Jump to content

Engine Diagnostics: Compression Gauge Versus a Cylinder Leak Down Tester

Recommended Posts

  • Administrators

Most engine troubleshooters understand and use a compression gauge to diagnose serious engine seal problems.  I use a compression gauge for a quick and general overview of engine condition.  In addition to the relatively inexpensive compression gauge, I have used a cylinder leak down tester for over thirty years now.


post-1-0-30045100-1370907183_thumb.jpg post-1-0-89231300-1370907188_thumb.jpg


A quality compression gauge is valuable. This gauge has been with me for many years. Ford truck buffs will appreciate the 18mm spark plug thread adapter. Triton aluminum head owners would value the iron FE and big-block heads—with 18mm iron spark plug threads and tapered seat ends—no need for a gasket, a proven Ford approach until the aluminum cylinder head era!  (Click on photos to enlarge.)


A compression gauge requires a cranking engine to force compression pressure into the gauge.  Used properly, the compression gauge reading indicates the PSI (U.S. and British measurement) of gases compressed into the cylinder head's combustion chamber.


Basics for using a compression gauge:  1) Disable the ignition to prevent engine startup, 2) hold or block open the throttle during cranking to assure adequate induction air volume for test accuracy, and 3) crank the engine for seconds per cylinder, not for a lengthy period.  As the engine cranks, watch for a peak reading at the gauge, which will occur within a few seconds of cranking.  A typical automotive or truck engine is designed to crank over at 80-200 rpm to achieve ignition/combustion, so several seconds of cranking is plenty for a compression check!


The compression gauge provides a reading of peak compression during cranking, not when the engine fires or is running.  Actually, the engine's compression while firing under load may be much lower than the cranking compression.  As cylinder pressures rise, like when climbing a grade or pulling a load—or crawling rocks along a 4x4 trail—the ability of the cylinders to seal will be far less if there is engine wear.


Note: Engine wear that I'm discussing here is not just worn valves, a head gasket leak or valve lift issues.  It's also the cylinder's ability to seal under load, and more specifically, the condition of the cylinder walls, pistons and piston rings.  In the case of pinpointing cylinder wall and ring seal, a compression gauge is simply not up for the task! 


post-1-0-09675800-1370907136_thumb.jpg post-1-0-64640000-1370907143_thumb.jpg post-1-0-44589100-1370907149_thumb.jpg


This is my Snap-On cylinder leak down tester, an MT324 model purchased in 1981.  Prior to that time, I used a compression gauge and an engine oscilloscope "dynamic compression test" mode to compare relative cylinder condition. The leak down tester provides pinpoint diagnosis!  (Click on photos to enlarge.)


When we move to the leak down test, our entire aim changes.  Instead of cranking compression, we're now addressing one thing: actual cylinder seal.  This means consistent valve seal, piston/ring seal, head gasket seal and no casting cracks.  The test is not a cranking test.  It's static with the piston at top dead center (TDC) of its compression stroke, valves closed.


This last point is very important.  The leak down test takes compressed air from your air compressor and an air hose source and runs that air into the spark plug hole (plug removed).  Typically, a leak down gauge will operate at 60 PSI or higher, enough force to push the piston downward if the piston is not precisely at the peak of its travel: TDC!


post-1-0-45304300-1370907172_thumb.jpg post-1-0-09414200-1370907178_thumb.jpg


This is one way to find TDC accurately: a two-stroke engine timing tool, purchased during my motorcycle repair shop days in the early 1970s! Note the spark plug thread adapters with a slotted section in the adapter to aim a dial indicator directly through the spark plug opening.  This tool finds TDC within 0.001" accuracy!  The setup is time consuming, however.  There are other, quicker ways to find TDC for the purpose of a cylinder leak down test—as long as the piston is at its peak and not likely to plunge downward when you apply compressed air!  (Click on photos to enlarge.)


It takes time to set up a leak down test.  On a Jeep engine like JJ_Jeep's 4.0L inline six, there are only two cylinders that can use the crankshaft pulley marks to set TDC: #1 and #6.  The other cylinders either require 1) looking into the spark plug hole to align the piston with TDC, 2) use of a two-stroke timing kit or similar fixture and a dial indicator in the spark plug hole to find TDC, or 3) marking the crankshaft pulley accurately for each of the other four cylinders at TDC.  Make sure the piston is on the compression stroke's TDC when running a leak down test.


Note: A crankshaft pulley 360-degree tape is available from sources like Summit Racing.  This can make leak down tests much easier.  The firing order helps identify the cylinders, and in the Jeep inline six, the order is 1-5-3-6-2-4, making #1 and #6 pistons reach TDC at the same time.  Remember, too, that we're working with four-stroke engines, and the piston must be at TDC on the compression stroke, not the exhaust stroke!


If you set the piston to TDC on its compression stroke, each of the valves will be fully seated.  Introducing compressed air at this point will fill the combustion chamber and the narrow space around the piston and above the upper piston ring.  The air will come in steadily (if you have enough air supply!) at the air system's line pressure that you set.


The principle of the leak down test is for the gauges to measure incoming line pressure and also the pressure bleed-off within the cylinder.  The bleed or "leak" is read as a percentage of the total air volume filling the combustion chamber, head gasket space, and piston-to-cylinder wall gap above the upper piston ring.


Example: Let's say that 100cc of air fills the combustion chamber, head gasket space and the piston to cylinder wall gap above the upper piston ring.  If the ring gaps and any other seepage allow 10cc bleed-off of compressed air, the "leak down" in this case is 10%.  8-10% is excellent seal for a production automotive engine with gapped piston rings. 


Caution: Always make sure the gauge couplings and the fitting at the spark plug hole do not seep air!  A leak at the gauge's spark plug fitting seat will throw off the leak down readings.


The biggest asset of a leak down test is the very fact that the piston is at TDC.  On an internal combustion engine, the wear of cylinders, or "taper", rakes outward toward the top.  This means that the widest wear point in the cylinder will be right below the cylinder "ridge".  The ridge is created as the taper grows, and the greatest wear point is where the top piston ring stops.


By pressurizing the cylinder with the top compression ring and other rings in the tapered zone, there is the highest likelihood of compressed air "leaking" past the piston rings:  The ring gaps, especially the top compression ring, are at their widest, and this is a direct cause of leakage.  With the piston in this position, the rings lose tension, their gaps are widest, and compressed air leaks by in greater volume.  The extent of ring wear or cylinder taper dictate the percentage of leak down.


One proof that this phenomenon works:  On a worn engine, back off the rocker arms and close both valves.  Keep them closed.  Raise the piston to TDC (maximum cylinder taper) and check the leak down rate.  Now drop the piston to the BDC (bottom dead center) point with the valves closed.  The piston is now in a "normal", minimally worn zone of the cylinder wall.


There will be a distinction in the degree of seal and the leak down reading, since the ring gaps close up at BDC, the ring tension increases, and the cylinder seal becomes normal or at least closer to normal.  If there is little leakage difference, and the leak down percentage is low or minimal at both of these piston positions, cylinder taper is minimal...Without tearing the engine down for inspection, you likely have an engine with minimal cylinder wall and ring wear!


Before loosening rocker arms to close valves, run a regular leak down test with nothing but the spark plugs removed.  Since JJ_Jeep's issue with the 4.0L is a #1 cylinder misfire, it's easy to set up the leak down tester for the #1 cylinder, using the timing mark on the crankshaft pulley and the timing cover marker to set the crankshaft at TDC.


Make sure the ignition is disabled to prevent engine firing...I bring the pulley to the "O"-degree position by slowly turning the crankshaft in its normal direction of rotation.  Make sure you're on the compression stroke.  If necessary, remove the distributor cap and note the rotor position.  (JJ's pre-'99 TJ 4.0L still uses a distributor cap and spark leads.)  Coil-on-plug engines are a bit more involved.  Make sure the piston is coming up on its compression stroke to exact TDC.


Read the tester instructions for recommended air supply pressure.  On gapped rings with inside gas ledges on the compression rings, supply pressure will determine the force that the rings will exert against the cylinder wall.  A cranking or running engine will typically have 120-175 PSI working compression in each cylinder.  The leak down gauge may want much less incoming pressure than this...See the manufacturer's recommendations for your leak down tester.


With the valves seated, piston at TDC and air pressure at the tester gauges set properly, we now have a steady stream of compressed air entering the cylinder.  Even if the valves and head gasket seal perfectly, a percentage of incoming air will "leak" by or "down" the cylinder and past the rings.  This is normal if rings have end gaps, which is the case for most production automotive engines.


Note: Overlapping end, interlocking or "zero gap" rings have minute leak down.  On a typical production engine, there is a controlled and predictable leakage, optimally around 8% to 10%.  Zero-gap rings on racing and performance motorcycle engines can seal as tightly as 2% leak down.  We need ring gaps on everyday automotive engines to allow for a wide range of heat cycling, expansion rates and varied operating conditions.


So, a reading of 8-10% on a newer engine is great.  Up to 20%, even 25%, on an engine with some wear can still function well if the engine also has good cranking compression readings.  On an engine with 20%-25% or higher leak, I do run a cranking compression test and make sure there is not excessive oil consumption.  Assume that a 20-25% leakage engine has a shorter remaining service life.


As with a compression gauge reading, uniform seal per cylinder is desirable under any leak down test.  If readings exceed 20% or so, start looking for leakage.  You will find leaks readily while running a leak down test. 


When an intake valve leaks, you will hear air escaping through the throttle body or a carburetor's air horn.  For exhaust valve leaks, listen at the tailpipe.  For piston ring blow-by, pull the dipstick or oil filler cap, and listen for the air entering the crankcase past the rings.  If you have difficulty hearing this air flow at any of these points, use an automotive stethoscope or even a piece of tubing, like rigid electrical conduit, as a sound amplifying device.


For accuracy, do not run the test for a long period.  Oil will leave the cylinder walls, and the reduced seal will be detectable as a higher leak down percentage.  Also, on rare occasions, a valve may have carbon or deposit buildup near its face, and the compressed air dislodges debris.  If debris lodges between the valve face and seat, there will be a false leak reading. 


Note: Such an instance would be an engine with serious carbon or deposit buildup.  One way to minimize this risk is to not apply air pressure until the piston is at TDC with the valves completely closed on the compression stroke.


Why do we care whether the engine has too much leak down if the compression cranking test with a compression gauge indicates "normal" or factory-recommended cranking compression?  There's a big difference, as I've shared.  In the case of testing for cylinder taper and piston ring gap issues, the leak down tester pinpoints specific wear at the cylinder walls and even the degree of piston ring wear—yes, a widened ring gap, even with negligible taper on the cylinder, indicates ring face wear, loss of ring tension and poor ring seal.  This raises the percentage of cylinder leak down. 


It does matter that the cylinder seal is within proper leak down specifications.  So-called "normal" compression gauge readings do not reveal how the cylinders seal under firing and engine load conditions.  There is a rise in cylinder pressures when the engine is under load, like accelerating, pulling a grade or even going through the gears.  Cylinder pressures also dictate the amount of spark voltage demanded.


Why is a compression gauge inefficient at diagnosing cylinder sealing troubles?  Because the real test of engine seal and performance is the ability of rings, valves, gaskets and castings to work under load


Piston compression rings have inside ledges that face upward.  Compressed gases fill this void and force the piston ring outward against the cylinder wall for better sealing: The higher the cylinder pressure, the more outward ring pressure.  As the piston rises in the cylinder, the ring pressure increases, and that contributes to cylinder taper.  As I shared, taper creates a loss of compression due to the widening ring gap and loss of ring tension.  Combustion pressure rise creates even more outward force at the compression rings, especially at the cylinder's taper zone.


Note: Ring gap is the measureable space between the ring ends, not space between the ring face and the cylinder wall.  Unless zero gap type, new rings call for a specific clearance gap in thousandths of an inch or mm, measured with a feeler gauge.


As one example of the difference between cranking compression and cylinder leak down testing, I recall the stock, FJ40 "F" engine in the 1971 Land Cruiser that I built into an OFF-ROAD Magazine project in the late 1980s.  The engine had considerable mileage, though it consumed minimal oil and ran "okay".  A cranking compression test showed adequate and uniform compression over the six inline cylinders.  Uniform compression, in particular, is always desirable.


When I ran a leak down test on this engine with my Snap-On tester, the results were astounding: While the engine had uniform cranking compression around 120-125 PSI, the leakage per cylinder was 40% or higher!  How is this possible?  In this case, it had much to do with the incredible integrity of Toyota iron engine block castings, with higher nickel content and the hardness to resist taper.


Tearing down the engine for a ring-and-valve job, I found the F-six had worn its top/compression piston rings to the point of razor edges!  I had nicks and cuts on my fingers from unsuspectingly handling the pistons and rings during their removal. 


In this example, there was minimal taper, actually almost none, as I could still see factory honing or "crosshatch" running to the top of the ring travel in each cylinder!  The ring gaps and diameters had worn to the point that air could leak past the top rings during a leak down test.  The lower oil control rings still kept oil from wicking up the cylinder walls and burning. 


The result: adequate cranking compression, no oil loss, yet high leakage and poor sealing.  The pistons and rings could bluff through a compression test by forcing enough air into the compression gauge during cranking—especially with a long stroke inline six.  This could not fool a leak down tester!


This engine ran sluggishly, the result of inadequate cylinder seal at the rings.  The valves seated, the head casting and head gasket were fine, however, the combustion gases leaked past the rings and into the crankcase.  Engines like this have a distinct "blow-by" smell in the crankcase oil.  While the PCV system can recycle blow-by gases, the remnant of combustion gases remains in the oil.  I have often referred to this odor, notable with the oil filler cap removed and the engine running, as the "smell of death".


Summing up, the compression gauge does have its place for quick diagnosis of major engine wear or damage.  A valve that will not seat, or is actually "burned" with a piece etched out, will not evade detection in a compression test—nor will a blown head gasket or severely warped/cracked head casting.  For that reason, an easy and fast check for severe damage can be done with a compression gauge.


The leak down tester is a pinpoint diagnostic tool.  This test can detect specific valve leaks and wear like cylinder taper or the widening of ring gaps.  It can simulate the engine under running load by continually applying compressed air into the cylinder—at the rate and pressure you select.  As an added benefit, you can detect a head gasket or casting crack leak, including leaks between cylinders and leaks into cooling ports, which will force cylinder gases into the cooling system, detectable as bubbles at the radiator.


Caution: If you suspect a head gasket or cracked casting leak into the cooling system, the leak down test will raise cooling system/radiator pressure to the extreme.  Remove the radiator cap before applying air to the cylinders, and start your air flow at a very low PSI, watching for bubbles in the radiator. Cooling systems normally operate in the 12-18 PSI range, depending upon make and model application!  If you suspect leakage into the engine's cooling system, set your leak down tester's air supply pressure accordingly.


Warning: Always make sure the engine and coolant have cooled down before running a compression test or cylinder leak down test!


There are affordable leak down testers.  If you do any amount of engine diagnostic work and want to get a better sense for what's  inside an engine, to make an informed decision about its condition, I highly recommend a leak down tester.  I still use a quality compression gauge to quickly test an engine for major trouble, often following up with a leak down test on the low pressure cylinder(s).


Here is just one example of an affordable leak down tester, I found this in two minutes with a Google Search:




Your test approach and tools of choice depend upon your needs and objectives.  I've owned my Snap-On MT series leak down tester since 1981 and use it when necessary.  The current generation of testers from OTC and others are more compact and easy to hook up, and they perform the same functions as my Snap-On tester.  (Even Snap-On has simplified its testers for today's market.)  There are also compression testers and leak down testers for diesel engine applications.


post-1-0-69095600-1370907155_thumb.jpg post-1-0-64374200-1370907161_thumb.jpg post-1-0-27993400-1370907167_thumb.jpg

I made this leak down tester adapter in minutes. My Honda XR350R and XR500R four-valve thumpers have deeply recessed spark plugs, tough to reach with a spark plug wrench!  I took an old spark plug, knocked out the porcelain electrode, removed and smoothed out the electrode strap end, then cleaned out and brazed the threaded plug shell to a piece of threaded black pipe. The brass "T" fitting provides for an additional gauge, and the air coupler nipple fits my Snap-On MT324 leak down tester.  You can make an "air hold tool" in a similar fashion.  (Click on photos to enlarge.)


If you want a very inexpensive leak down simulator without measuring actual percentage of leak, I've made testers for obscure motorcycle applications with peanut spark plugs and deep plug access.  (See the photo above.)  In this case, you still run the piston to precise TDC on the compression stroke, with the valves completely closed. 


You can listen and hunt for air leaks at the common points.  For the crankcase, listen at the oil filler opening.  For an intake valve leak, listen for air at the throttle body or carburetor air horn.  Listen at the exhaust or tailpipe for an exhaust valve leak.  For a head gasket leak, try soapy water applied at the head gasket edge, compress air into the cylinder at TDC on the compression stroke, watch for air bubbles.  Or for head gasket seepage or a cracked casting into a cooling port, watch the radiator for air bubbles—always using low supply line pressure as I've discussed


The commercially available "air hold tool" is essentially to keep OHV valves closed during valve spring replacement without cylinder head removal.  This tool can serve as a leak tester—without gauge readouts for the percentage of leakage.


K-D and others have made air hold tools for years if you want a ready made tool.  The principle remains for leak testing:  Run the piston to exact TDC on its compression stroke and apply compressed air into the cylinder.  You can set air pressure at your air system's regulator.  Listen for escaping air.  While you won't know the volume or percentage that leaks off, you can detect major air leaks quickly.


I've made air hold tools as simply as knocking the porcelain and electrode from an old spark plug.  Removing the electrode ground strap, I braze an air coupling nipple into the remaining spark plug shell—the threaded metal section.  Clean up the piece thoroughly, removing all debris and loose metal, before inserting your new "tool" into the engine's spark plug threads...


Trust this helps you make the right tool choice!



Link to comment
Share on other sites

  • 1 month later...

Moses, one thing i would like to point out, without stepping on any toes, about a compression test, for the people who have never really done one, is that most compression testers, and also most repair manuals say to remove all of the plugs on the side you are testing, such as the right or left head, for v engines, such as v-6, v-8, v-10, and v-12, and to remove all the plugs in a straight 4 or straight six engine, before doing the test, and to make sure the throttle is fully opened. Also, on most engines, depending on the internal wear in the engine, that the readings will be different when the engine is cold versus warm. As for a comparison between the two tools, a leak down test is the much better way to test for internal issues. One thing i always forget to mention in tool posts, is that tools like these can be rented, at least most of the time, from places like Autozone, and Advance, for when you are only looking to check out one or maybe 2 vehicles, and don't want to spend the money to buy the tools.

Link to comment
Share on other sites

  • Administrators

All good points, biggman100, thanks!  The throttle open is to provide sufficient air volume, I've always performed this test with the engine cold to "tepid" to avoid burning my fingers while removing the spark plugs!  Hot air expands, so too warm is not practical for accurate readings, although your point is well taken about seal and various temperatures.


I prefer removal of all plugs to take a load off the starter.  When immobilizing the ignition, do not long-arc the ignition spark wire discharge on late ignitions, you can damage the coil and other parts. 


Don't crank endlessly, either.  More is not better, in fact it wipes oil from the cylinder walls and drops the compression seal attributed to the oil film under normal engine operation.  Intuitively, what works for me is a burst of 4-5 seconds, no more, to stabilize the volume output...You want an "honest reading", not wishful thinking.



Link to comment
Share on other sites

I have always done compression tests cold, not so much to avoid burning myself on anything, but more because as things heat up in the engine they tend to expand, and to me it seems like that would give a false or inaccurate reading, especially since i have seen vehicles that were a nightmare to get running cold, but once they are warm start right up, and barely give any indication that there is anything wrong. I know that there are 2 schools of thought on it, at least as far as i have seen in either the shops i have worked at, or guys i have spoke to about it, but i think cold is the better way to go.

Link to comment
Share on other sites

  • Administrators

Cold works for testing compression...That's time honored, as a compression test was historically performed as part of a routine tune-up, while the spark plugs were out and other work was underway like changing the points, condenser and rotor...One way to further test with a compression gauge is to take a low compression cylinder and squirt a small amount of clean motor oil into the upper cylinder—just enough to establish a seal around the piston crown to cylinder wall.


Crank the engine now, and if compression jumps right up with this oil seal around the piston-to-wall gap, the valves are likely okay, and the rings have wear. 


Of course, this diagnosis can be confirmed with a cylinder leak down test.  Ring leakage will be audible from the dipstick tube and oil fill hole at the valve cover.  You'll hear the steady hiss of tester air past the rings, into the crankcase, then up through the dipstick tube or valve cover.


With the leak down test, a bad exhaust valve is air hissing out the tailpipe.  A bad intake valve is air hissing out the air intake.  In any case, make sure the piston is at exact TDC on its firing cycle before condemning valves or rings!



Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Create New...