RomRaider

I've been chasing some fueling problems recently. Running ID1050X's, AEM 340lph hardwired, stock FPR, on a V7 EJ207, been tuned on speed density for about a year now. Recently it began stalling when letting off the gas and coasting to a stop which I suspected was the front O2 going bad. My cruising AFR's have also been pretty erratic. They hover around 14.3 and jump around between 13.8-14.7. No amount of closed loop VE tuning seems to be able to bring it up to a stable 14.7. I've attached a short data of some cruising that shows the rapidly changing O2 sensor readings.

Before buying a new front O2 sensor, I tried running full-time open loop. This made the AFR more stable during cruising, but my AFR's were way off target throughout most of the cruising range, going anywhere from 13-16. This makes sense as there is no AFR learning. I tried tuning the VE table in open loop but I couldn't manage to get it on target everywhere throughout the table even after a few revisions. I was even getting drastically different percent errors for AFR at the same RPM and MAP values which was confusing.

I've since replaced the O2 sensor with a Denso unit, and now it doesn't want to stall as much, but the cruising AFR's are still erratic and rich as can be seen in the log and I'm out of troubleshooting ideas. Is it normal for the front O2 sensor to have rich dips like this while cruising to the point where you can feel them?

Some thought's I've had so far:

-The hardwired fuel pump is overrunning the FPR during cruising (I already know it does during idle)

-My injectors need cleaning/flow testing

-The minor manifold-to-uppipe exhaust leak I've had for years is causing incorrect O2 sensor readings

what fuel pump?

check the fuel pump output graph for flow and pressure.

they are rated generally for 3 BAR of fuel pressure, but remove 10 PSI for idle, and you're looking at 33.5 PSI fuel pump pressure.

not all fuel pumps, but some of the larger ones, will override the stock FPR, if the flow at idle is greater than 274 to 300 LPH.

DW300, AEM340, can be adjusted without much work to not override at lower MAP
anything bigger requires hard wiring because of the amperage being more then the stock fuel pump controller can handle, and by the time you do that, you need a larger FPR.

if the pump doesn't take more than 15A full out, then I'd go back to the primary fuel pump circuit and control the pump against overriding the FPR.

Yes I did just replace the fuel pump with a new AEM 340lph pump which takes up to 16A I believe. I hardwired the pump, bypassing the FPC, so it is definitely overrunning the FPR. I can actually hear the FPR open up after idling for a minute. Can the FPR be overrun during cruising as well, or just idle? If so that is likely the issue. My next step was to upgrade all the stock fuel lines and add an aftermarket FPR.

I'm still struggling with this issue if anyone has suggestions. I rewired the fuel pump for the 3rd time to be switched by the FPC so it runs at 33% and 100% now my FPR is no longer overrunning, but the hectic front O2 sensor reading is still there. For reference, this car ran great on this map and the fueling tables were dialed in perfectly, then the fuel trims suddenly went crazy. Every time I try to retune my VE table I end up pulling so much fuel that the car runs terribly and chokes during cruising from being so lean. My cruising fuel trims are around -15%, so I pull reduce the VE table in those cells by 15%, and log again. The car bucks because it's now too lean, yet my fuel trims are still calling to reduce fuel. My AFR's appear to be on target, with the front O2 and my wideband roughly matching up, yet it still wants to pull fuel for no reason. During cruising the car has very minor hesitations that you can feel, like it loses a little power for a millisecond every few seconds. It run fine during WOT.

I've attached a log of some cruising after I tried retuning VE. When I retune it, my nicely smoothed VE table ends up having huge peaks that don't make any sense in the cruising cells and the car runs poorly. I also tried logging a bunch of different sensors to see if anything was behaving weirdly, and everything looks ok, although the front O2 does jump around a lot more than my wideband. The only other thing I noticed was that my FFB would randomly jump down into the 13s and into the 16's while cruising for a a second, and you can see the AFR jump a bit when that happens. I'm thinking this is causing the brief hesitations I'm feeling during cruising.

For background, I recently replaced my fuel pump with a new AEM 340 unit, and changed the fuel filter out with an OEM equivalent from a local auto parts store. The car was running fine with the pump hardwired, then after a few weeks the cruising AFR's started to go rich and jump around from 13.8-15 erratically. Suspecting the front O2, I replaced it with a new Denso unit. This didn't change a thing. The idle AFR also rose a bit from 14.7 to 15-15.3. Could this all possibly be the result of a pre-O2 exhaust leak?

Pulling my hair out over this issue. Any help would be appreciated. ROM and logs attached.

I have been having unstable idle problems since I have had the car. I was watching the stock o2 sensor at idle and noticed it jumped up and down alot. I ended up comparing the stock o2 with my wideband and noticed it wasn't accurate. I rescaled my o2 sensor to reflect the output from the wideband taking into account the delay from being farther down stream. The idle is rock solid now.

I attached pictures of my scaling compared to stock. I would suggest verifying your o2 scaling with your wideband and adjust as needed. You might zero out your closed loop fueling /load comp table and adjust your closed loop fueling target to really fine tune it.

revisiting this post, I remember going over this issue on an ej207 with ID1300x.

minimum injector pulsewidth, on pump-gas.

above a certain temp, you're going to run into a problem where on pump gas, the vehicle would want less pulsewidth than the injector can be open at a bare minimum.

if requested pulsewidth is less than minimum, the injector can only do minimum.

also, fino, you're trying to do speed density (manifold stuff) with temperature taken from not in the manifold. you can't do manifold air density with temp taken from before a two temperature changing devices. that's like trying to plan re-entry on earth using Mars' atmosphere.

any time the temperature delta changes between the turbo inlet and manifold, your VE table will scale differently than the last time.

the only time that VE changes will be accurate is if your temperature delta is the same at all times, which is impossible - and cramming a bunch of extra VE into the higher cells where your temperature delta is the worst, is technically the wrong way to compensate for inaccurate density from bad temperature reference.

having calculated airmass be as accurate as possible is the way, the only true way.

Makes sense. I'm using the default value from the Cobb spreadsheet for these injectors which is 0.308 ms. I also have a GM IAT sensor welded up about 4 inches from the throttle body. I imagine the compressor heats up the air a bit but my manifold temp should be pretty close to the IAT reading, especially during cruising where the flow should be relatively in steady state.

Did you find a solution to this?

I'm having a similar problem with my car.

Closed Loop Fueling target jumping from 13.6-14.7.
When targeting 13.6, Fuel learning dives to pull fuel out, but Fuel correction still targets 13.6 resulting in a stale mate. This happens at idle and cruising.
It seemed to start happening after my car was EcuTek tuned. I've since come back to Open source and it hasn't solved the issue.

Yes, in my case it was a pretty bad exhaust leak right at one of the header flanges on a recently installed HKS header. I also replaced my wideband sensor because it was failing, but I think that may have been related to the fueling problems caused by the leak.

Great news!

So you think that the leak was causing the ecu to panic and choose a richer AFR? and the leak also caused the wideband o2 to fail? Was the OEM front o2 sensor effected?

Did you notice any of the strange learning patterns i mentioned?

I've been having this issue with my car as well... kinda thought it was just normal, didnt think it was supposed to be that stable in closed loop. Always fluctuates whether reading in RR or on my Wideband.

I also did rescale my Front O2 sensor a little to guide it closer to the wideband's readings.

I noticed that on occasion, the ECU doesn't do a great job in correcting in the right direction, or at least quickly enough. I've seen the car rich by half a point, still adding fuel by correction for a few seconds or so, and the same with lean. Typically it's only a half a point in either direction from around 14.7 but it's never a stable number in closed loop and it's always changing quickly.
I have done several rounds of MAF scaling as well. I haven't changed any sensors yet, O2 or MAF since I got the car, but I've only put a few hundred miles on it so far myself. I do believe I could use a MAF sensor though because my scaling at idle is much higher than it should be to acheive correct results. Stock fuel, but with AEM 320lph, CAI, FMIC, VF52, catless up and dp.

EDIT: just re-read the post and saw you mentioned your -target- AFR was changing. I don't know if my actual closed loop target AFR is changing or what, haven't logged that.