DC VII Something more mechanical than electronic

[thcardoc]

Evaporative Systems

One of the diagnostics and repair routines that we do quite often these days has to do with the evaporative emissions system on the car. Codes like P0440, P0442, and P0456 (when used) are usually all system leaks or system performance tests based on feedback signals that suggest a leak or component failure and they require a specific set of tools and technician experience while component codes such as P0443, P0449, P0452 and P0435 require a totally different approach. There are many more codes for all of the different systems across the vehicle fleet depending on make and model and combine that with the fact that a lot of the components can be difficult to reach evaporative system issues can be challenging. But like every other thing we have to do today you can cut the task down to a manageable size by having a good game plan and following it through.

Every occurrence of one of these codes can be dealt with through the use of a scan tool, voltmeter and/or test light, and many shops also use a smoke machine to try and find leaks in the hoses or components. Whatever method you choose the first thing to do is NOT TOUCH THE GAS CAP. If by chance it is or was loose and you touch it first you just succeeded in making it impossible to confirm that was really the cause of the check engine light. A good number of the failures are caused by components that are just starting to fail and are presently working intermittently, touching the gas cap first causes it to also be an unknown because if it was leaking, that could be proven by starting to test the system and observing an active leak, then tighten the cap and see if you made a change in the system leak rate.

For this exercise we will use a 2003 Chevrolet Malibu. It is setting a P0440. The outline for the routine is to first get the code’s description and it’s “code set criteria” which is the conditions and sensor values that have the computer make the decision that there is a problem. So let’s get started, how do you want to approach this code?

(Erasing it and hoping it doesn’t come back on doesn’t count as a first step ) VBG….

[thcardoc]

The first thing you would need to do is look up the description of the code.

P0440, general evaporative emissions failure, weak vacuum test, or gross leak. Three common names or descriptions for that one code. The code sets when PCM attempts to pull the cannister and the fuel tank into a vacuum that exceeds 8" in water. Most vacuum gages we associate with automotive use measure in inches of mercury. Eight inches of water is less than 1/2 of one inch of mercury vacuum. You can associate this to how much effort it takes to take a drink of soda with a straw, where the top of the soda in the glass is about eight inches below the top of the straw. If the PCM can pull the system down that far then first of all that confirms that the purge valve works and allows a vacuum to be pulled, the vent valve must have closed and there are no major leaks. Not being able to pull that much vacuum of course means that the purge valve either didn't work for one reason or another, or there is in fact a large leak.

Now that you know what could be the problem, how do you start to narrow down what is and isn't working correctly this time??

[thcardoc]

The PCM uses the fuel tank pressure sensor to determine the tank pressure and it controls the purge and vent valves to carry out a testing routine. Under normal operation the cannister purge valve is duty cycled to control the flow of air and fuel vapors into the engine to purge them from the cannister and tank. The vent solenoid is normally open and get's closed to allow the computer to test the system. Armed with a few jumper wires and a volt meter you can if you choose to try and watch the computer run it's natural test which under the right conditions it should accomplish about once a day, or you can operate the solenoids yourself while you monitor the fuel tank pressure with a volt meter. For the record this is essentially how it's done in the shop but the scan tool with bi-directional controls and the ability to monitor the fuel tank pressure at the same time is a big time and effort saver, this exercise is about how you would do this without having a scan tool.

The schematic shows that the cannister purge valve control wire is dark green / white, (dk-gn/wht) and goes to pin 76 connector #1 of the computer, the vent control valve control wire is white (wht) and goes to pin 2 of connector #2, the fuel tank pressure sensor is dark green (dk-gn) and goes to pin 55 of connector #2.

Choose the easiest location to make your connections and back probe the connectors, don't pierce wires unless you absolutely have to. If you do pierce a wire, it must be sealed again and tape alone doesn't do the job. Nail polish, works well as do certain adhesives that dry quickly, then tape.

Connect the voltmeter to a good ground and to the fuel tank pressure sensor, dk-gn wire. The voltage will be about 1.5v with atmospheric pressure, and will go down with an increase in pressure and rise with a decrease (vacuum pulled). With the engine running and the purge valve commanded open by connecting a jumper wire to ground and the purge valves dk-gn/wht should cause a slight increase in the voltage from the pressure sensor indicating that a slight vacuum being pulled in the tank. It's common to notice a momentary change in the engine's idle too as the computer has to adjust for the extra air and possibly fuel vapors. (one note, the four cylinder eco tech may use ported vacuum so the idle speed may need held around 2000 rpm to have vacuum for the purge system.

Grounding the vent solenoid wht wire will cause the vent solenoid to close and stop allowing fresh air into the cannister, and all the way to the engine. This will cause the tankspressure to drop quickly so be ready to remove the jumper from the purge valve as soon as you get to 3 to 3.5v on the pressure sensor. If you pull too much vacuum on the system you can collapse the tank and potentially damage it. A scan tool usig bi-directional prevents that because the computer will shut down the test if the vacuum in the tank exceeds a specific level. You have to do the same that the computer does in being diligent to not exceed a safe vacuum level.

With the vent valve closed, and the purge valve open the tank and cannister pull into a vacuum. If you would unplug the purge valve at this time, the tank and cannister, and their lines should retain that vacuum. (Not bleed it off).

With your testing you see some vacuum start to build even with the purge valve open and the vent closed but not a lot. Your pressure sensor voltage doesn't go above 2v. When you unplug the purge valve to try and seal the system the pressure quickly returns to 1.5v atmospheric pressure. This confirms the code P0440, you cannot pull a strong vacuum on the tank and you clearly have a leak beacuse it cannot hold any vacuum that is being created. Your next step is???

[3boystoys]

Looks like your playing with yourself on this one.

[thcardoc]

Originally Posted by 3boystoys

Looks like your playing with yourself on this one.

This work is such that "experts" are everywhere but when it comes down to it few are willing to put their ideas out there because they fear being shot down and exposing themselves as not having all of the training and experience that they claim to have. In trying to say that another way, there is no room for failure when working in a shop and these exercises which do draw a number of readers are in fact sufficienty complicated that responders are much more likely to be incorrect than correct so they hold back from trying to answer and simply read and wait for the solutions. Either way as long as I write these they get a little more knowledge about doing this kind of work and gain some insight as to what it really takes to fix these kinds of problems. Now sure it would be great if some did throw out their ideas whether they are good or bad, but I'm not going to hold my breath waiting for that to happen.

There is a reason that there are no shows presenting material at the level that threads like this one will, the DIY market simply wouldn't respond to it and that's what pays the bills. The problem is your not going to attract and keep the pro's around when someone tries to pretend that even this little evaporative system problem is simple to solve. Lot's of shops and their techs get they lunches handed to them on cars just like this one. (This is based on a real car that beat three shops, and two DIY'ers)

Not having a sound approach to this will allow it to beat the next person up to bat as well, IMO.

BTW, did you have something constructive to contribute? Any idea what should be done next?

[3boystoys]

Guess you don't have enough work to do.

[thcardoc]

Tons of work, this is something I'm doing for fun when I need a break.

[3boystoys]

Originally Posted by thcardoc

Tons of work, this is something I'm doing for fun when I need a break.

Maybe your need to re-evaluate.

[Daniel Wood]

Hey boy, how come you don't figure the solution out? Doc is pretty sharp with this stuff, give it a chance.

[thcardoc]

Originally Posted by Daniel Wood

Hey boy, how come you don't figure the solution out? Doc is pretty sharp with this stuff, give it a chance.

Thanks Dan. When I post these on professional websites the techs put in the effort to figure them out. Doing stuff like this without the pressure to get it right the first time allows them to develop better thinking skills. Then when they have to work through one of these cars for real they are much more efficient at it. The same does hold true for the DIY crowd.

FWIW, In the last month I've done four similar repairs as to the one referenced in this thread, they all had different causes, and all were different from this one. All five of these essentially reached the point in the diagnostics that we are at right now with the exception that I use all of the capablity of the scan tool instead of doing everything by hand in the shop. But that doesn't mean that it can't be done by hand as this thread is intended to demonstrate. I'll give it a few days now unless someone takes a stab. That will give them some time to consider what needs to be done next in order to identify exactly what the cause of the failure is "this time". The best part is this will set the stage to also identify any of the other solutions without any more effort.

[thcardoc]

At this point you are able to play the role of the computer and actively test the cannister and fuel tank and the rest of the system for leaks. By leaving the vent valve grounded it stays closed, and pulsing the purge valve to ground you can pull the system into a vacuum, which you can watch with the voltmeter. The voltage rises when you ground both valves, but quickly falls as soon as you stop grounding the purge valve.

I asked what the next step was, and the answer is to start isolating portions of the system. A very common leak area is the fuel filler neck, and of course the gas cap. This is the first moment that touching the cap is a valid step, if you first prove that the system has an active leak, then you tighten the cap and do a retest you have the ability to prove if it was really a problem or not. If once you tighten the cap the fuel tank pressure sensor voltage rises quickly, and then holds when you stop pulsing the purge valve then the cap was in fact leaking. For this car, tightening the cap made no change in the rate of vacuum decay (how fast the pressure transducer voltage dropped).

Now take a clamping tool and try pinching off the hose between the fuel filler neck and the tank. If you made a difference, then the leak is in the filler neck assembly, if that made no change then test further. The vent valve itself may work mechanically as in you hear it close but it could still not seal completely, by clamping that hose you could prove/disprove that portion of the system.

With this particular car the leak remained with every check so far and since it is a large leak you cannot even be sure if there is also a leak in the filler neck area or the vent valve and so they will need to be rechecked once you locate and correct the major component failure.

Now what is the next step?

BTW, in a way 3boys was right about one thing. Reading these and benefitting from what is being shared is fine but the lack of interaction between members has to be seen as reinforcing why you rarely see diagnostics at this level on any of the TV shows. No interaction "looks like" no-one really cares. Lot's of responses on the other hand would serve to support these kinds of articles, or stories both here and on air. JMHO

[Daniel Wood]

Well don't give up on us, you never can tell when someone will come along and solve your problems.

[carsandcycles]

We are yielding to your expertise; this is not easy stuff, and if like me a lot of members are more of the 'shadetree' mechanic, this stuff is over our head(s) (at least, it is to me). Sorry for the interruption.

[thcardoc]

Originally Posted by carsandcycles

this is not easy stuff, and if like me a lot of members are more of the 'shadetree' mechanic, this stuff is over our head(s) (at least, it is to me).

For people who make their living's repairing cars many of these things we have to do today aren't easy at all, and yet almost everyday we see or hear someone try and tell consumers that what we do is simple, and that the computers tell us what is wrong. Do a google search for the P0440 and you'll find tons of advice about "it's just the gas cap" and only occasionally you'll see other final answers such as a leaking component but you'll almost never see how you prove what component is leaking without simply throwing a grocery cart of parts at the car.

In the last month I had one car have a leaking filler neck, two cars with bad vent valves, one car had a leaking cannister, and then there was the one this case study is based on. Up to this point the system has been tested basically intact with only the wiring connections made. Clamping off the filler neck, and vent valve hoses showed a leak elsewhere in the system. Shops use a smoke machine to help with the diagnostics at this point, but you don't have one of those. Is there anything you can do?

Start testing individual sections. You'll need a vacuum/pressure gage preferably one that measures in inches of water rather than inches of mercury, but either will do. A hand vacuum pump will help but remember the fuel tank is a rather large chamber to try and pull into a vacuum and fuel evaporates easily so testing with the tank connected to any other component with just a hand pump will be difficult. But that doesn't mean its impossible.

You need to first open the system at a point that will allow you to test for leaks on each side of where you opened the system. The purge line from the engine is easy to remove from the cannster and it is a 1/4" nipple which means you can find some vacuum hose that will fit it nicely and then with a vacuum "T" you can attach some type of a vacuum or pressure source as well as a gage to watch the system pressure if necessary. Keep in mind once you start isolating sections of the system you might be testing components that are no longer connected to the pressure sensor in the tank so it won't be able to help you them.

Using a vacuum brake bleeder tool you can generate a pretty good vacuum flow. At this point we are testing the tank, the cannister, the vent valve, and the filler neck as a group. Pulling a vacuum on them it climbs quickly, and clamping the hose from the vacuum bleeder we see the vacuum holding.

We now know that the leak is between the purge valve at the engine and the connector that attached to the cannister. Using a piece of 1/4" brake line as the nipple testing that line by itself confirms that it is leaking. These steel and nylon lines are repairable (it's three pieces from the factory). Corrosion is viisible in the rear of the car alongside the fuel tank but is that where the leak is? The connector at the front of the steel line is almost impossible to reach since it is between the steering rack and firewall and no matter how hard you try it simply cannot be removed again due to corrosion.

We're almost there, what would you do now?

[thcardoc]

Originally Posted by thcardoc

We now know that the leak is between the purge valve at the engine and the connector that attached to the cannister. Using a piece of 1/4" brake line as the nipple testing that line by itself confirms that it is leaking. These steel and nylon lines are repairable (it's three pieces from the factory). Corrosion is visible in the rear of the car alongside the fuel tank but is that where the leak is? The connector at the front of the steel line is almost impossible to reach since it is between the steering rack and firewall and no matter how hard you try it simply cannot be removed again due to corrosion.

We're almost there, what would you do now?

Divide and concur. By dropping the line brackets under the drivers floorboard there is room to cut the steel line in half at any accessible location. Now you can test the rear half and the front half of the line seperately. On this vehicle, the front half had the leak, so nylon line was installed from that point all the way up to the purge valve connector. The purge valve connector itself isn't a typical part and if you don't want to purchase the O.E. line must be reused.