|In last month's
article, I discussed performance and oil consumption issues. This month I'll
take a look at compression and leak-down tests, and discuss the merits and potential
downfalls of each of them. We'll also talk a bit about how carbon deposits can affect your
One of the most common tests
that can be performed on a engine is the standard compression test. This particular test
measures the amount of pressure that is built up inside the combustion chamber when the
engine is turned over. The typical compression tester is a pressure gauge that is attached
via a short hose to a plug that is screwed into the spark plug hole. As the engine turns
over, the compression gauge will read the maximum pressure exerted within the combustion
chamber. The overall value is one method of testing your engine to determine the condition
of the rings or valves.
Your 911 needs to be setup
before you can start the compression test. With the car cold, loosen the spark plugs with
a spark plug socket and extension. Then tighten them up very lightly. You want to test the
engine when its warm, yet if the spark plugs are very tight in the heads, you can
damage the threads in the heads by removing them when the engine is hot. Loosening them up
a bit when the engine is cold will minimize any damage you could possibly do to the
threads in the heads. Although you might think that its good practice to use
anti-seize compound on the plug threads, Porsche specifically recommends against this. The
anti-seize compound seems to interfere with the proper grounding of the plugs. Also,
temporarily remove any heater hoses that might get in the way of removing the spark plugs.
Warm the car up to operating
temperature and then turn it off. Wait about 5 minutes or so, as head temperatures tend to
spike right after you turn the engine off. At this point, the engine fan has stopped, and
the heat tends to build up with no place to dissipate to. Removing the spark plugs right
after turning off the engine can cause the threads in the aluminum to gall. After about
five minutes, remove the spark plugs from their holes. Now, disconnect the cable from the
capacitive discharge (CD) unit (1969-1983 911 and all Turbos). If youre working on a
1965-68 911, then simply disconnect the power line (+) from the coil. If youre
testing a 1984-89 911 Carrera, then remove the small square DME relay from under the
passenger seat. Doing this will disable the cars ignition system, and prevent the
spark plug wires from firing. Its also a wise idea to remove the fuel pump relay at
this time (for 1969-83 911s). You are going to be cranking the engine over several times,
and you dont want raw fuel to be dumped into the system.
Having a helper around is
useful, as you can watch the gauge while he or she cranks the engine. I recommend that you
attach a battery charger to your battery to avoid running it down. Dont fire it up
at 50 Amp, but instead leave it on about 10 amps, which should help it recover when
its not cranking.
With the engine warm, install
the compression tester into the spark plug hole. A bit of patience and skill are required
in order to properly manipulate and screw in the compression tester so that you dont
cross thread and damage the threads in the cylinder heads. With the compression tester
installed, crank the engine over 12-16 times. Make sure that you place your foot all the
way down on the throttle. This will allow maximum air flow into the engine, otherwise your
compression readings will be off. The engine should make six to eight full complete
compression strokes (12-16 turns of the crankshaft). You can tell when the engine is on a
compression stroke because the compression gauge will jump and show an increase when the
cylinder is compressed. Carefully watch how the compression tester gauge increases, and
record the maximum value when you have completed the last compression stroke. The gauge
will jump at first, and then increase slowly until cranking the engine over more and more
has no additional effect on the reading. Remove the compression tester and repeat for each
of the other cylinders.
So what to do with the results?
In general, compression tests are limited in what they can tell you. It is important to
remember that different compression testers may give different readings as well. Cranking
the engine faster (with a stronger battery or high powered starter) may also skew
readings. The most useful piece of information that you can glean from them is how each
cylinder compares to the others. All of the cylinders should give readings that are very
close to each other. This would generally indicate an engine in good health. A good rule
of thumb is that each cylinder should read a minimum of 85% of value of the highest
cylinder. So, if the highest reading is 150 psi, then the minimum acceptable reading would
be about 128 psi.
It is important to note that
this would be an acceptable figure, but not necessarily ideal. In all practicality, all of
the cylinders should be very close to each other (within about 5-10 psi). On a newly
assembled and run-in motor, compression numbers are usually within this range. As the
engine ages and certain parts wear faster than others, one or more cylinders may
experience a bit more wear than the others. This will definitely show up in the
compression tests. Needless to say, if you have all of your cylinders in the 150 psi
range, and one cylinder is down around 120 psi, that should give you cause for concern.
The important thing is to remember is that you want to gather consistent readings across
all of the cylinders, without focusing on the actual values. If a reading is significantly
off, go back and test that cylinder again to make sure that the measurement was not caused
by some sort of fluke, which is often the case.
So what causes variations in
compression tests, and why cant they be used as the final word on engine rebuilds?
The problem is that there are several factors that effect the final pressure read by the
tester. Engines running with very aggressive camshafts have a tendency to give low
compression readings. This is because there is significant overlap between the intake and
the exhaust stroke on the cam. During high-rpm operation of the engine, this overlap works
to give the engine more power. However, when turning the engine at a low RPM, as with a
compression test, the overlap causes some of the pressure in the combustion chamber to
leak out before the valve is closed. An early 911S engine, for example (with its
high-overlap cams ) has a tendency to give lower compression readings than the 911 CIS
engines (1974-83), despite having a higher compression ratio. This is caused by the
aggressive overlap of the camshaft.
Altitude and temperature also
affect the compression readings. Manufacturers specifications are almost always
given at a specific altitude (14.7 psi at sea level), and 59° Fahrenheit. Both
temperature and barometric pressure change as you go up in altitude, so you will need to
correct your measurements if you wish to compare it with a factory specification. The
following chart provides conversion factors for correctly compensating for changes in
Altitude Compensation Factors
compression reading of about 150 psi at sea level in Los Angeles would measure
significantly less in the surrounding mountains. For example, at an elevation of 6000
feet, the expected reading would be 150 psi X .8359 = 125 psi. The cylinders would be
reading low if compared to sea level measurements, yet perfectly fine at this altitude.
Another factor that can alter
compression test readings are incorrectly adjusted valves. If the valves are not opening
or closing at the correct time, then one cylinder may read vastly different than another.
Make sure that your valves are adjusted properly prior to performing the test. For the
complete procedure on adjusting your valves with the engine inside the car, take a look at
Project 18 in the book, 101 Projects for Your Porsche 911. Along the same
lines of thought, premature camshaft wear can also lead to variances in compression
readings, however, this type of wear is not normally common on the 911 engine.
You can determine if the rings
are causing low compression readings by squirting about a tablespoon of standard 10-30W
engine oil into the cylinder. Crank the engine 2-3 times to spread the oil around inside
the combustion chamber. Then retest the compression. If the readings shoot up
significantly (45 psi or so), then the problem is most likely with the piston rings
seating to the cylinders. Squirting the oil inside the combustion chamber in this manner
allows the rings to temporarily seal quite a bit more than they would dry. If the
compression readings do not change, then most likely culprit is a leaky valve.
Without a doubt, the most
comprehensive test that you can perform on your engine is a leak down test. While somewhat
similar to the compression test, it eliminates nearly all of the extraneous variables that
may alter the final compression readings in a typical compression test. In simple terms,
the leak-down test involves pressurizing the cylinder and measuring the amount of air that
is leaked out past either the rings, the valves, or out a gap between the heads and the
The leak-down test equipment
uses an external air compressor to pressurize the cylinder. The engine is held stationary,
and the test is not dependent upon outside variables like the cranking speed, altitude,
temperature, or the camshaft overlap. In fact, the leak-down test can be performed on just
about any engine, whether or not it is inside the car or not.
Unfortunately, the leak-down
test equipment is somewhat specialized, requires an air compressor, and is not exactly
inexpensive. Most local repair shops have a leak-down tester, but its not common to
find one in your neighbors garage. The good news is that most shops will be able to
perform a leak-down test on your engine for a nominal fee. The 911 engine doesnt
require any special leak-down adapters, so you should be able to take your 911 to any good
foreign repair shop, and they should be able to do the test for you. Similar to the
compression test leak-down test should give you information on the condition of the rings
and valves, but from a slightly different perspective. The leak-down
test can be performed on an engine that is not installed in the car. However, if the
leak-down test is performed on an engine that isnt warmed up, then the test may not
give accurate results.
The leak-down test is performed
by initially setting the engine to top-dead-center (TDC) on the compression stroke for the
piston that you are checking. Make sure that its exactly at TDC, otherwise the
engine will begin to turn over as soon as you pressurize the cylinder. You want to make
sure that both the intake and exhaust valves are completely closed (as they should be at
TDC) otherwise air will immediately leak out of the cylinder. To make sure that you are at
TDC for cylinder number 1, remove the distributor cap, and rotate the engine clockwise
until the rotor is lined up with the small notch.
When you are running the test,
it is a wise idea to make sure that the crank doesnt turn at all. Have an assistant
hold the crank steady or place a flywheel lock on the engine if its out of the car.
Connect the leakage tester to the engine in the same manner that you would with the
compression tester. Pump up the cylinder and let the leakage tester measure the amount of
air lost. The gauge on the tester should give readings in percentage numbers. A newly
rebuilt engine should have leak-down percentages of around 3-5%. An engine in good running
condition should show 10% or less. Numbers around 20% indicate some wear of the engine,
but are still adequate for good engine operation. Leakage numbers of around 30% indicate
that there are problems brewing, and that a rebuild may be necessary. Needless to say a
large leakage amount like 90% indicates that there is a hole in the combustion chamber,
and the engine is probably not firing on this cylinder at all. Rotate the engine
crankshaft clockwise 180° when youre done, and check the next cylinder. Repeat the
process for each of the six cylinders.
Another good quality of the
leak-down test is the ability to pinpoint the exact problem with the engine. When the
cylinder is compressed with air, you can usually hear where the air is releasing from.
Leakage past the intake valves can often be heard at the intake manifolds through the fuel
injection. Exhaust valve leakage can sometimes be heard through the tailpipe. Leakage past
the rings can sometimes be heard in the crankcase breather hoses. The most obvious leakage
occurs when the cylinder heads have broken or pulled, and the air leaks directly out of
the combustion chamber in-between the cylinder and the head.
While the leak-down test is
probably the best indicator of engine condition, it shouldnt be the final word in
your evaluation. I have heard from many people about great running engines that for one
reason or another do not test well on the leak-down tester. Its important to
remember that the leak-down tester does not test the engine when its running
it only does a static evaluation. As with any air cooled motor, its operating
characteristics vary widely. Use the leak-down test as one indicator and back it up with
other tests and observances.
I thought it important to
mention some things about carbon deposits build up inside engines. Just about every single
engine I have ever seen torn open has had a significant layer of carbon buildup on the
pistons and the inside of the heads and valves. Particularly with todays ever
changing formulations of gasoline, the additional carbon build up appears to be a problem
in almost all air-cooled engines.
The 911 engine has a few
problems of its own, specifically related to carbon build-up. Carbon deposits will form
naturally inside the combustion chamber as a natural by-product of the combustion process.
Both engine oil and gasoline are hydrocarbons, so burning either of them incorrectly can
result in a buildup of excess carbon deposits. These deposits are often caused by
excessive oil burning in the combustion chamber, which is a sign that your engine needs a
rebuild regardless. In addition, a rich mixture setting can also introduce more of the
black soot that creates the carbon buildups in the engine. Short-trip driving and extended
idling (not ideal running conditions for an engine) can also increase the buildup rate.
While excess carbon deposits can be cleaned and removed without a complete overhaul, very
often they are yet another sign that something else on the engine needs attention (like
rings and guides).
Carbon deposits can cause the
engines valves to become shrouded, and covered with carbon. In an opposite manner to
porting and polishing the heads, the carbon buildup actually disrupts the flow of fuel
mixture, and can restrict the airflow into the combustion chamber. The horizontal layout
of the 911 engine in a boxer configuration also lends itself to being susceptible to
problems with carbon deposits. It is not uncommon to find a 911 engine that has not been
run for a long time that has low compression. Even if the engine has had a relatively
short number of miles put on it since its last rebuild, you may discover that it has very
low or zero compression in one of its cylinders. Often the reason for this is carbon
deposits. When an engine is left idle for a long period of time, moisture has a habit of
getting into the combustion chamber, and gets absorbed by the carbon deposits. This
absorption results in the carbon becoming loose and flaking off. The boxer orientation of
the 911 engine means that the exhaust valves are located at the bottom of the engine.
Carbon deposits that flake off have a bad habit of getting lodged in-between the exhaust
valve and its seat. This creates a compression leak in the combustion chamber.
Its important not to drive
the car for extended periods of time (hundreds of miles) if you think that a piece of
carbon might be lodged in-between the exhaust valve and its seat. The reason for this is
simple. The exhaust valve (unlike the intake valve) becomes very hot, and needs to cool by
coming in contact with its valve seat. If the valve doesnt seat properly, then it
will be thermally isolated from its heat sink (the seat in the head). Prolonged driving in
this condition will cause the valve to become burned, and will develop a typical pie-piece
shaped notch in the valve. Valves damaged in this manner are basically destroyed, and will
not seat properly even if the carbon is removed. In the worst-case scenario, the valve
will become so hot that the head of the valve can break off. Having the head of a valve
dance around the inside of your combustion chamber will usually destroy the piston and
send chunks of metal circulating throughout your motor. Needless to say, this is not a
As mentioned previously, worn
valve guides, or worn out rings allow excess oil into the combustion chamber that vastly
increases carbon build up. Of course, the solution to this problem is a full rebuild, or
at best a top-end valve job. In addition, how you drive your car can affect the build up
of deposits. Short drives around town have a tendency to increase the level of carbon
buildup. Slow-speed, short-trip driving has a tendency to not let the engine heat up to
normal operating temperatures. Excess carbon deposits can often be burned out
by driving on the highway for about an hour or so. This should allow the combustion
chamber to heat up enough to burn away the carbon deposits.
If your engine has been sitting
for an extended period of time, you may want to try using a gasoline additive to your
fuel. Berryman B-12 Chemtool and Techron both have good reputations for helping to
dissolve and remove deposits. One of the best things to do is to take your 911 on an
extended, spirited drive of at least an hour or more along the freeway. Try to vary your
RPMs, but make sure that you keep them relatively high to help raise cylinder head temps.
The cleaning process combined with the heated cylinder heads should be enough to clean out
any excess deposits. When you return from your drive, run the compression or leak-down
test again, and you may be surprised at the improvement in the numbers!
Next month, take a breath, we'll discuss the various
costs involved with tackling a rebuild of your engine!
|Want more technical articles like this one, but
don't like viewing them on the computer screen? Pick up a copy of Wayne Dempsey's
book, 101 Projects for your Porsche 911. See www.101projects.com for more details. Now available!