Adjustments to Pass Emissions Testing (Fuel Injection)
by Dennis Haynes
"What is your opinion on adjusting the spring tension inside of the air flow meter? I've been experimenting and have noticed that on an air flow meter that wouldn't respond to CO (mixture) adjustment, tightening the spring tension allows the CO adjustment to become possible. The spring tension can be tightened or loosened and is indexed by a plastic gear which is held in place by a spring. I figured you may be familiar with this. I have done this in the past with an old BMW 5 series that was not responding to adjustment. in this case it seemed to help a lot.I think that so far I was able to get the 2.1 wasserboxer to be within specs only after tightening the spring tension inside the air flow meter by three clicks. CO about 1.00-.095 % and HC around 200 ppm. I noticed that I encountered slow speed surging while traveling in gear 1,2 or 3rd without giving any gas at slow speeds. After richening the mixture a little bit, it is a little better, but it still exists. I am sure the throttle switches are properly adjusted. Could this be a result of changing the spring tension inside of the air flow meter? Also I find that trying to get the 2.1 to run within spec is a time consuming process and I find myself having the think hooked up to the exhaust analyzer for hours at a time. Any ideas? I have seen some info on the largest version of the 2.1 that was offered in Germany and it was a 2.1 with 10.1 compression and running Digijet injection, no Lamda and no catalyst. I'm sure that setup really would run very smooth at idle and it could be tuned to pass the MD emmisions. My 2.1 with digifant is nice and runs good, but I just wish I could get it to run a bit smoother at idle, but I seem to feel that these engines especially with the newer FI setup run rough."
As for adjusting the air flow meter, "KEEP YOUR HANDS OFF". Put it back where it was. The whole purpose of the Lamba system is to continuously adjust the mixture to keep it on the money. A mixture too rich will cause the Cat to be over worked, ineffecient at reducing CO, and overheated. I've seen the CATs burn paint off the rear valance. A mixture that is too lean will cause engine miss, which will also overheat the Cat, and increase the production of Oxides of Nitrogen, a mojor component of smog. Assuming your FI and Lamda circuit is working properly, you must have the airflow meter way out of adjustment if you can now adjust the CO without the system re-correcting it. When adjusting (actually testing), the mixture should be measured before the Cat. If your Van is running correctly, you should have near zero CO and Hydrocarbons after the Cat, with it warmed up and ignited. It is normal for your engine to miss slightly at idle, however, 200PPM hydrocarbons is excessive. In NY, you fail at 220PPM and 1.2%CO.
EMISSIONS TESTING:
Basically there are two types of Emissions tests in use. Although they appear in many different forms and different manufacturers and types of equipment is used, all the tests are looking for the same gasses. All official test procedures also call for a physical inspection to determine that all of a vehicles emission control equipment is still in place. This is enforced to varying degrees but should be considered when thinking about removing equipment to save repair costs (removing Fuel Injection) or looking for more performance, especially changing camshafts and compression ratios.
The most common type of test is the one that measures the exhaust gas content with the engine idling. The test equipment is based on what is known as a Four Gas Analyzer. The machine tests for Carbon Dioxide (combustion efficiency), Carbon Monoxide (Incomplete combustion), Oxygen content, and unburned Hydrocarbons. This tester is usually combined with a computer or other operator interface so that test parameters such as vehicle year, test limits etc. can be set. This machine also uses an ignition pick up or other device to monitor engine speed during the test. This insures that the engine is at a proper idle speed as a high idle will enable a poor running engine to pass the test. This type of test will eventually be required everywhere.
The other test is more complex and requires much more complicated equipment including a Chassis Dynamometer. This test is often refereed to as the IM-240. This test actually simulates driving the vehicle in real world conditions for four minutes. As the vehicle will be undergo changing speeds and loads, the test will be able to more accurately determine the performance of a vehicles emission control systems. This test measures everything the above test does with the addition of being able to measure the actual volume of certain pollutants being produced (grams per mile). It will also be able to measure the oxides of nitrogen (component of smog) ( 3 way cat and EGR system test) as the engine will be operated under load. This is the test system that will be used for most state centralized emissions testing programs. This test will be required in many areas due to the failure of the more reasonable testing procedures to accomplish their goals, mostly due to consumer pressure on the station operators and their willingness to cheat.
Now, what do all these tests mean and how do I get my car to pass? Luckily, believe it or not, VW has done an outstanding job in my opinion of keeping their emission control systems as simple as possible. Although this was probably done with the idea of keeping production costs down, it has made systems relatively simple. Many late model VWs do not have EGR valves, Air injection pumps etc. Except for early fuel injection systems, the VW product line has always had much less non-sense stuff than other makes. What is there, has a purpose and is usually easy to diagnose and maintain. Sometimes, as easy as something may be to repair, it may not be cheap. One problem with this simplicity is that it is more susceptible to improper procedures causing nasty results. For example, if a GM car can idle at all, it will probably pass the first type test with ease. If a Vanagon doesn't have the timing and idle speed set by the book, it will fail horribly.
Let's look at the different gases and see what abnormal readings on these gasses mean. Whether or not your vehicle passes the emissions test, it is a good idea to keep a record of the readings from year to year. This information can be used to spot trends such as fuel system degradation or an engine getting tired.
Carbon Monoxide:
Measured in percent, this is the result of incomplete combustion. In other words, the fuel mixture has been compressed and ignited, but could not complete the combustion process due to the lack of oxygen. This failure is always mixture or ignition timing related. The ideal reading for CO is 0%. A normal reading for CO is usually around .25 to .75% with 1.2% being the failure mark in most areas for vehicles 1984 and newer. Earlier cars are allowed higher readings. Any car equipped with an Oxygen sensor will try to run under 1.0% as measured before the Catalytic converter (use that test connection for set-up). A vehicle with a CAT, O2 sensor, and an AIR system (Air Injection Reaction) will read almost 0% with everything working properly.
Most '70s cars should run between 1.0 to 2.0% with 2.5% being the usual failure point. I find that 1600 dual port will run fine in this range and with Fuel Injection, should be able to run about .5 to 1.5 %. Remember this is at idle so it is not an issue with overheating the heads and burning valves. '50s and '60s cars will need to run richer still to remain smooth running. Anything over 3.5 to 4.0% in any car is cause for concern although most test programs allow up to 6.0% for early vehicles. This is a lot of forgiveness in an inspection program. Over advanced ignition timing will also cause high CO readings at idle due to the idle speed control being set low in order for the engine to idle at a reasonable speed. VWs are especially sensitive to this. Keep this in mind when using different distributors.
Hydrocarbons:
Measured in PPM (parts per million), this is the result of unburned fuel making it out the tailpipe. This is different from the CO problem in that here the fuel has not started combustion at all. This failure is usually caused by an ignition (misfire), or an engine problem such as a bad valve. An engine running too lean will often have excessive HC due to what is called a "lean misfire". Any leaks in the intake system or bad injectors will also cause excessive HC due to uneven fuel distribution.
Vanagons are very susceptible to the "lean misfire". Here the ideal reading again is 0 PPM. Most late model cars should run under 100 PPM with readings between 10 an 50 PPM being typical and a reading of over 220 PPM being the failure point for must inspection programs. Again, older cars are allowed more freedom although I find that most any car should be able to run under 400 PPM. My 1923 Haynes runs at 300 to 350 ppm HC and 4.0% CO. High HC readings are a little more challenging to trouble shoot because there are so many possible causes including a mixture or worse, a cylinder balance problem.
Carbon Dioxide:
Carbon Dioxide is the end product of all Fossil Fuel combustion. CO-2, measured in percent is an indication of an engine's combustion efficiency. The ideal is 14% with most VWs running in the 12.5% to 13.5 range. CO-2 output is affected by timing (valve and ignition), compression, mixture, engine condition and temperature. The IM-240 tests will also measure the actual volume of CO-2 produced and limits will be set according to vehicle design. If centralized testing is coming to your area, be very careful of any modifications made to your vehicle. Even changes to your tires sizes and gear ratios will affect this test.
Oxygen:
Measured in percent, is actually excess oxygen left over from the combustion process. If this reading is very low, then you will also experience high CO. In boilers and other equipment, we avoid high readings as that indicates excess air which reduces heating efficiency. In automobiles, we usually are looking for about 18%. Normal atmosphere is 21%. This is controlled in late model vehicles by the Oxygen Sensor circuit. The O2 sensor actually measures the Difference in O2 content of the atmosphere and the exhaust. When the atmosphere has more O2 than the exhaust, a small positive voltage is generated.
The Vehicle Computer (ECM), senses this signal and then adjusts the mixture accordingly. You can measure the fuel mixture of an O2 sensor equipped vehicle with a Digital Multimeter (DMM). If the mixture control system is working properly, you should see the signal fluctuate from .1 to .7 volts. This voltage correlates closely (not by design, just luck) with the actual CO percent (Example: .5 volt is approximately .5% CO). This is great news as we could install a universal O2 sensor in any car and use a DMM to set the mixture. If you wish to do this, ensure that you mount the sensor in a "hot " area of the exhaust as it has to be heated to 600 degrees F in order to function. This is why many cars included 86 and later Vanagons use a three wire sensor. The three wire sensor actual has a heating element built in to preheat the sensor. This enables the sensor to go on-line on a cold engine much sooner and keeps weather and driving conditions from effecting the sensor. I find that earlier Vanagons often have problems with sensor performance as the sensor is mounted far enough down stream in the exhaust that they don't get adequately heated during cold weather or when the engine is idling. This also creates havoc with emissions tests as vehicles are often tested without being thoroughly warmed up (car must be driven to properly warm up all emission equipment, especially the Cat).
So what does all this mean? Well we went for our test and we failed. Our van seems torun okay but it failed. Lets review why.
Failed the CO!
Check the Mixture and Ignition Timing. Especially on VWs, over advance timing will cause high CO readings. When setting up ignition timing, mixture, and idle speed, follow the book. On Digijet Vanagons, disconnect and bypass the Idle Stabilizer when setting the timing and idle speed. Resist the temptation to "advance" the timing for performance. Also make sure both the vacuum and mechanical advance mechanisms are working properly. On Digifant Vans, Disconnect the Temperature Sensor before setting the timing. Again, resist the urge to over advance the timing. Check the O2 sensor system to ensure it is working properly. Resist the urge to adjust the Airflow box. If the O2 sensor system is working properly, it will compensate unless some one has already made a mess of things. Also check the oil, diluted oil will also cause high CO readings. Worn rings and valve guides will contribute to this also. However this will only be a problem on very tired engines. Setting the idle speed around 1000 rpm will also help keep the CO readings down.
Failed the HC!
Check the ignition system first. This usually caused by an ignition miss. Then check the engine. Poorly seated valves, leaking head gaskets, or other compression problems will put the HC readings through the roof. Also check for vacuum leaks and the injector spray patterns. Remember, HC failures are usually caused by compression or ignition problems. Improper fuel distribution/vacuum leaks will cause misfires that will cause excessive HC.
Failed both CO and HC!
If the HC is very high and the CO is close, you probably have an ignition/compression related problem. If the CO is very high and the HC is close, than you may only have a timing/mixture problem. Use the above procedures accordingly and you should be able to resolve your emissions failure blues. Next time, we will talk about all that infamous junk they put on our cars to control the emissions. I will explain the purpose of these devices that you can better troubleshoot your cars and also to help you decide which equipment is worth maintaining or discarding when searching for more performance. One thing to keep in mind is that much of this equipment is put on our cars because we don't maintain them. That's why some car manufactures are shooting for 100,000 mile maintenance intervals.
