Need some help, don't know what I screwed up

[Ajzride]

So my dad is coming out next weekend for the holiday, and since he only comes about once every 2-3 years, I really wanted to be able to take him for a go-kart ride. However, I'm needing some help getting the car back to running again, I'm not sure what I've screwed up because it was running fine, and now it won't run at all once it warms up unless you keep your foot on the gas.

It seems like this started when I put the exhaust on, but since I haven't been driving the car much, I'm not positive. I did have some problems with it running rough a few months ago, and cleaning the MAF appeared to have cleared that up.

What I've do so far:

Since the car is DBW, it has an idea re-learn routine you can go through, I did that this morning. The car started off rough and then smoothed out until it got to 180, and then it stalled and won’t restart until it cools off.

The MAF seems to be working, I can blow a fan up into the intake and it will increase in voltage (as read by the ECU and reported through ROMRAIDER). It reads about 1.3v at idle, jumps to 1.8 if I blow air up in there.

The O2 sensor is brand new. Lambda starts at 1.0 when cold, then once it has warmed up (about 120 engine temp, but its probably based on sensor temp) it drops to 0.76 and just stays locked on there. Once it was solidly locked and not moving, I used a propane bottle to blow some propane in the intake, and the O2 jumped up to 0.82, and then started moving around until it closed in on 0.95. I think the O2 is working based on this, but I'm not positive. I'm not sure if the fact it moved off of 0.76 and started working towards 1 was related to me blowing propane in there, or just coninciden based on temperature map and moving from open loop to closed loop.

Could my map be that far off after adding exhaust, or could be a sensor issue? Knock count is 0 on all 4 cylinders, even when running rough.

Thanks guys.

[q4stix]

Can you take a look at the fuel learning to see if it's trying to adjust the fuel a lot? It could be something as simple as it's trying to compensate for a vacuum leak and ran out of adjustment

[Ajzride]

Actually, for now, it just flat ran out of fuel.

I'm a complete moron who failed to consider the coincidence that I might have ran out of fuel within a few seconds of swapping from warmup to closed loop. I've got 5 gallons in it now and it is running again. I'm letting it cool back down to cold with the battery disconnected so I can re-do the idle learn and see how it behaves after that.

Thanks for the prompt response.

[q4stix]

Even better of a solution! Glad to hear that was probably the case.
Enjoy the drive with your dad!

[AZPete]

I did that several years ago, but worse. I whined to my wife that my 818 wouldn't start and she doesn't know a lug nut from an air filter, but she asked, "did it run out of gas?" It turned out she was right and I'm still hearing about that day.

[Ajzride]

Today I started it up so I could turn it around and back it into the garage, and it was running like crap again. Literally couldn't finish turning it around before it died and wouldn't start again. Unplugged the MAF and it fired up and ran like a champ. Definitely something wrong with my closed loop sensors. Need to figure out which one is the culprit, this is annoying.

[Just puttering]

From ford experience (my wrx runs well enough i have not messed with it) the lambda should not just be stuck on one number. The computer should be constantly varrying the fuel so the number should be moving around some??? When cold, i assume the lambda is 1 because the computer does not care about the o2 numbers untill things warm up. O2 sensors do not work untill they reach about 600 F, also, i got a brand new o2 from the auto store and it was DOA. It just floated around .2V!

Can you force it to stay in open loop? Unplugging the MAF should force open loop but may hide the real culprit!

Not to long ago, i fought with the computer and not starting ....... Untill I added more fuel!!!

[Ajzride]

Thanks for the tip on forcing open loop, I did some reading on romraiders and it looks like I can do that by modifying the ROM file. I will try that this evening. While on the forum I also read this is frequently caused by a vacuum leak. I did go around the engine with a propane bottle and didn’t seem to find one, but I do remember removing the TMIC to take pictures for another forum member, perhaps I don’t have it back on properly. Will dig into both of these after church.

[Just puttering]

Note: they say an exaust leak is just as bad as a vacuum leak for the efi. I am interested if it continues to run ok when forced in open loop

[Ajzride]

I decided not to dig into this today because I couldn't find an IACV gasket local. If I'm going to yank the TMIC off again to make sure it is installed correctly I want to clean the throttle body and IACV while I am in there. That should come in Tuesday and I'll tear it open then.

[Ajzride]

It turns out I didn't need to wait for an IACV gasket, because I don't have an IACV. Working on a 2006 can be difficult, the WRX actually had a newer engine than the STi, and lots of sites have wrong data. Anyway, I did clean the throttle body really well (which stopped the annoying buzzing with the key on but engine not started) and discovered that when I had taken it all apart to take some pictures of connectors, that I had not put the gasket back in where the BOV connects to the TMIC. After cleaning the TB and making a gasket for the BOV, I put it back together and performed the idle re-learn again.

- The engine started off pretty choppy in warm up (status 7), with the AFR at 14.7 and the MAF around 2.3 g/s. Once the O2 sensor reached temperature, it immediately plummeted to 11.14.
- The engine swapped to closed loop (status 8) and smoothed out quite a bit, but the AFR stayed at 11.14.
- The engine swapped to open loop (status 10) for some unknown reason, and that is when it got really bad. The car could barely stay at idle and was so rich it was blowing black smoke and making the whole neighborhood stink. AFR stayed at 11.14
- After a few minutes it swapped back closed loop and smoothed out nicely. AFR stayed at 11.14.
- I shut down for 20 minutes, as instructed and started it back up.
- Started in warm up (status 7) and AFR 14.7. Once the O2 sensor hit temp, it dove to 11.14 and stayed there.
- Engine swapped to closed loop (status 8) and ran fine, AFR stayed at 11.14 for about 8 minutes, then started to float up around 12.
- Shutdown after 10 minutes and let sit for 20 minutes.
- Started back up, warm up, AFR 14.7.
- Once O2 reached temp, it dropped to around 12.5 then started working back towards 13 - 13.5.
- Car stayed in closed loop for 5 minutes idling fine, went for a drive.
- Car ran well, popped in and out of closed loop and open loop, never died when coming to a stop, no bucking or surging.
- AFRs were between 13 - 14 mostly. Once hard throttle pull they jumped to 15.69, and at one point they dropped to 12.25, but stayed in the 13-14 under normal driving.


Thoughts

I think the 11.14 is as low as the O2 sensor I have installed will read, and the car just runs really rich during the idle learn process. Once the idle learn was completed the car seems to run pretty well. I'm not sure if the missing BOV gasket was really causing any issues or not, but I think so, because after the last time I tried to learn, it still wouldn't drive well enough to go up and down the driveway.

Hopefully all this information will help someone in the future.

[Bob_n_Cincy]

AJZ,
I don't know if these charts will help.
This is mt 04FXT 2.5L TDO4 single AVCS with a stock tune. (198 WHP)
If you look at the very bottom of the second chart. There in a line that indicates closed loop. As you can see it jumps out of close loop when I nail the throttle.
Let me know if you have any question?
PS: Charts were recorded using BTSSM phone app with a Bluetooth OBDII adaptor.

bobs sat 1st pm run.jpg

1min.jpg

[Ajzride]

Thanks Bob, that gives me a good comparison when I start trying to dial things in. I really need to rescale the MAF, but that is hard to do without a dyno or being able to take a long drive. Hard to take long drives in a go-kart, so I might need to track down a dyno.