[Efthreeoh]

Originally Posted by gT-BMW

thats my homemade 12v smoke machine sir

That's what I thought. Innovative. Most use a paint can and blow the lid off and scald themselves with hot mineral oil. Nice unit.

[aamodtpw]

Update to my earlier post: Dealer identified the charcoal canister as the culprit in my car's case.

[kkasson]

Originally Posted by kkasson

I also finally decided to do something about this - I've had the gas nozzle problem since I got the car but I've never cared enough to do anything about it. My canister was also stuck but after a few quick smacks on a plastic shelf it started to loosen up.

I went back to a pump today that used to cause this problem - unfortunately it still did the same thing. The pump shut off when then tank had about 9 gallons. I started it a couple more times and made it to about 11 or 12 gallons in the tank before I couldn't get it to stay on any longer.

[E90Daddy]

Please do not TOP OFF your tank! The reason is if you do liquid fuel will go into the charcoal canister and ruin it and you will have a check engine light and costly repairs.
I am a certified Tech for VW but the same stands true for all cars.

[rt7085]

I have been studying as I suddenly have this issue now this and it is likely what @E90Daddy pointed out already. I have a tendency to fill to the top in MA.

See the pictures below. When you are refueling, the gas cap is off and so the filler pipe is open. The air in the tank is displaced by the fuel coming in and the air just escapes via the red arrow. However, when the nozzle is fully inserted the filler pipe is largely closed off again and so the displaced tank air and vapors must follow the blue path into the charcoal filter. This is the reason why the vent valve goes up to the filler neck, to catch the vapors before the release in the atmosphere. If this line or the filter is blocked, the air cannot escape and thus the filler pipe backs up with fuel thereby shutting the pump off. You cannot really see the backup as the nozzle blocks it from view. If you pull out partly, air can escape a bit more and you may get around the shutoff issues. It can also be seen that when you overfill, the blue vent pipe can take on fuel as the filler pipe backs up to the top and this likely ends up in the pipe/canister. This likely coagulates the charcoal pellets and blocks air flow in the end. Thats why banging the filter canister likely helps as you loosen the pellets. Not sure if coagulation and excessive liquid fuel absorption permanently damages the charcoal filter in the long run; but it likely might. And thus the filer must be replaced. It may even get so bad that the DTML pump takes on fuel; it imagine it might break the DTML pump. I will probably just replace the filter soon and be done with it.

So... no more overfilling for me. This is poorly designed in my opinion.



Its noteworthy that there are actually two vent paths in the first diagram not reflected in the bottom diagram. So this is the green and blue paths. But it inadvertently proves that it must be a canister blockage as otherwise the green path would be the bypass for the blue path.

The light blue lines are the purge line to the engine (via engine vacuum generated while running) and through the DTML pump leak test dust filter to the athmosphere. But engine vacuum is the primary way. Burn the vapors.

[TheMidnightNarwhal]

Originally Posted by rt7085

So... no more overfilling for me. This is poorly designed in my opinion.

Isn't this how all modern fuel refuel work?

And why in the hell would you want to overfill your tank. Once it clicks off just stop, you tank is full enough.

[type-dRew]

Originally Posted by TheMidnightNarwhal

Isn't this how all modern fuel refuel work?

And why in the hell would you want to overfill your tank. Once it clicks off just stop, you tank is full enough.

It is, this system is identical to my old 2000 Ford Focus, which also clicked off early until i replaced the charcoal filter. Also identical to my old 2013 Sonata turbo, which did not have this issue but would sometimes fight me if I started filling up 'too quickly' after turning the car off. Let it set for another 30 seconds and its fine. The system basically needed to purge the vapor longer when it was lower on fuel.

No clue why you would purposely fill the neck up with fuel. It's gas, theres more of it available when you run out, the extra couple miles don't matter.

[KillaManjango]

Originally Posted by TheMidnightNarwhal

Isn't this how all modern fuel refuel work?

And why in the hell would you want to overfill your tank. Once it clicks off just stop, you tank is full enough.

I think most people overfill to get a nice round number as the final price...I am occasionally guilty of this as well....grinds my gears when I see it stop at $29.72

[NFiftyWon]

Suction is applied to this pipe using a venturi. When the tank is not full, air is being drawn through the hole by the vacuum, and the air flows easily. When gasoline in the tank rises high enough to block the hole, a mechanical linkage in the handle senses the change in suction and flips the nozzle off.

Its almost certainly a fault of the handle.

[vilord]

It will also shut off early if one of the purge valves sticks, it isn't always the charcoal filter (but that is common at the age of our cars).
On mine, the purge valve that failed was mounted to the intake manifold under the throttle body (total PITA to get to). There's an ISTA/INPA test for it that causes it to open/close/open/close in quick succession so you can listen for it clicking. Mine did nothing.

I got so used to it shutting off early that I've started always topping it off just in case it wasn't really full.

[gT-BMW]

Originally Posted by vilord

It will also shut off early if one of the purge valves sticks, it isn't always the charcoal filter (but that is common at the age of our cars).
On mine, the purge valve that failed was mounted to the intake manifold under the throttle body (total PITA to get to). There's an ISTA/INPA test for it that causes it to open/close/open/close in quick succession so you can listen for it clicking. Mine did nothing.

I got so used to it shutting off early that I've started always topping it off just in case it wasn't really full.

the purge valve has nothing to do with the gas nozzle shutting off. is just an emissions device. its designed to stay closed when engine is off. when the engine running it opens and takes the vapors out from the charcoal canister.

if it sticks open when engine is off you would have a CEL for it most likely evap large or small leak.

also i would like to add if it was sticking closer to the closed position you would also have an erratic idle. i know this cause this was one of the issues i was battling with. many told me it was the vanos sticking but it was this part that messed me up. got smooth idle now i dont even know the engine is on.

[vilord]

Sorry, the valve I meant to refer to is the Tank Vent Valve, which is technically a purge valve.
It is #9 on rt7085's second image.
It opens when the engine is off to allow the tank to vent though the charcoal canister to the atmosphere.

the CEL would come on if it stuck open when the engine was *on*.

When the engine is off, there is no vacuum or evap checks because the engine is off.

Thanks for telling me what's what though.

[rt7085]

Originally Posted by vilord

Sorry, the valve I meant to refer to is the Tank Vent Valve, which is technically a purge valve.
It is #9 on rt7085's second image.
It opens when the engine is off to allow the tank to vent though the charcoal canister to the atmosphere.

the CEL would come on if it stuck open when the engine was *on*.

When the engine is off, there is no vacuum or evap checks because the engine is off.

Thanks for telling me what's what though.

I think @gt-BMW is still right. You are talking about the same valve, the "tank breather valve". It must be shut when the engine is off. The is because the DTML pump activates after a drive and pressurizes the fuel tank. It then measures a pressure drop through the current the DTML pump is drawing trying to hold the pressure. Would there is a leak, the current draw of the motor would be too low. If the purge valve would be open with the engine off that pressure could not build and you would get a CEL.

[rt7085]

Originally Posted by gT-BMW

the purge valve has nothing to do with the gas nozzle shutting off. is just an emissions device. its designed to stay closed when engine is off. when the engine running it opens and takes the vapors out from the charcoal canister.

if it sticks open when engine is off you would have a CEL for it most likely evap large or small leak.

also i would like to add if it was sticking closer to the closed position you would also have an erratic idle. i know this cause this was one of the issues i was battling with. many told me it was the vanos sticking but it was this part that messed me up. got smooth idle now i dont even know the engine is on.

Interesting note about the tank breather valve contributing an erratic idle. I have this idling issue on warm starts; not cold. It hunts a bit. And it goes away after I simply rev the engine a bit or drive. I believe the purge valve only opens on idle when vacuum is highest and closes when the revs go up and vacuum drops. I think it is also closed initially on cold starts, at least that is what my M5 used to do. Perhaps this little rev push actually unsticks the valve to a fully open position instead of half open or something like that.

I have replaced the VANOS actuators without result. I might throw in a new purge valve for giggles.

[vilord]

The tank vent valve opens when the engine is shut off. That is how it works.
The *purge* valve closes.

There are two valves in there, and I was calling the tank vent valve the wrong name.

Perhaps there is a slight delay so it can measure pressure drop, but if it didn't open then the air going into the charcoal canister wouldn't be able to get out to the atmosphere.

Here's a generic page about EECS/EVAP:
https://www.samarins.com/glossary/vent-valve.html
"A clogged up vent valve can cause trouble filling the gas tank." <-- this would also happen if it was stuck closed, as mine was.

The BMW system isn't any different.

[vilord]

Originally Posted by rt7085

I might throw in a new purge valve for giggles.

If you're going through the trouble, I suggest pulling out both the vent valve and purge valve and testing them with a vacuum gauge as shown in the article I linked. If one of them is leaking a bit when closed, you would most likely encounter some of the idle hunting you're describing.
Could also be a leak or crack in one of the vacuum hoses, or in the tank vent piping, or in the gas cap, or... yeah vacuum leaks are a pain.

This makes me wonder about the E46 coupe my ex had years ago where the engine would die when returning to idle... the dealer never could figure that one out...

[rt7085]

Originally Posted by vilord

The tank vent valve opens when the engine is shut off. That is how it works.
The *purge* valve closes.

There are two valves in there, and I was calling the tank vent valve the wrong name.

Perhaps there is a slight delay so it can measure pressure drop, but if it didn't open then the air going into the charcoal canister wouldn't be able to get out to the atmosphere.

Here's a generic page about EECS/EVAP:
https://www.samarins.com/glossary/vent-valve.html
"A clogged up vent valve can cause trouble filling the gas tank." <-- this would also happen if it was stuck closed, as mine was.

The BMW system isn't any different.

The only valve is the tank breather valve. There is no purge valve. There is only a dust filter and a DTML pump which is a pass-through when it is off during engine running. (There is no check valve either). The dust filter allows air to backfill the charcoal canister with fresh air when the engine draws the hydrocarbons out of it with its vacuum. The filter allows this without clogging the system / engine with atmospheric dust. If that was closed, the whole fuel tank would suck vacuum and collapse. (This was actually an issue on early M5 E60 where tanks literally crumbled up like balloons as charcoal filter clogged so badly and the vacuum of a V10 was so high). There was report earlier in the thread where somebody could not get the fuel tank cap off. Guess what, it is likely due ti vacuum. That person should throw his eyes on the tank as his dust filter is likely clogged.

[vilord]

Okay, guess I'm wrong.

All I know is that when I replaced my faulty tank vent valve (as tested with ISTA, and triggering a CEL) with a correctly working one, all of my problems with the pump shutting off early stopped. *shrug*

[N54Society]

That's a safety feature to make sure you're not on your phone while pumping.

[Stealthflipboy]

Thanks everyone!! Been so frustrated with this for the better part of 3-5 years been so irritated parked the car for months at a time. Banging the charcoal filter has been such a breath of fresh air, who woulda known?

[QuestFor10s]

My recent experience has given me some insights that I thought I would share to future people in my situation.

I had recently added a Radium dual pump surge tank to my E92 335IS and at my first fill up the gas pump clicked off at about half tank. It acted like it was full but was clearly not. Unfortunately I tried all the "tricks" to get it to take more but to no avail, unfortunately those tricks contributed to flooded my carbon canister and DMTL pump in the process. It made no sense until I spent the next several days learning the BMW EVAP system, like a master tech, and solved my problem.

Cutting right to the chase:
During installation the internal tank breather vent hose came off its fitting and allowed gas to enter it. As we know, pumping gas into the tank displaces the air, which normally leaves via this breather hose and goes to the carbon canister. But this breather hose gets submerged at half tank and, for me, began displacing fuel into the carbon canister, flooding it, and blocking any airflow. This caused the pump handle to click off, even at half tank as in my situation.

RULE: A flooded carbon canister will cause the gas station pump to click off.

So disconnecting the carbon canister dumped about 1 liter of fuel out and the DMTL pump also became filled with gas. It also allowed gas to flow into the vapor purge line which runs up to the engine. So even if you replace the CC and DMTL pump you will still have a problem! The symptom is the engine will stumble every time the EVAP purge valve solenoid is energized, which is every 30 seconds or so under manifold vacuum. It stumbles because the fuel mixture is suddenly very rich.

So the solution is you must blow out all your evap vent lines prior to reinstalling a new CC and DMTL pump. Now the car runs perfect and the gas station pump works as you would expect, clicking off when it detects its full. Like everyone that said its not the pump its the car, they are correct.

For those who top off their tanks, replaced their CC and still have problems, I believe you may still have gas *in your lines*. Depending on how much or little could make the difference between more sensitive gas station pumps working or not.

A breather hose with fuel in it could cause the CC to get flooded, since it could act like a siphon because the CC is lower than the fuel tank.

For future reference the line which is important to clear is the large one on the passenger side fuel pump hat. The other end which must be disconnected is the lowest hose on the carbon canister.

Good luck! Let me know if I can help anyone facing a similar problem.

[S65Hunter]

Originally Posted by QuestFor10s

My recent experience has given me some insights that I thought I would share to future people in my situation.

I'm having a very similar issue. Found leak at the FPR, replaced it. Couldn't get more than half a tank no matter what gas station/pump I used. Tried to force it to pump, that led to charcoal canister and DTML pump to flood. Removed them both, drained all the gas and let sit over night. The bogging down at cruising speed seems to have gone away. I'm wondering if during FPR install I knocked one of the vent valves loose or something of that nature.