Interesting. It took me no more than 1.5hrs to do the job and i did not need any jelly. I guess i am more special. :haha:

 

Sorry, I'm not that expert but what does it do?
and if it is broken, what are the symptoms?

thanks in advance.

 

Maybe rough iddle but otherwise just the cel.

Mike

 

OK, so I broke the old Evap Purge Valve apart...

As you can see, the only real "moving" part is the little black plunger on the bottom. Even so, it only moves about 1mm under very light pressure (obviously air pressure, considering its on the bottom, where the two hoses connect). Having this valve go in/out may only "activate" the upper portion of the valve.

Can one of our electrial experts interpret what's going on with this thing? I don't see how this part can "wear out" since the only moving part doesn't get clogged with anything except air.

 

I guess since I am a solenoid valve design guru (Not my terminology!. I actually was given this on a business card at a previous employer! <lol>), and used to design Canister Purge Valves (in Automotive industry jargon) aka Evap. Purge Valves, for a large Tier One, maybe I can explain. I've spent 25+ years designing mostly fuel injectors, fuel pumps, other fuel system valves, and a couple of oddball A/C magnetic clutches.

As someone else posted, these valves allow air to escape from the tank when you fill, and allow outside air to re-enter the tank as you drain it of fuel, The connection line goes from the tank, to this valve, and then on to the charcoal canister, and then through the carbon to the outside air. That ensures that air can move in and out of the tank, but no hydrocarbons escape, as they are trapped by the carbon.

The valve is a normally-open device. When it isn't energized, the valve sits in the open position, allowing air/vapor to flow into and out of the tank. It is generally only energized (i.e. closed), after key off, in order for the ECU to perform a shutdown leak test, which is part of OBDII diagnostic testing done by the car. Assuming the ambient temps aren't too cold, then the ECU in your car, at shutoff, closes this valve. Due to normal boil off of fuel in the tank, the pressure will build up slightly. There is a sensor that can measure that pressure rise inside the system, and the ECU expects a certain amount of pressure rise depending on the outside ambient temp, and a few other factors. the higher the temps, the bigger pressure rise it expects. If it measures the correct pressure rise, at shutdown, then it knows that the leak integrity of the fuel system is within government mandated standards, and you are not a raw hydrocarbon gross polluter!

These valves must be pressure tight when closed. They tend to fail due to leakage across the valve seat when the valve is closed. If they leak, the boil off gases bleed off to the canister during the test, no pressure rise occurs, and the ECU assumes there is a leak.

SO.... Even though they look very simple, they are quite prone to failure due to contamination across the seat. I have gotten the E8 code on my 2002 M5, and I simply pulled the CPS valve, took it into work, flushed it with stoddard solvent, and re-installed it. No more contamination. No more leak. No more code!

These valves have NO impact on the performance of the car, UNLESS they happen to stick closed. They are designed to fail OPEN. I've never seen one stick closed on any that I've designed, and put into production (GM, Ford, and VW products, plus some others), but it could theoretically happen. However, if it did, you would also have a hard time filling the tank, as the resulting back-pressure would cause the fuel fill nozzle to kick off as soon as you started flowing a few tenths of a gallon of gas.

 

Unlike some others, I did not replace the carbon canister at the same time. According to a BMW Tech that I call frequently for advice, replacing the valve is all that is necessary (since its the valve that breaks down internally).

My car is booked in on 13th March to have the valve changed. (had Peake E8 for about 4 years!)
I'm trying to decide whether to do the carbon canister at the same time as my indie said it may be that that's causing the code - opposite of what was said above! Is there any more advice out there? Have most people had both changed at once or is this tech right? Decisions decisions!

 

The purge valve basically allows air to escape from the gas tank as the fuel/air volume inside the tank changes. I haven't had a problem with it yet on my M but changed it twice on my old Nissan. I've never heard of the canister being the cause of the problem for any vehicle. It's supposed to last the lifetime of the car and has no moving parts. The only time I could see a tech wanting to change the canister as well is because sometimes they will strip or break the bolts that connect the valve to it and just want to replace the whole thing as a unit so they can just disconnect/reconnect a few hoses and one electrical connector. Never looked into the location of it on our cars yet but on most vehicles I've found that the canister with valve are usually located somewhere near the fuel tank under the rear of the vehicle, most often tucked into the rocker panel behind a rear wheel. This area is very exposed to temperature, road conditions, salt, etc. and the valve is just a weak little solenoid so it doesn't take much to kill it.

 

So I had the Purge Valve done last Friday. I had had the code for years as I mentioned, but also got occasional fuelling codes too. As I'd recently changed all the sensors, and checked for exhaust leaks etc, I was beginning to suspect a vacuum leak. While the valve was being replaced, we checked over all vacuum hoses etc, and they were all fine.
M5 London started a crazy thread about the car feeling better with a full tank and loads of people agreed, despite the fact that there was no logical explanation. I got to thinking that maybe somehow the vent valve behaved differently with a full tank and sure enough I only got fuelling codes and E8 when it was below half a tank.
The car now feels snappier and more responsive. It's smoother in traffic and changing from 1st to 2nd as well. Also the idle is smoother, sitting almost dead on 620RPM instead of fluctuating by 50/60RPM like before.
I've had no fuelling codes since then either so I'm thinking that a faulty Purge Valve is not just an SES light problem but can actually affect the mixture. Maybe it can leak air if it's failed in a certain way?

 

That's funny that this popped up. I just got this code this weekend on a 1,300 mile trip up north to see the Grand Canyon and Monument Valley. Along with this code I received "filler cap open [wasn't] and a couple codes on the exhaust CPS bank #5-8. I'll likely replace the CPS and see what happens, but at least here's some very timely info for me.

Thanks!

 

WORD!hmmm

I got b2, b3 and E8 after I kind of hit the top of a speed bump (coming down and off the speed bump). B2 and b3 have since reset and have not reappeared, but E8 has showed its amber face 3 times in the past 4 or 5 days.

Group Buy? HAHA! Jk

Good thread. tstaples and others, thanks for your insight!

 

Great info Tstaples. I've got the initial inconsistent idle during cold starts, no codes though. Others have mentioned this is caused by a bad purge valve, and needs to be replaced.

Do you know how the bad valve causes the rough idle? I assume something to do with the ECU getting a inconsistent reading from the system?

Thanks
Biz

 

I really don't think this valve can cause rough idle. It either works, or it doesn't, and if it leaks, then the ECU throws a code inidicating that there is a leak detected by the CPS. Of course, the fallacy of all this is that the leak is usually IN THE DARN VALVE that used to run the leak test, and not in the actual system itself. Hmmm, I wonder if any of my fellow valve designers have been designing in planned obsolesence in order to generate more replacement valve sales? hmmm

 

YES this is a Pain in the *&% too do I did it two months ago. The short tube(very short) and the connector took me 30 min and one broken side to get it off

 

Cheers tstaples. I'm not questioning your knowledge for a moment here, but maybe I have misunderstood the process. Your description only mentions allowing air/vapour in and out of the tank via the cannister. If this was the case, wouldn't the valve be a lot closer to the tank itself, rather than miles away in the engine bay? I understood that the ECU uses it to allow vapour into the intake manifold to be burned at idle - hence a fault can lead to a vacuum leak, hence the idle issues. I will happily stand corrected.

 

It's quite possible that I am wrong as to the strategy. I am "assuming" (and we know what kind of trouble that can cause) that the evap purge valve functions as I am accustomed. The valves I have designed normally reside atop or near the fuel tank. It's possible that BMW employs a different control strategy than I am familiar with. I have not studied the actual routing/control strategy used on the M5. I assumed that my valve was located in the engine bay area due to packaging constraints atop the fuel tank, due to the rear seat. Interesting thought.

 

Pretty sophisticated

Below is a description of the Evap control leak detection system used in the X5, which I think is very similar to the M5. It differs a bit from the previous poster (tstaples), in the order of where that valve is located (after the carbon canister, between the manifold and the canister), and that there is s tricky pressurization of the whole system done by a pump. Leaks are measured by comparing the load current on the pump motor relative to that of a reference orfice. I quoted the BMW explanation below:

The DC Motor LDP ensures accurate fuel system leak detection for leaks as small as
0.5mm (.020”). The pump contains an integral DC motor which is activated directly by the
engine control module. The ECM monitors the pump motor operating current as the measurement
for detecting leaks.
The pump also contains an ECM controlled change over valve [the vent valve] that is energized closed during
a Leak Diagnosis test. The change over valve is open during all other periods of operation
allowing the fuel system to “breath” through the inlet filter (similar to the full down​

stroke of the current vacuum operated LDP).

The DME only initiates a leak diagnosis test every second time the criteria are met. The criteria
is as follows:
• Engine OFF with ignition switched OFF.
• Engine Control Module still in active state or what is known as “follow up mode” (Main
Relay energized, control module and DME components online for extended period after
key off).
• Prior to Engine/Ignition switch OFF condition, vehicle must have been driven for a minimum
of 20 minutes.
• Prior to minimum 20 minute drive, the vehicle must have been OFF for a minimum of 5
hours.
• Fuel Tank Capacity must be between 15 and 85% (safe approximation between 1/4 -
3/4 of a tank).
• Ambient Air Temperature between -7OC & 35OC (20OF & 95OF )
• Altitude < 2500m (8,202 feet).
• Battery Voltage between 11.5 and 14.5 Volts
When these criteria are satisfied every second time, the ECM will start the Fuel System Leak
Diagnosis Test. The test will typically be carried out once a day i.e. once after driving to
work in the morning, when driving home in the evening, the criteria are once again met but​

the test is not initiated. The following morning, the test will run again.

LEAK DIAGNOSIS TEST
PHASE 1 - REFERENCE MEASUREMENT
The ECM activates the pump motor. The pump pulls air from the filtered air inlet and passes
it through a precise 0.5mm reference orifice in the pump assembly.
The ECM simultaneously monitors the pump motor current flow . The motor current raises
quickly and levels off (stabilizes) due to the orifice restriction. The ECM stores the stabilized
amperage value in memory. The stored amperage value is the electrical equivalent of a 0.5​

mm (0.020”) leak.

PHASE 2 - LEAK DETECTION
The ECM energizes the Change Over Valve allowing the pressurized air to enter the fuel system
through the Charcoal Canister, The ECM monitors the current flow and compares it​

with the stored reference measurement over a duration of time.

Once the test is concluded, the ECM stops the pump motor and immediately de-energizes
the change over valve. This allows the stored pressure to vent thorough the charcoal canister​

trapping hydrocarbon vapor and venting air to atmosphere through the filter.

TEST RESULTS
The time duration varies between 45 & 270 seconds depending on the resulting leak diagnosis​

test results (developed tank pressure “amperage” / within a specific time period).

So there is quite a bit more to it if you have an EVAP fault besides just the vent valve. Any interconncet tubing that is leaking more than the .02" reference orfice will produce a fault, even if the vent valve is tight.

 

That's great info, DHL93449. As I mentioned before, I was not familiar with the BMW approach to running this test. This is very different than the methodology used by several other OEM's for whom I have designed valves. I don't know why such a complex method would be employed, when the standard can be met with a much simpler approach. A case of Uber-Tech run amok, I suppose. (I was educated in the K.I.S.S. school of design.) Unfortunately, when I look at this from a systems engineering approach, there is a LOT more that can go wrong, and trigger a code, than just a leaking valve.

BTW, I take back my earlier statement that I don't think this valve can cause unstable idle. The way BMW have implemented this OBDII test, this valve can certainly cause idle problems. (Again, why would you do it this way?????)

However, back to my personal M5 situation. As I wrote previously, my car was throwing the E8 code awhile back. I pulled the valve, flushed it, and all was well; until recently. The dreaded E8 code returned. Interestingly, it fired off while the car WAS SITTING ON THE CHASSIS DYNO the morning I was getting the car smog checked for it's bi-annual renewal. The car failed due to the SES light, but the guy who does my smog tests ran it anyway. The car passed all categories except HC. It was ONE PPM over the max allowable limit. So, as the clock was ticking, and I couldn't afford to waste the last couple of weeks before my registration would expire, I popped into the local BMW dealer, ordered a new valve, got in about 3 days later, and installed it. (Geez that's a lot of nuts to take off for one lousy valve!) I then reset my codes, and drove the car for 4 days to ensure everything was reset, and the timers cleared out (Otherwise the car will still fail smog test. They can tell if you have just reset the codes right before rolling up for the test. It's something like 10 starts, with the car running more than a certain number of minutes each start, spread over a minimum of 3 or 4 days, to clear the "reset". My numbers are likely a little off. If anyone knows definitively, please chime in.)

Back to the smog tester for the re-test, and Voila! It passed. But wait! There is more to the story. The car again looked great in all categories, except HC. This time, the car passed, but was ONE PPM below the max allowable limit for HC. Alright!. I spent $70 for a new valve, and several more hours of my labor (if I include the time necessary to make the two frozen Strawberry Margaritas necessary to complete the valve swap.) and all I got was a lousy 2 PPM improvement in f___ing HC? :dunno: I highly doubt if the resolution of the sniffer used by the smog test station is that accurate, so my two tests are most likely just numbers based on bouncing around within the noise envelope of the tester.

The good news is that the car passed. I suspect that there is still something amiss somewhere, with just enough leak to trip the light under certain conditions. I haven't gotten another one since, but the weather has also been cooler, and that should/will have some impact on the test. I guess I've got another 2 years to find it and fix it!

Thanks again for the info. Based on the information posted (and again assuming the M5 is similar to the X5), I just figured out a great way to "fool" the test via tweaking the pump output current. Now, I need to go study the diagrams, and figure out where that pump is!

 

I just threw an 0441 OBDII code. Would someone point me in the right direction of the purge valve location? I'm gonna try the old clean it to make it work better trick first and then buy a new one if necessary. Thanks anyone/everyone,
Jim K

 

It's located in the valley of the motor, under all that beautiful air induction hardware, towards the rear. It's a tedious, and time consuming, but relatively easy job. Check out this thread for pics:

http://www.m5board.com/vbulletin/e3...52-z8-discussion/94891-replaced-my-fuel-tank-breather-valve-pictures-job-2.html

Check out post #17 in this thread for a handy guide.

http://www.m5board.com/vbulletin/e3...lletin/e39-m5-e52-z8-discussion/92809-replacing-vapor-ventilation-valve-10.html

It's part 14 in the diagram:

http://www.realoem.com/bmw/showparts...21&hg=13&fg=15

 

My test findings

As part of the last project, which involved stripping the top of engine down to bare heads, one was replacement of the
evap purge valve due to fault code. This solenoid valve is normally closed and energized open. It does nothing more
than the name implies, it purges the carbon canister of absorbed hydrocarbons, plus providing a "system closure valve"
for leak detection test. Evaporative control on any car never vents to ambient, that is the sole purpose of the any evap
system; storage, then re-introducing them back into the engines intake system. It has been that way since enactment
of the federal "Clean Air Act of 1968".
On our M5, the hose connection on the side of the valve connects to the vapor storage system and the lower hose goes
to the 90 deg elbow that connects into the rubber "tee" just aft of the idle air control. With the solenoid de-energized,
valve should be closed. In my case, there was a constant "open" between the two hose connections, consequently the
leakage test for the system failed, initiating the CEL. The new valve was fully closed when I checked it prior to it's
installation. In actuality, the ECM does not "energize" it closed, it simply cuts-off the DC voltage, allowing the valve to
return to it's "normally closed" position.
NOTE: With only two terminals on the valve it can only be one of two alternatives, energized open or energized closed.
Based on findings above, it is energized to the open position.

Regards,
Alan

 

I just got this code yesterday

Is this thing covered under the emissions warranty or do I just need to DIY?

 

Mods alert

There is some great info in this old thread.... suggest adding it to the fuel tank breather valve subheading in the DIY sticky.