logray

There are a lot of indicators for a bad Air Oil Separator. 1. Excessive white smoke out the tail pipes. 2. Excessive vacuum on the oil filler cap, making it very difficult to remove, and once it is off the engine idle will become very rough and might stall. 3. Excessive oil in the tube leading from the AOS to the plenum behind the throttle body, and even excessive oil in the intake plenum. (a light film of oil is considered normal) 4. High pitched screeching noise from the AOS unit itself. If you have any of those issues then more than likely your AOS is shot and there is only one solution which is replacement. Proceed with the measurements! However, if you are suspecting a marginally bad AOS or perhaps just validating that your AOS is fine (ruling it out as being bad), there is a very definitive and inexpensive test to determine whether the AOS is the culprit for your poorly running engine. You can use a slack tube manometer to measure the crank case vacuum. These devices are available commercially for around $60 or more, but you can make your own for about $20 or less using some readily available parts from your local hardware store. You might even have some of these materials already on hand. From Home Depot: 1.) 20' of 1/4" inside diameter clear plastic tubing for about $3.00. You actually only need about 4 or 5 feet of tubing and the actual inside diameter doesn't matter much. 2.) $3-10 for various brass hose fittings to adapt the clear tube to a "donor" oil filler cap (the fittings are in the same section as the clear plastic tubing - plumbing and irrigation dept.). If you don't want to get fancy like I did, all you need to buy is one 1/4" by 1/4" barbed hose "coupler" often referred to as a "hose repair kit". It looks like this: 3.) Optionally for ~$7 a simple ball valve so you can start/stop the crank case suction (a good idea if you believe your AOS is bad, you can quickly stop the flow of water before it enters the crank case. Although an ounce of water into the the engine probably wouldn't hurt anything, some people use ATF or even a light weight motor oil (which changes the measurement slightly, not sure if significantly enough though). I chose to use distilled water with a few drops of food coloring). 4.) $2-5 for some epoxy to glue the brass fitting to the oil filler cap. 5.) A donor oil filler cap from Porsche, P/N 99610703552 for about $10 retail. The one I bought came with a new gasket as well. How to build the slack tube manometer. 1.) Take the oil filler cap and drill an appropriately sized hole in the center of the oil cap for the brass tube fitting. For example if you selected 1/4" tubing (the size of tubing should not matter), you can use a bit slightly larger than 1/4" so the barbs will barley fit through the hole with some resistance. The hole should be just large enough so you can snugly fit the hose fitting into the oil filler cap, but not so loose that there is a large gap around the fitting. 2.) Next using water and compressed air clean the oil cap inside and out of plastic shavings or debris. You can use any other method to clean it, the point here is you don't want any debris going into your engine. Then generously epoxy the brass fitting to the oil cap and allow it to dry overnight. You'll see that I got a little fancy and bought a brass fitting that had a threaded end so I could adapt both a cap for when I wasn't using it or the ball valve for the manometer for when I am taking measurements. 3.) Next take about 6' of tubing and affix it to a pole or board so it forms a U at the bottom. There will be about 3' on each side of the U. The actual length of the tubing is not important, but the more tube you have the larger of a measurement you can take (or have more time to stop the flow of liquid should there be extreme vacuum). One side of the tube is open to the air (hence the term slack tube). The other end of the U attaches to the oil filler cap. To attach the tube to the plywood or board I used U nails, but you can get creative (tape/wire/nails/screws/etc.). At the bottom of the board I attached a jar lid so the tube had something to form the U, however this is not important as long as it makes a "U" shape (does not have to be a perfect U). Mark off 1" indicators on the board up to about 12". Instead of making these marks you can also just measure the distance the water is drawn towards the engine after the fact. I've seen some write ups where people just drape the tubing over the side of the engine bay and hold the other end near the oil cap so it forms a U just by gravity alone (if you don't want to get fancy like I did). 4.) Fill the tube with your liquid (I chose distilled water with a little red food coloring). The level of water is not important, but you should have at least 8 inches on either side. This will permit a measurement of up to 16 inches of water which would be a pretty extreme crank case vacuum scenario. I lined up the water level with a starting point at 6" and marked the starting point with a star. 5.) At this point get the car up to operating temperature (about the 8 on the 180 in your temp gauge). Here is a quick video on how this actually works. With the engine idling, you'll see the water move towards the engine once I open the ball valve at the oil filler cap and the pressure between crank case and atmosphere/water stabilizes. Then you'll see it drop back down once I close the ball valve. You need to add the distance the fluid traveled on both sides of the tube or just multiply one side by two in order to get your reading. A normal reading is between 4 and 7 inches of water drawn (I'm sure your elevation has an affect on this, but don't have the formula handy and it probably isn't too significant to worry about). An abnormal reading indicating the AOS is bad or on it's way out would be 9-12 inches or more. The measurement I took was 2.5 inches traveled (multiply by two to get 5" which is normal). The marks made it easy to see quickly where the measurement was, but you could just just mark the before and after measurement and use a ruler or measuring tape after the fact. Thanks to Wayne @ Pelican Parts for information about this diagnosis method located in the following link. http://www.pelicanpa...l_Separator.htm