the fuel side of things, but that would be cool if it was something as simple as that damn recirc tube in there.

 

[Colt4G63 also]

At 67.75 seconds the FuelFlow and InjOn drop to 0.00 and the car jumps to OpenLoop.

I don't see anything abnormal about that, considering you were under decell with the throttle fully closed. If you didn't let off there you better replace that TPS.

The other thing that looked funny to me was the fact that MAF raw never went above 200hz. The whole thing looked a little flat to me but maybe you were just driving really mellow.

 

[UofACATS]

I don't see anything abnormal about that, considering you were under decell with the throttle fully closed. If you didn't let off there you better replace that TPS.

The other thing that looked funny to me was the fact that MAF raw never went above 200hz. The whole thing looked a little flat to me but maybe you were just driving really mellow.

Shouldn't the injectors still be on though? The car still has to idle, and I'm accelerating, cruising, and decelerating the entire log and the injectors are otherwise on the entire time. The log is me driving down my driveway, I wasn't out to capture a log, I just saw the weird behavior and captured it.

If the cars idling, shouldn't it stay in closed loop? It has no reason to go Open Loop unless somethings drastically wrong.

It's super glued in there from the factory I think, no cutting required. Its just an idea/theory of something that could be causing the issue. Granted, it is after the MAF but sits close enough to it that somehow it could be creating turbulence up in the MAF itself... Pull it and datalog the car idling fully warm. Odds are its something else, like on the fuel side of things, but that would be cool if it was something as simple as that damn recirc tube in there.

I captured two logs.

Number one is a quick idle log at operating temperature, and two is another idle log, with the dump tube removed.

1:

You must be logged in to view this image or video.

2:

You must be logged in to view this image or video.

 

[gofer]

The injectors shut off on decel to prevent flooding the engine and it kicks into OpenLoop because of the "load" being below a certain point. Similar to being in gear and going down hill...

I posted a log of a 1/4 mile run I did, at 191 seconds you can see my injectors shut off and I go into open loop for awhile until my LoadFactor comes back up above 0.10-.

Thats normal.

 

[UofACATS]

Good to know. At this point I'm looking at everything sideways..

 

[gofer]

Now tune your idle to that last log, what do you think? Raise your 0Hz and 50Hz slider up to 15.6% to bring up AirflowPerRev to 0.25 gm/rev then flirt with deadtimes a bit to get CombinedFT's to cycle around 0 (+/- 8%)?

Then do a cruise log with your OpenLoopThresholds set where they are at. 15 minute drive warmed up!

 

[Colt4G63 also]

Shouldn't the injectors still be on though? The car still has to idle, and I'm accelerating, cruising, and decelerating the entire log and the injectors are otherwise on the entire time. The log is me driving down my driveway, I wasn't out to capture a log, I just saw the weird behavior and captured it.

If the cars idling, shouldn't it stay in closed loop? It has no reason to go Open Loop unless somethings drastically wrong.



I captured two logs.

Number one is a quick idle log at operating temperature, and two is another idle log, with the dump tube removed.

1:

You must be logged in to view this image or video.

2:

You must be logged in to view this image or video.

No, the fuel should not be on all the time. If the RPM is above the preset fuel cut limit Approximately (1200rpm) with the throttle fully closed it will cut fuel and be in open loop during this operation. Your log is kinda short to make any real judgment on how well it's running, but here's a sample log that came with the software. You'll see that it's not only in open loop during the WOT pull, but it also goes open loop again and cuts fuel for the decell.

Edit: I type slow, and I couldn't find one of my elg. logs, so Gofer beat me to it.

 

[UofACATS]

Now tune your idle to that last log, what do you think? Raise your 0Hz and 50Hz slider up to 15.6% to bring up AirflowPerRev to 0.25 gm/rev. Then flirt with deadtimes a bit to get CombinedFT's to cycle around 0 (+/- 8%).

Then do a cruise log with your OpenLoopThresholds set where they are at. 15 minute drive warmed up!

I know the values are a bit off from optimal. It's here I get hesitant because I like to keep my fuel settings as calculated. I usually don't deviate from those settings because they're usually a known-correct value. However, the use of ethanol changes that assumption with varying percentages of ethanol week to week, station to station.

In your opinion what's more important, AirFlowPerRev at idle or calculated deadtime? I can see merits for both

Thanks again, I'll head out for a MAFcomp cruise.


EDIT:


Clipped a 27-minute log. (too large).

I didn't know ECMlink did this, but if you select a small portion of the log the MAFcomp tool only uses the selected information. I selected a good portion of the log, which I've attached. I can also take a screen shot of the MAFcomp recommendations from the entire 27 minute drive.


700Hz looks rogue to me??


LOG:

You must be logged in to view this image or video.

SCREENSHOT:

You must be logged in to view this image or video.

Maybe I should try swapping the MAF sensor with another? They have to be showing different values because this one shows a 20% correction across. Maybe the problem is in the MAF, although I really don't think so. Visually it looks normal, combs are fine, etc. and I don't suppose they just start operating erratically without cause.

I've had most of the fuel system apart recently. I pulled the filter to inspect/clean. The pump's obviously been pulled. The lines were all replaced a couple months ago with new fittings as well. And I also installed new vacuum lines to everything, giving every part a dedicated source. The injectors are the only part left unknown, I suppose it's worth yanking them to look for ethanol residue.


The entire point of this is to switch to SD as seamlessly as possible. I'm starting to feel like it might be worth using the default VE table for SD, and fighting with that tune instead. Although I feel like that'd be an unwise move.


Any ideas?

 

[UofACATS]

bump thanks for looking

 

[UofACATS]

Tired of spinning my wheels with dialing in a MAF with fuel trims floating around like drunken sailors at sea, I set fire to that bridge and went speed density.



I'll edit the first post 'cause this is staying unresolved. gofer and Colt4G63also I appreciate the help.

 

[gofer]

The problem with you editing your last post is that it doesn't "bump" the thread to let me know something new has been posted, and then you bumping the thread is against our site rules. Please don't do that either, just start a new post if you've got new info!

The inconsistency of the FT's could only be the MAF at this point, if you ask me. I haven't seen/done it all either so maybe someone else could think up of something else it could be. Either way though, the SD VE table will get your car running fine without any adjustments. To get a good log and adjust your VE table you'll want to do lots of closed loop driving on the interstate varying throttle position. Try going 50% throttle like you would to pass a car and build some boost... Right click the log and click the SD VE Adjust (CombinedFT) tool. Be sure to select the MAF type to SD in your MAFComp as well as locking IAT and Baro in the ECU inputs tab.

 

[UofACATS]

The problem with you editing your last post is that it doesn't "bump" the thread to let me know something new has been posted, and then you bumping the thread is against our site rules.

Right, it was a situation where the 24-hour rule worked against me. I did post new info when I had it, but it auto-merged based on those rules, and didn't bump it. (Post #15 is actually three separate posts). I knew no one would look at a thread with no new posts (even though there actually was a new post) so I bumped it.

The reason I put "EDIT:" there in bold is so the reader would know it's new information. It would have been beneficial to simply delete my last post and post again with the update.

The inconsistency of the FT's could only be the MAF at this point, if you ask me. I haven't seen/done it all either so maybe someone else could think up of something else it could be. Either way though, the SD VE table will get your car running fine without any adjustments. To get a good log and adjust your VE table you'll want to do lots of closed loop driving on the interstate varying throttle position. Try going 50% throttle like you would to pass a car and build some boost... Right click the log and click the SD VE Adjust (CombinedFT) tool. Be sure to select the MAF type to SD in your MAFComp as well as locking IAT and Baro in the ECU inputs tab.

I switched yesterday, the car went from MAF to SD and was up and running well in 30 minutes.