I bought my 1999 Z3 Coupe with my eyes open - obviously, a car with 196,000+ miles could have some serious mechanical problems. The superb condition of the body, and largely, the interior, coupled with some obvious mechanical needs, made the car really desirable to me - one that I could work on without having to become a paint and body expert.
When evaluating the car before buying, I took heart from the fact that it ran smoothly and strongly, with no obvious smoke or bad noises. I think that the high mileage was actually a plus for that - cars stay in good shape longer when they are regularly used, and it would have taken some highway mileage to reach that big total.
As the months progressed, I was gratified to find a couple more things. Even with the radiator leak, the car never overheated, and even when left without starting for two or three weeks, it always fired instantly (I keep it on a Battery Tender while parked). That means that the fuel system held pressure. It is very common on older cars for the injectors, or some other component in the complex fully closed fuel system, to leak a bit. When that happens, it takes several seconds of cranking to re-pressurize before the car will start. This one, turn the key, and vroom!
Still, as I began my disassembly for the first major batch of maintenance, I wanted to do what I could to assess the condition of the engine. The first task I performed was an oil change, capturing a sample to send off to Blackstone Laboratories (www.blackstone-labs.com). For about 25 bucks, this group will analyze your oil sample to look for contaminants and signs of wear.
Of course, I didn't know how long this oil had been in there, nor the mileage. It was blackish, but not tarry, not did it smell burned. Chris, the prior owner, asserted that he ran Mobil I synthetic oil, and changed it around every 5,000 miles.
In terms of wear, the sample showed about what I expected - some trace of metals from the bearings and other metal parts, and perhaps some aluminum from the pistons. But, the very good news was that there was not even a trace of water or antifreeze in the oil, nor were there combustion byproducts that would be present if the head gasket was getting weak.
I will send another sample after around 3,000 miles on the new oil (Mobil I 10W/30). That should give a better indication of how quickly the wear is happening.
After that good news, I started removing bits of the cooling system. I was very apprehensive about one thing: BMWs are very sensitive to having the wrong antifreeze. If you guess and get it wrong, you can really damage the aluminum parts of the engine, and the typical Prestone stuff is definitely the wrong choice. I had seen pictures of severely pitted surfaces under the water pump and thermostat, but my Z3 was very clean. Another good checkmark! From now on, I'll definitely use genuine BMW antifreeze, and change it every year or so.
Next, a lot of degreasing - that oil leak at the valve cover had slimed the front and passenger side of the engine. Once that was cleaner, if not clean, I could remove the valve cover. More good news there - everything was clean and shiny (as seen in the next-to-last picture in this post), and no "goo" that spoke of long service intervals or coolant leaks. The difficulty in shaking that cover loose, and the extremely brittle nature of the gasket, makes me believe that the gasket was the one that came with the car when it was built in September, 1998 as a 1999 model. More good news, though - I was able to get the gasket off without dropping bits of it down into the engine!
Next, I carefully inspected the deep wells where the six spark plugs resided. It is very common for the valve cover to leak there, and fill the well with oil. This can cause misfires, and also deteriorate the "boots' that connect the individual coils with each plug. All six wells were clean and dry! The boots were in such good condition that I didn't even have to replace them. Here's a picture of one, still attached to its coil:
Out came the plugs, and... more good news! All six of them had very consistent wear and coloration that indicated excellent combustion. The plugs were well-worn (look at the center electrode in the picture below) but showed no signs of burning or contamination. Clearly, these combustion chambers are still pretty tight.
Interestingly enough, though, these are the wrong plugs for the car! Notice that it has a single electrode, just like the plugs in a 1963 Chevy. They are supposed to look like this NGK BKR6EQUP, with four electrodes:
There is also a Bosch plug with two electrodes, but the Internet experts seem to prefer the NGK. Of course, those experts also claim that running a plug like the first photo would cause the car to misfire and otherwise run badly, but I'm telling you, even with the wear of many miles the car ran very sweetly with those other plugs. But, I splashed out for the NGKs at six bucks a plug. Those suckers better last forever at that price!
SO, everything was looking good, but I wanted to do one more test. I had never owned a compression tester, and now seemed to be a great time to invest. Harbor Freight, 40 bucks on sale, for the 94190 leak-down compression tester. Now, I know what you're saying, and you are right - no way that a $40 compression tester will be accurate. Indeed, the reviews on the HF website had some folks comparing its readings with a Snap-on tester, and found variances up to 25%.
But I didn't care. I wasn't going to spend hundreds of dollars for the Snap-on tool, and this would at least give me a way to compare from cylinder to cylinder in my car. The instructions said to do it with the engine hot, but it was just so easy to do it while I could see both cams to determine when the valves were closed for each cylinder. I decided to do it cold, and see what happened.
Check out the pic below. There are two cams, and four valves per cylinder. In the picture, you can see the rubber adapter hose sticking out of the spark plug well. The arrows I added point to the cam lobes for the four valves for that cylinder. I reasoned that having them point so that the cam follower was on the "base circle", where the valve is closed, would be the place to check each cylinder.
I tried to imagine a line from the centerline of the cam, through the point of the lobe, and then turned the crankshaft (with a rachet and 22mm socket) until the angle between the intake and exhaust lobes was around 90 degrees. Seemed to work, and... more good news! The cylinder leakage across all six cylinders was very consistent, within about 5 percent of each other. The gauge reading was around 40% leakage, but that number could be wrong anyway, and would certainly be lower if the engine was hot and had recently had some oil splashed around.
You test the cylinder by adjusting the right-hand gauge to zero (all the way to the right), and then connect it to the adapter hose screwed into the cylinder. This is not how the instructions read, but the useful reviews on the HF site made the correct procedure clear.
So, I'm very happy with my diagnosis, and hope this engine will run a long time before having to come apart. As long as I can avoid a sudden failure (for example, the oil pump seizing) I have hopes of another 50,000 or even 100,000 miles from this engine!
No comments:
Post a Comment