Hi all. I’ve run into a peculiar problem with my RalliArt lately and so I’m presenting it here to hopefully get some feedback and assistance. I realize it’s not exactly 100% tuning related, but this is generally the place where the more knowledge savvy individuals tend to frequent, and it is engine-management based, so here goes.

Disclaimer: I (admittedly) do not currently have full-time access to a Tactrix cable and am currently using a handheld OBD scan-tool and physical gauges to monitor my readings, so my data has to be presented verbally for the time being. I have since ordered a Tactrix, however, and will be able to update with logs and other data as soon as it arrives.

This all began a couple of weeks ago, up until that point the car has been running well within specs and I have not changed anything on the setup for at least the last 6 months, other than what I am about to talk about. So I was cleaning out the garage and decided to re-install my subwoofer (which I have swapped in and out of the car frequently, for track days and whatnot, for the last 2 years). I had the battery disconnected briefly to install the inline fuse power wire and then plugged everything else back in. Drove the car for a day and all was well. As mentioned, this is the only potentially changing element on the vehicle is that the battery was disconnected and the subwoofer was installed.

Then, I went on a longish trip (85 miles) and had no issues on the way, but coming back we stopped for dinner and when we came out of the restaurant the car had a bit of a rough idle on startup. I thought nothing of it, as the car tends to run a bit “off” when not warmed up fully, and we went on our way back home. I noticed as we entered our town that every time I would come to a stop, the AFR would move toward rich (12-13) but the car was still behaving normally, so I dismissed it as a possible boost/vacuum leak along with the rapid warming of the temperature outside over the previous day or two.

The next day, when driving to work, the car would consistently go rich every time I would come to a stop (once again) and there was even one occasion where it went all the way full-rich (9 on my AEM gauge) and the car stalled.

When I got to work (in a shop thankfully) I checked for any vacuum leaks and found none. I then disconnected the battery again (this time leaving the subwoofer unplugged, just in case) and performed the throttle-position relearn procedure since, after discussing it with Devin (Boostin Performance), we figured that since it was only happening at “idle” it was worth a shot. I realize that after doing so there is still a bit of a learning period for the car, but a couple days later and it still had the same issue. I'm going to go back and spend some more time with Devin at Boostin soon, but our schedules are both so busy that it's tough to get together and figure this thing out, which is why I'm also presenting it here.

I have yet to perform a boost-leak test, as I haven’t really had time, but that is on the slate for (hopefully) this upcoming Friday 3/23.

I did manage to go on one ride with a Tactrix cable and tried to monitor the condition to see what was up. It appears that when coming to a stop, the STFT just stops working, completely. STFT is at 0.0 and holds there, even though the wideband is reading anywhere from 11-13 and my front O2 sensor is showing a corresponding reading of right around 1.0v (O2 voltage switches normally when under any other driving conditions, mirroring my gauge AFR readings and maintaining stoichiometric 14.7ish). Usually after about 10-15 seconds (sometimes up to 30), STFT starts moving around again and the car will attempt to correct itself based on O2 readings and can usually find its way back around stoich. Occasionally, though, the car will briefly go even further rich and start to stumble but, unfortunately, I haven’t been able to catch it in-time to see if this is happening while STFT is trying to correct the AFR or if it’s during the time when it is inactive.

The only other discrepancy that we were able to find in the log is that while stationary, the MAF signal was moving around erratically (anywhere from 0.7 to 1.7) which, when you look in the table on ECUFlash, changes the corresponding table value from around 1-2 at ~0.7 to a few hundred (iirc) at ~1.7. (For the record, yes I have an intake with an oversized MAF tube diameter and yes I had the proper scaling done to accommodate the increased size last year when I installed it and I haven’t had any issues up until this point) I realize that this voltage irregularity may have a dramatic impact on fueling and is probably being compensated for while STFT are active, but I am wondering why STFT is inactive right after coming to a stop. Regardless of the STFT issue, based on these readings, I am also planning on swapping out my MAF sensor with a known-good part to see if that corrects the issue (also on Friday) as well.

This condition does not occur during normal driving or cruising where AFR stays around stoich, and at WOT the AFR holds strong right where it should in the 11’s.

As I said, I will update with what little progress and updates I can give, but I would love if someone had some other input or ideas regarding this issue. I have done plenty of searching regarding idle issues, fueling issues, MAF scaling issues, you name it, but have come up relatively empty-handed. The one other thing I saw was an instance of a bad exhaust manifold gasket, but I have yet to venture down that avenue as I think the resulting condition was actually opposite mine and I would like to avoid that labor-intensive headache if at all possible.

I think that is about everything there is to say at this point, but if there is any other information or data points that I can provide via generic OBDII readings, please let me know and I will retrieve them, otherwise I will hopefully have some logs generated after this weekend as long as my cable arrives in time.

Thanks in advance.

 

Mod list:
Custom SRI with CBRD velocity stack MAF housing
Nisei 1pc UICP
CBRD RalliArt-specific FMIC
UR LICP
CP-E short downpipe and catback
AMS test pipe
Grimmspeed 3-port
Mini battery
ECUFlash tune on 93 octane
Analog AEM UEGO
Analog AEM boost gauge

 

Maybe the mini battery its getting near the end??? When your driving the alt is able to keep the battery going enough to keep all your electrical systems in check but at idle maybe its just not enough and for some reason dropping out engine mgmt rather than sub etc?

Or maybe the relays are toast.

I'm having a similar issue, rough startup after leaving the car for a while but then once i get some charge into the battery its all good. I also found the my fuel pump relay was toast( shook it and sound like a salt shaker) got the replacement ones from SSP

 

Possible bad MAF/MAP scaling. Car learned trims before. Now you erased them by disconnecting the battery and it's re-learning.

 

So if I originally scaled the MAF on learned trim values, resetting it could cause a discrepancy? Would it be a good idea then to disconnect the battery again and start from scratch (or would it make a difference, at this point, since I don't have stock intake maf values to multiply/compare to?) Collect all new log data and plug it into the scaling calculations (I think there was an Excel spreadsheet that someone put together)?

Also, I'm still concerned as to why STFT is disabled/inactive right after coming to a stop when fuel cut condition ends. Is that normal? I never really had need to monitor that before, but now that I'm watching it, it seems odd to me.

 

No codes at all?
Ask Devin to look at two things:
1)The cross-over for Open to Closed Loop (rpm and hz)
2)Re-scaling you MAF not on tube size (i.e. 3" to 3.5") but by actually logging STFT, LTFT Current, and MAF Voltz.
My car idles right at 1.4-1.5 voltz. If you car is idling at 1.0, that means with your velocity stock it is that much EASIER for the engine to draw in air. With something like a velocity stack I would want the longest straightest possible piece of tubing, and place the MAF sensor a good 9" away from it.
One day when I was at RRE the same thing happened to an 09 Ralliart. Guy was fully tuned for a few months, everything was fine; came back to get some gauges installed, car was perfect when it left. He took his girl to dinner about 70 miles away, and when he got back in the car the idle was going bonkers. Super rich, over correction, super lean, over correction, super rich, etc, etc. Chalked it up to something in the LTFT memory that was erased when the negative was disconnected.
The only thing that makes no sense is that it takes 15-30 seconds for your STFT to start adjusting; that s/b like 2-3 seconds.

I have a theory that our cars store a "anticipated LTFT % adjustment value" on a per rpm basis. I can't prove it (damn I'm turning in Mulder), but I've noticed some weird behavior when making significant changes to airflow/fuel. The car "should" only take 4 key cycle to re-learn, but there have been times where certain rpm conditions will prevail for a week or two....weird.

 

post up an idle + cruise log

check your FPR vacuum hose...

 

The one thing I just can't wrap my head around is that, as I've mentioned, I've had the subwoofer in and out of the car repeatedly (which means connecting and disconnecting the battery repeatedly) and I've also had the battery completely out of the car for much longer periods of time (like during SST fluid service) and this has never happened. Sure, there were always some hiccups and getting into the right groove, but nothing this severe. I do believe this issue is related (in some way) to the MAF, but that's why I don't think its necessarily a scaling issue, nor related to having the battery disconnected and relearning. If it is, then why this time and never before?

Steve, at idle, my MAF readings are fluctuating so erratically that it seems like it HAS to be the cause, but I don't know if that' secondary to another issue or what. I'll have all those numbers more readily available in the next couple of days with some full logs, as I'm sure that will help over my rambling descriptions.

tephra, FPR hose has been zip tied since forever, lol. No other loose hoses or obvious vacuum leak areas. I realize that it's easy to sometimes overlook the simplest of solutions, and the truth is you don't know me from Adam, but just to give you a bit of background on myself (in-case you're worried about getting too technical or something along those lines) I have been a dealership tech for the last 6 years with 4 of those being with Mitsubishi. When it comes to basic drivability and factory-spec diagnostics where I have known reference values and physical mechanical skills I am very capable, it is just that I am still learning the intricacies of diagnosing a modified and performance-based platform where OEM specs and concepts are thrown out the window (Trust when I say that I have seen more than a few people here, including yourself, who's knowledge of finite details on these cars far exceeds anyone I've ever known or worked with in a dealership setting). Boost-leak testing tomorrow. I'll get those logs up asap as well.

Thanks all for the input so far. Once this all gets figured out, hopefully this may benefit someone else in the future, as well.

 

Hmmm, one thing to consider, relative to the battery, is possible you drained the reserve charge in the system. IIRC when pulling the battery there is still a "charge" there for the ecu/whatever (atleast it was for Subies). Perhaps this last time you managed to burn out that reserve charge, thereby completely wiping all saved trims.
For instance on Subies to re-set the fuel trim manually you disconnect the battery, and pump the brake several times.; this uses up the remaining charge in the system.
One thing that sounds really scary, is you saying that the MAF registers at less than 1.0. I didn't think that was possible, as anything less that ~1.2-3 is 0 (and I've never had a logged condition less than 1.27).

 

I tried getting some logs using the Tactrix standalone method and am getting some wonky values on a few parameters, so I think I may have some incorrect addresses in my logcfg (specifically IAT and MAT, but with those being wrong, I'm not sure if I have the proper calculations for the others, either, so I feel like posting those logs up would be frivolous). I'll try to get some standard "plugged-in" evoscan logs and see if that works out better.

 

tephra, I got a decent log with EvoScan on the ride home from work today, but I think I have the wrong addresses for IPW and IDC since it was only displaying 0 values for those. I don't know it that is vital information for this type of diagnosis, but I'd imagine that more data is always better, so perhaps I'll have to figure that out and log again before I can progress any further. I think I can solve that issue, though, since my Standalone failure this morning was able to at least read those successfully. Basically, if I'm understanding correctly, if I use the addresses from the logcfg and replace the "0x" with "23" before the other 6 digits in the paramid, it should route me to the right location in EvoScan, yes?

I also think that I have the IAT and MAT addresses flipped, since I'm showing near-ambient MAT and mid 100's IAT (though it could just be my super hot SRI and super awesome FMIC right? lol), but what do I know? I'm still new at this.

As far as the log I have, it does at least demonstrate the inactive STFT well after I go off-throttle and come to a complete stop and it also has the WIDE RANGE of MAF values that I was talking about. If you're interested in seeing what I've got, I'll gladly send it your way. I say that because the log I have is relatively huge (1.9mb) and even when compressed in .zip format, it still exceeds the site maximum, so is there an email address I can send it to?

 

I also plan on setting up wideband logging with the pin8 method this weekend, so I'll have that to add to the log data as well.

 

Ok, I think I finally found something substantial. Here's a short log from the other day driving in stop-and-go traffic: traffic 3_29_12.zip

Granted, my IAT values are incorrect (I'm still trying to find the proper address/scaling) and my MAFv values are calculated incorrectly (right address, wrong scaling), you can see that the MAF is either very slow to change, or completely zero's out in the sections where the STFT is disabled and the AFR goes rich. Once the MAF starts moving, the AFR tries to straighten itself out and STFT becomes active again.

Before, I couldn't get it to do this while logging, which is why I was so stumped because even though the values didn't seem right, they never zeroed out like this.

Trying a MAF swap this weekend since I didn't get to last week. I'm thinking that should clear it all up (hopefully)




Also, for logging MAFv in standalone, I'm using the same info from EvoScan, logcfg looks like this:
paramname=MAFv
paramid=0x808F58
scalingrpn=x,1024,/,5,*

EvoScan shows:
Display: MAF Volts
Request: 23808F58
Function: x/1024*5

Values appear to be correct in EvoScan but not in standalone. Does that look right? Or is there something I should change in the formula to make it work properly?

Thanks.

 

Ok, so not that anyone is really responding to this thread, but I'll use it as a journal to document my progress and findings and if anyone wants to chime in, feel free.

I swapped out the MAF with another car, and the problem still occurs, though it did seem to correct itself a tad sooner, but I still have never seen anything like what is happening. Very strange indeed. The car itself doesn't actually appear to be running poorly most of the time when it is happening, either, just running in open loop with super rich fuel trims.

For S&G's, I also tried swapping out the Engine/Fueling relays with ones from the other car and there appeared to be no change with that, either, though mine DID have a slight rattle as though something inside had become dislodged, but only if you give it a good solid shake. I may end up buying another set just for insurance sake.

One other thing I found today that I was overlooking before was that while my boost gauge shows a fairly consistant 20(ish)"Hg vacuum at idle, my OBD scanner shows that the MAP sensor is only registering 9-10 inHg. I wasn't sure about this by looking at the logs because I believe the boost/vacuum readings are all PSI and you need to multiply by 2.xx to get a valid inHg reading, but for some reason it kept registering in my head the other way around and I kept confusing myself. Looking more closely, I see that while steady-state cruising I am still showing nearly identical values as I am getting at idle (within 1-2 units, whatever they may be, instead of the gauge reading actual ~5PSI/10inHg difference).

Damn is this frustrating...

 

But then why would I be registering "0" values for MAFv, a multiple failure issue, or maybe the PCM is just taking a poop on me...