I Don't Believe in Leprechauns
We've all had that sinking feeling about a job we think has just gone sour. You would rather believe in leprechauns than admit the thought that is currently doing belly whoppers in your stomach acid right now.
Case in point: A shop had just finished a timing belt change on a 1998 Honda Accord with a 2.3L four-cylinder engine. After starting the vehicle, the engine immediately started misfiring.
The shop owner and the technician assigned to the job checked the timing belt alignment repeatedly. They reported to me that the belt alignment was "spot on." They also verified several times that the valve adjustments were correct.
A cranking compression test was performed on the vehicle. The only thing they noticed was that No. 4 cylinder was 174 psi. This was about 25 psi lower than the rest of the cylinders, but they surmised that 174 psi should certainly be enough to fire off that cylinder at idle. And generally speaking, this is true. I have found that most engines will have a compression-related misfire at idle when compression gets below about 90 psi.
Eventually a code P0304 (cylinder No. 4 misfire) was set and the malfunction indicator lamp (MIL) was illuminated. The shop then tried swapping ignition wires and spark plugs, with no improvement. They also tried swapping the injectors on the rail. The misfire did not follow the injector. Since they were indeed getting spark to No. 4, they thought perhaps the spark was happening at the wrong time, so they installed another distributor.
After all these maneuvers failed to correct the misfire, I was called onto the job.
A confirmed miss
When I arrived on the scene, the engine surely had a dead miss in the No. 4 hole, but it did improve some when the rpm was raised up to about 2,500. You can see in the scan data capture that the very negative short-term fuel trim (SHRTFT) and abnormally high load calculation did seem to improve at higher rpm. But with the misfire causing an increased load calculation at idle and the oxygen (O2) sensor being overloaded with raw fuel and ambient oxygen, this system was working way too hard at fuel control.
We knew we had the three basics of spark, fuel delivery and sufficient cranking compression. Next, we needed to check the other aspects of these three basics, including their timing and levels. I began with a simple mechanical integrity test using a vacuum gauge. This dynamic test can indicate many defects in the engine's ability to pump air in and out of itself.
Sure enough, the vacuum gauge picked up a problem. The engine vacuum gauge needle fluctuated about 1.5 inches mercury (Hg), from 20.5 inches Hg to 19 inches Hg. The vacuum needle fluctuations appeared to be about the same frequency as the misfire.
The fact that at times we had 20.5 inches Hg of intake vacuum told me that the overall valve timing was probably correct. We also had enough compression for all the cylinders to contribute, but still No. 4 refused to do so.
I began to get that same sinking feeling the shop owner had showing on his face: This looked like a valve sealing problem.
Time to verify
To double-check my suspicions, I hooked up a lab scope to the No. 4 cylinder ignition wire to compare it to the other cylinders. The capture of the No. 4 secondary pattern at idle shows lots of turbulence in the spark line. At idle, this line should be relatively smooth and even, just like the other cylinders. This much turbulence indicates that the air/fuel (A/F) charge in the cylinder is causing a varying resistance across the spark plug gap because some of the charge was escaping that cylinder during the compression stroke.
At this point, I informed the shop owner that the cylinder head would have to be removed and disassembled for an inspection of the valve components. Understandably, this was not news the shop owner wanted to hear. After all, he did have sufficient compression in that cylinder. I explained to the shop owner that compression and spark and a proper A/F mixture are not the entire battle. Having all those components present at the correct time and in the correct amounts are also required.
In this case, we had a valve problem causing a misfire, not because of low compression, but because of an insufficient A/F charge density problem. The leaking valve was allowing some of the A/F mixture to escape the cylinder on the compression stroke before it could be used during the combustion process. When enough of the A/F charge escapes before the power stroke, that cylinder will not contribute enough power.
Because the valve leak is a somewhat fixed size, it affects the combustion process more at lower rpm. As a result, the leaking valve has more time to do its damage than at higher rpm. This is why the misfire seemed to improve at higher throttle openings when there was less time to leak off its A/F charge. The increased velocity of the valve at a higher rpm also probably helped the valve seal better in those ranges compared to the valve speed at idle.
Time for class
When I am called into a shop to perform diagnostics, I am also sometimes asked to do a bit of informal on-site training as part of diagnosing the particular vehicle that day. This shop owner is one who is eager to learn new things, so I took the time to show him another dynamic test of the engine's ability — or in this case, lack of ability — to properly handle its required pumping activities as part of an efficient combustion process.
I performed a running compression test. I suspect this test has been around as long as there have been four-cycle engines. But to this day, it remains an underutilized or downright unheard-of test.
Basically, it consists of removing one spark plug at a time and installing a compression gauge in each particular cylinder in turn. Then while the engine is actually running, take your compression readings. This is a test of the engine's volumetric efficiency (VE), which is an engine's ability to pump air through under actual running conditions.
At cranking speed, air has all the time in the world to get past the valves and manifolds; however, at actual engine running speeds, there is less time for air transfer in and out of the engine. This means the mechanical pumping systems of the engine must be fully operational and without defect if the required amount of air is to be ingested, held for the proper amount of time and then expelled. If the engine cannot perform this function, it will result in a VE deficiency, or what is known as a "density misfire."
The running compression test is done with the spark to the tested cylinder disabled. I also like to disable the fuel injector to that cylinder so I do not damage the catalytic converter or wash out a cylinder.
Running compression test readings are taken at idle and at snap throttle. You then compare readings to your cranking compression test readings. If all the cylinders' test results fail the test, then an exhaust or intake restriction is indicated. If one cylinder's test results fail, then a valvetrain problem for only that cylinder is indicated.
As you can see in the test chart, cylinder No. 4's cranking compression reading was 19 percent lower than the best cylinder, and the idle compression was 20 percent lower. However, the snap throttle portion of the running compression test was only 5 percent lower.
After hashing out the theory and analyzing the test results, the shop owner agreed to pull the cylinder head and send it out for inspection. The machine shop was kind enough to send back the rebuilt head along with the four bent intake valves in a bag for our inspection.
I am a big believer in verifying my repairs with secondary scope tests because it is so effective, as well as quick and easy. The spark line looked just like the others at idle, and as you can see in the last scan tool screen capture, the fuel trims are also much improved. And of course, the vacuum gauge reads a steady 20.5 inches Hg.
Why did the valves bend to begin with? I am afraid that will remain only a speculation in this case. At this point, I hope the shop owner looks at this case as a very expensive learning experience not soon to be forgotten — and not as an occasion to start believing in leprechauns.
JIM GARRIDO of "Have Scanner Will Travel" is an on-site mobile diagnostics expert for hire. Jim services independent repair shops in central North Carolina. He also teaches diagnostic classes regionally for CARQUEST Technical Institute.