Stubborn Purge Valve codes P0441 / INPA 250

[80stech]

It almost looks like that leak detection spool has 3 positions with center being off. It's not really clear though the way the diagram is drawn on how the coils are wired.

This would have been good info on that thread about the gas tank getting crushed.

[e39_touring]

Originally Posted by 80stech

maybe the leak detection valve is not spool is not fully going into the run position(against spring).

I'm going to rule this one out because the way the DMTL works is it monitors motor current while pumping through the reference orifice and then fully closed, comparing the result. In INPA, you can see the motor current at each stage of the test (reference, 1mm leak test, .5mm leak test). In my case, the motor currents all make sense and are within acceptable range - immediately high at closed, lower at reference measurement, and gradually rising to match the closed rate during the 1mm and .5mm test cycles (gradually rising as the fuel tank and associated EVAP lines pressurize).

I know the path from the purge valve through the canister and out the DMTL pump filter is not blocked as I was able to blow lightly compressed air with smoke through it. What I don't know is if it's partially restricted. I think the culprit has to be either the canister or DMTL filter. I was hoping to avoid pulling the fender liner because I have to order more of those plastic rivets to reinstall the trim, but it may come to that.

Haven't had a chance to try and measure vacuum/flow between my two X5's to compare, and I think that will shed some light.

[80stech]

This is a very interesting thread! As far as I can tell from your diagram on how the spool is laid out, just because the DTML can go through all it's tests doesn't prove the spool is all the way back in the non-test position when it's done.

If that diagram is right then I wonder where the 1mm orifice is ??

[e39_touring]

Originally Posted by 80stech

This is a very interesting thread! As far as I can tell from your diagram on how the spool is laid out, just because the DTML can go through all it's tests doesn't prove the spool is all the way back in the non-test position when it's done.

If that diagram is right then I wonder where the 1mm orifice is ??

Even though the diagram doesn't specifically state it, I'm guessing the 1mm orifice is at the pump outlet itself.

That valve/plunger (is that what you're referring to when you say 'spool', or do you mean the coil/solenoid?) looks to me to only have two positions. The diagram shows the default (de-energized) view. Air can enter the filter and takes the right (up) fork at the y and goes on up to the canister. This is the position during engine operation.

The EVAP leak test is only done after the engine is shut off (and other conditions are met). It starts with the change-over valve in the same position, but the pump is activated to take the reference reading (pump through the .5mm orifice and store the current reading for future comparison).

Then the plunger is energized, and it moves to right in the diagram. This blocks off the port where the right (top) fork in the line from the filter meets the valve and opens the port at the valve where it meets the line heading straight up from the pump.

By monitoring motor current over time (the test takes a few minutes), the system can detect 3 conditions:

1) Leak > or = 1mm as the motor current falls and then doesn't rise.
2) Leak > or = .5mm but < 1mm as motor current falls then rises as tank pressure builds but eventually levels off at a level < or = the current stored from the 0.5mm reference reading.
3) No leak (or, more accurately, no leak or a leak less than the allowable 0.5mm standard) as the motor current falls then rises as tank pressure builds and ultimately surpasses the current of the 0.5mm stored reference measurement.

OK, whose head hurts after reading this?

[80stech]

I'd put my money on the problem being with that change-over valve. You could almost eliminate the filter because the test pump uses it as well.

[e39_touring]

Originally Posted by 80stech

I'd put my money on the problem being with that change-over valve. You could almost eliminate the filter because the test pump uses it as well.

I don't think it's that because there's a separate code for the DTML pump failure. That, and the test is run with the engine off, and I can clearly hear that valve click when it switches from reference to testing.

Keep in mind that to run the test, the pump draws very little air (probably only a little more than is needed to build pressure in front of a 0.5 or 1mm orifice). However, when the tank purge valve opens, I think it draws quite a rush at engine vacuum (judging by the size of the lines and ports on the tank purge valve). So, a clogged filter could conceivably work fine for the purposes of tank testing but be too restrictive to move the air required or expected to purge the canister.

I know this feels like bench racing, and, hopefully, I'll be able to run some more tests tomorrow!

[80stech]

The pump would need to deliver a fair volume of air to pressurize the gas tank. Either way if you eliminated the purge valve and line I think you're going to need some plastic rivets!

[andrewwynn]

You can drive without the wheel arch trim for a spell if needed. I didn't have the rivets when I changed my rear brake lines and for about a week I had no fender teim. The actual wheel well is screwed on so that wasn't a problem

[e39_touring]

Well, I had a few rivets left from an earlier trim swap so I dug in a little bit. It was very clean behind the fender liner so I didn't dismantle the DMTL pump.

Unfortunately, I also couldn't hook up a vacuum gauge because the only tees I had in my toolbox were for a much smaller hose diameter. which would have caused a restriction in and of itself.

What I did do is hook up my mityvac hand vacuum pump to the canister-to-purge valve tube, and I was not able to draw any vacuum with it. And, just for good measure, I blew more air (~20 psi) back through that tube to try and knock anything that might have built up out of the way. I didn't sense any restriction and could hear the air flowing pretty well.

Finally, I buttoned everything up and drove the car for a few days and for about 50 total miles mixed 50/50 between city and highway. Knock wood, no CEL or pending codes, and all emissions readiness monitors are OK. I'm hesitant to say completely that the problem is solved since I never found anything I could point to as the culprit, but I'll take what I can get.

Thanks, everyone, for the ideas and suggestions!