So now we know that you have either a failed closed evap vent solenoid, a plugged charcoal canister and or plugged lines. The canister may be soaked in fuel and/or there may be a leak at the tank/cap. Normally when the purge valve opens it pulls fresh air into the vent solenoid and through the canister. I bet we are getting mostly gasoline vapors. Disconnect the vent solenoid from its tubing remove the gas cap and tape the fill opening and let’s see what we get.


George

 

It's about to start raining on me so can't poke too much at it today. I looked at the canister on the back. I can't tell which line is the one for the vent solenoid. I assume it would be the top one? It looks like that line takes off to the front and I can't actually see the vent solenoid. I can only see the "Tank" label on one of the lines.


Edit: Pretty sure it's the top one that goes to the vent solenoid. I should be able to pop that one off and cause a big enough vacuum leak to tell me if the vent valve is stuck closed? Would just opening the gas cap and press in the fill opening to see a change or does the vent valve need to be disconnected?

 

Im trying to figure out where the gasoline vapors are coming from. Since you have trouble filling the tank I wanted to first see if its a stuck closed vent solenoid but then there would also have to be a leak at the tank and the gas cap is the most likely. If you just open the tank then there should still be a command lean fuel trim issue because then we know that there will be gasoline vapor.

George

 

So let me make sure I'm understanding you correctly here. I disconnect the vent solenoid from the charcoal canister at the back of the vehicle leaving the charcoal canister open to the air, simulating an open vent valve? Remove the gas cap, and tape over the fill hole to create a seal? What should I be looking for in my fuel trims during this test?

 

We are looking to see if the fuel trims still nose dive into negative territory. If they do then repeat with the vent solenoid and the fuel tank vent line disconnected from the cannister. If the fuel trims are still going very negative then you can disconnect the cannister from the purge solenoid line and at that point the fuel trims should become normal.


George

 

This may help for evap line ID:




George

 

Well got some confusing results.

I disconnected the vent solenoid from the canister. I disconnected it at the vent valve itself, surprisingly more easy to get to. After disconnecting the vent valve I unplugged it from it's electrical connector and tried to blow through it. I was able to blow through the vent valve. I then taped a seal at the filler neck and started the car. Here are the LTFT and STFT for the number 1 bank. Bank 2 were similar so I have omitted them.

This was at operating temperature and ran for about 10 minutes.

After this, I turned the car off and disconnected the gas tank from the charcoal canister at the charcoal canister and left the vent solenoid disconnected and unplugged as well. Here are the trims for that.

This was about 12 minutes. I have included both banks here as it's a bit different. It still trims negative with the vent valve and gas tank disconnected from the canister. But Bank #1 goes to about -4 while bank #2 goes to past -8. Towards the end of this idle, I disconnected the electrical connector from the canister purge valve in the engine bay and it only seems to make a small difference in the short term fuel trims, going a bit positive, but long terms remained where they were.

EDIT: I was not able to disconnect the evap line from the purge valve, got too dark on me and I maybe wrongly assumed disconnecting the electrical connection would have the same effect.

 

Couple of things.

We are making progress because the fuel trims in the first set of graphs are already way better. STFT + LTFT is always -5 to -8 or so which is pretty good for an older engine. The curious thing is that eventually the computer thinks that something has changed and goes to a different LTFT level but the O2 sensors are saying no nothing has changed so they offset the inappropriate more negative LT trims with positive ST trims. Was this at idle the whole time? Did the idle speed change much? You should watch the relevant sensor levels in live data when the LT trims go so negative at idle to see if any of them drift: MAF, ECT, MAP, BARO, TPS. Another possibility is that something starts delivering more fuel when warm like an injector.

How did it run at idle and as it warmed up? Did you change rpm for any of these last tests? If so, does it run worse like before?

In your second experiment that implies that you do not have a gasoline soaked cannister. It also suggests that there may be something extra going on with bank two causing it to run a little more rich. That said these net fuel trims are not bad. Again, how did it run and did you try raising the rpms?

This all suggests to me that you may have a leak in your gas tank/filler neck/cap and possibly a stuck closed vent solenoid when powered. If I am right, you should get an evap code for the leak which you dont have so I am not 100% on this. One cheap thing to try is a new gas cap. For now, I would leave the purge valve disconnected from the cannister and go for a test drive and see how it runs. This should eliminate the source of the excess fuel vapors and correct the rich condition if this is the only problem. This should also eventually generate an evap code (large leak), Again watch those sensors because a rich condition can be a actual extra gas or it can be a wonky sensor that confuses the computers initial fuel delivery calculation by asking for extra fuel when its not actually required.


George

 

I don't think it would be an injector considering both banks go super rich when everything is connected like it should be. And the fact that I switched injector sets and had the same exact issue. I do realize that both sets of injectors could be bad, but it would be very odd for both sets to fail on both banks in the exact same way.

With just the vent valve disconnected from the canister and the LTFT in the -10, I could smell a stronger "exhaust" smell. I did hold the throttle at 2k towards the end of the second test, but my my RPM logging did not take, so I can't tell you exactly where, but towards the end before I unplugged the canister purge valve.

If the gas cap was leaking, wouldn't the be a vacuum leak and show as positive fuel trim?

So I should go ahead and connected the vent valve and tank back to the charcoal canister, but disconnect the purge valve from the charcoal canister? With everything connected like it's supposed to be, my fuel trims go -25 after about 10 minutes of driving.

I'll go ahead and add all those sensors to my logging. I have them on my live view, and I never noticed anything going out of whack, but maybe something will show on a log.

EDIT: The throttle felt way better with everything disconnected. With everything connected like it should, my RPMs would dip when I got to about 2k and I would have to give it more pedal to get back to 2k and hold there.

Edit Edit: And with the vent solenoid disconnected, and the tank disconnected from the charcoal canister, how was I not trimming more positive if I essentially creating a vacuum leak?

 

Bad injector was just an example. I am saying that something bank specific is also a little off which might be worth a look after we solve the -30 fuel trim problem. Injector, ignition, compression. This may be unimportant in the scheme of things once the big problem is solved.

No just leave everything disconnected and go for a test ride with the purge valve connected to an open pipe and see how it runs. I am trying to see if eliminating the fuel tank and cannister vapors feeding the purge valve solves this problem.

In normal operation the purge valve opens to pull some vapors into the intake along with fresh air from the vent solenoid. The computer knows this and compensates accordingly. My theory is that it is not expecting a high fuel vapor content which I believe we have here..

If the LTFT went more negative with increased rpm then forget that whole discussion. The computer stored some hefty negative trims before at rpm that it does not need anymore and will get unlearned. In the meantime the STFT offset the now obsolete large negative LTFT's at higher rpms.

You said that this started after a fill up?

George