In Front of the Flywheel - October/November - 2017

Intermittent Compression

Can a vehicle have a misfire caused by intermittent compression? If so, can a compression gauge catch it? The answers are “Yes,” and “Only if you’re lucky,” respectively. In this issue, we’ll tackle a vehicle with this very condition.

We’re going to need some history for this one. The vehicle is a 2010 Ford F150 equipped with a 4.6 liter, 3-valve engine. The customer’s complaint is the vehicle runs rough most of the time and the check engine light is on. Diagnostic trouble code P0303 — misfire cylinder number 3 — is in memory.

Where do you go next? This question could have multiple answers or paths. The hope is that the correct path would lead to the fault the fastest. After all, time is money, right? Time to diagnose this vehicle:

After retrieving code P0303, the technician checked the three things necessary for combustion: ignition, fuel, and compression. He removed the ignition coil and installed a ST125 spark tester. With the engine running, the spark easily jumped a gap equivalent to 40,000 volts, confirming sufficient spark.

Next, he swapped spark plugs between cylinders to see if they were the cause. No change. To be thorough, and because Fords are known for ignition coil failures, he swapped the ignition coils to see if the misfire moved; it didn’t.

Next step was to check the fuel. Fuel pressure shouldn’t be a concern because seven other cylinders were functioning correctly. But cylinder #3’s fuel injector or wiring could be at fault.

He connected a noid light to cylinder #3’s fuel injector connector; not the best test in the world, but it blinked rhythmically while the engine was running. In addition, because it was easy to do on this engine, he tried switching injectors between cylinders: The misfire stayed with cylinder #3.

This suggests that two of the three required components for combustion exist: ignition and fuel. Logically, compression is the next thing to check.

He installed a compression gauge and checked compression. He repeated the test several times; it passed every time. He moved the gauge to other cylinders to compare compression; the other cylinders were equal to the cylinder that was misfiring. So what did he miss? Ignition, fuel, and compression all tested good.

This is when I got involved. In the technician’s defense, he did everything he could correctly and approached the issue logically. Now we have to do the same things he did all over again… approach it logically.

My first step was to retrieve the codes. There were a few network codes that had nothing to do with this condition, so I chose to focus on the P0303 code.

I checked freeze frame data for the P0303. The total fuel trim correction for bank 1 is about –2% and bank 2 is around 5%. Given what we know about total fuel trim (see October/November 2016 issue of GEARS), the problem isn’t a fuel injector-related misfire.

In addition, the technician swapped fuel injectors and the misfire remained at cylinder #3. So we can take fuel off the table, because two tests eliminated that possibility.

The next step would normally be to check the ignition system, but the tests the technician performed were pretty thorough for eliminating that as a problem. So I chose to skip ahead and check compression.

The easiest way to check compression quickly on a Ford is a relative compression test, using an OE IDS scan tool (see the July 2017 issue of GEARS). This is a perfect example of an advantage of an OE scan tool over an aftermarket one.

I performed the relative compression test and the engine passed. Then, just to be sure, I performed the test again; this time it failed (figure 1). I ran it a third time and it passed again.

So the engine passed the relative compression test twice and failed once. That could indicate one of two things: the tool’s broken or the compression problem’s intermittent.

The next step was to prove the issue one way or the other. I performed another relative compression test, using an oscilloscope and a high current probe (see the January/ February 2017 issue of GEARS).

The results are confusing. There seems to be a compression issue with cylinder #3 for about the first two thirds of the capture and then it goes away (figure 2). Which leaves two questions: Why didn’t the problem show up on the technician’s compression gauge, and what could be causing the problem?

To answer this, it’d be helpful to understand how a compression gauge works. Pressure from the compression stroke pushes into the gauge; the Schrader valve holds it there. Subsequent compression strokes will add to the pressure.

So the gauge displays the maximum pressure achieved on any one of the many single compression strokes. It won’t distinguish between marginal, or low strokes. It only displays the maximum. In this case, a conventional compression gauge wouldn’t have found the problem, but the relative compression test, using either a scan tool or an oscilloscope, did a wonderful job.

In most cases, the relative compression test is easier to perform than a physical compression test anyway. Maybe that’s why I haven’t taken my compression gauge out of its case in years.

The second question — “What could be causing the problem?” — is easy and tough. The easy part of the question is why the compression is intermittent. Intermittent compression is almost always a valve-sealing issue. If there were a hole in a piston, we’d have no compression at all. If piston rings were broken or cylinder walls extremely scored, we’d have consistently low compression.

So it’s relatively safe to say that we have some type of valve-sealing issue. The tough part of the question is determining what’s causing it.

What could cause intermittent valve-sealing issues? Valve-to-seat issues, broken valve springs, valves sticking in their guides, or pressed-in valve seats that are loose, just to name a few. Regardless, we’re at the point where we need to remove the cylinder head to confirm the problem.

There is another test to narrow down the problem before removing the cylinder head: You could connect a vacuum transducer and use software to analyze the capture.

Vacuum isn’t as smooth as you might think. A vacuum gauge, much like a compression gauge, is limited. With modern equipment, you can watch each intake stroke on an oscilloscope. Automotive engineers have used this technique for decades.

Over the past decade or so, equipment to accomplish these measurements has become more affordable for the average shop or technician. The vacuum trace (in green; figure 3) coupled with a chart that identifies piston position and direction, allows us to go even further.

In our F150 example, the vacuum capture indicates the intake valve is sticking open. How we came up with this final analysis is an article for another day. But the point of this article still remains:

• Can we have intermittent compression? Yes.

• Can we prove it? Yes.

• Would an analog compression gauge show us the data required? No.

• Is it a valve problem? Yes.

• Does the cylinder head have to come off? Yes… or at least the cam cover does. Work smart and use the tools you own to their full potential.

If you don’t understand the vacuum portion of this article, hang on to your seats: In the next few issues, we’ll cover pressure/vacuum transducers in more depth.

Engine or electrical diagnostic issues you’d like to see addressed? Let Scott know. Send him an email at and you just may have your question covered in GEARS Magazine.