This job turned out to be interesting as it the concern in the vehicle was caused by not one but three different problems. Discovering them one after another was pretty, should I say, educative and took a lot of time and effort.
The concern was described by customer as “Engine is difficult to start, requires long crank to start and has a rough idle when running”. Customer tried to fix this himself by replacing spark plugs and “cleaning throttle body”. Vehicle in question was 2011 Ford F-450 Super Duty service truck with 6.8 V10 gas engine.
Upon starting up an engine I verified it takes up to about 5 seconds before engine fires up. I also noticed cranking sound is uneven and similar to a situation of loss of compression on one or more cylinders. Mechanically damaged engines are not rare to see when you dealing with working trucks operating in hard conditions of our area so right away I wanted to perform relative compression test with IDS scan tool to see if I got another case of such a problem.
When turned off and then restarted right away engine was starting normally. Right away I suspected fuel delivery problem as likely reason for this condition.
On a test drive I noticed some lack of power although was not quite sure about it – it was a heavy truck with not the most strongest engine and I did not have any chances before to drive truck with this combination of gross weight and V10 6.8 engine. Most of the trucks of this size would have diesel engines in our area and I used to the feel of those engines with their high torque in low RPM range. While driving I have been monitoring set of parameters with scan tool and noticed that fuel trim values had a significant shift to positive side, way above normally expected especially in high engine load situations (when accelerating or driving uphill). This was second clue about possible fuel delivery problem.
I also noticed that fuel trim values at idle had too much difference with about -7% total trim on bank 1. Negative fuel trim value, contrary to positive one indicates there is either too much fuel or too little air goes to only one bank of cylinders. Power balance test was showing sporadic reduced contribution from different cylinders. Misfire test was also showing misfire counts in different cylinders, although none of them was misfiring constantly. Nevertheless there was no fault codes stored in PCM memory.
When I attempted to perform relative compression test engine unexpectedly started in test mode when fuel supposedly should not be supplied to cylinders. I had to do four consecutive attempts in order to complete this test. Relative compression test did not show any problems with cylinders compression so engine was considered mechanically OK, at least for now.
Upon visual check I noticed one of the vacuum outlets on intake manifold was capped with something that did not look like a right part.
Turned out it was a plastic cap from something like a caulk or glue bottle. To my disappointment this setup did not cause any vacuum leak. It was not the right way to plug manifold vacuum outlet but it was not the cause of concern. I had to look further.
For the next test I performed fuel pressure test with external fuel pressure gauge and verified fuel pressure is way below specs all the time regardless of engine operation mode or fuel pump voltage. I also noticed fuel pressure have been dropping rather quickly after engine was shut off.
At this stage I had strong indications of fuel pump failure. Upon removal and disassembly of the in-tank fuel pump module I found lots of dirt in fuel pump bucket and sock filter.
Surprisingly fuel tank itself was reasonably clean inside, obviously all the dirt that ever was in it was pulled inside the fuel pump bucket and filter causing significant loss of fuel pump performance.
With new fuel pump fuel pressure got restored to perfectly normal levels. On the test drive I noticed that engine power definitely got better. Also fuel trim values were no longer high as it was before.
Despite having all these improvements original concerns of long crank before start and rough idle were still there. Also fuel trims values at idle were far from normal and still had a difference between banks. Diagnostics had to be continued further.
Prior to replacing fuel pump I had injectors removed in order to check for possible leak that would explain fuel pressure leak-off after engine has been shut off. This test had inconclusive results: minor fuel leak-off was present at every injector tip. Possibility of all ten injectors being faulty was considered as unlikely but for the purpose of further diagnostics I had injectors swapped between engine cylinder banks. I also noticed number of the cylinder with the highest misfire count. Upon reassembly I noticed that the highest misfire count followed the injector from one cylinder to another on the opposite bank (from cylinder 7 to cylinder 2). This lead to the decision to replace injector in cylinder 2. This action reduced difference in fuel trims between banks, now fuel trims values were about the same for both banks but still were too much in negative.
Low fuel trim value at idle could serve as an indication of MAF sensor contamination. I completed another visual check that revealed signs of intake contamination with mud and water on the inside of air filter cover (normally this area is perfectly clean on any vehicle)
Visible traces of contamination were also present on the MAF sensor itself
This condition was missed on initial visual check as I only had paid attention to the condition of the air filter itself. As I found it to be neither suspiciously very clean (as if it was just replaced) or too dirty or damaged I did not suspect there could have been any problem in this area. Unfortunately there was a problem. Apparently at some stage this air filter has got damaged by water ingress and replaced by customer or another shop. This is typical problem on the gas engine SuperDuty if they got driven carelessly over deep puddles of water due to location of air intake duct. Actually, driving any vehicle like this
… is quite a risky exercise that could result in as much trouble as major engine mechanical failure.
MAF sensor got replaced with a new one. With new MAF sensor I got fuel trims at idle went from – 20…-25% to about -15%. It was better but still out of normal range. I also had both original concerns still present. Having them after confidently proving as faulty and replacing three different components was quite discouraging. Having multiple problems on old vehicle is not unusual but in the dealership it is rather rare occasion. Anyway there was something else wrong.
I removed spark plugs from the cylinders with the highest count of misfire and compared them to good cylinders spark plugs. Difference in color of the insulator tips indicated unevenness in fuel mixture between cylinders. I also noted that spark plugs were not of the type specified for this engine by manufacturer.
For the next check I used improved external fuel gauge connection setup that allowed me to split fuel system in half with valve and monitor fuel pressure drop separately on fuel pump side or on the injectors side of the system. I isolated pressure leak-off to the injector side of the system and had to remove them again to repeat leak test. Although this time I had higher pressure in the system thanks to a new good fuel pump and one new injector to use as reference point. Now I was able to see clearly difference between injectors. New injector nozzle at cylinder 2 stayed almost completely dry (only minor fuel “sweating” was present)
… while all the other injectors had noticeable leak of fuel with tiny air bubbles present in some:
Testing injectors under full pressure clearly revealed that not one, but all ten of the injectors were faulty. New injectors were installed. This fixed long crank before start concern and almost completely fixed rough idle. Fuel trims returned back to normal in all RPM and load ranges and got almost the same between banks. Occasional miss was still there on idle but unfortunately I had not got a chance to deal with it as customer decided not to proceed with any further repairs.
As a final note I have to say that any unaddressed issue on a vehicle is not going to go away or “heal” by itself. Instead of this unresolved problems would only accumulate leading to increased repair costs as it takes much longer to diagnose all the problems on a vehicle with multiple issues. In this example overall time between diagnostics and replacing faulty parts got up to almost 20 hours (although some of that time was spent on collecting diagnostic data as reference for future diagnostics or working around lack of suitable equipment situations). At least none of the parts were misdiagnosed and replaced without a need.