Diagnosing a misfire on an Infiniti I30

Oct. 16, 2014
A rough idle leads to diagnosing a misfire and replacing Bank 1 injectors on a 1999 Infiniti I30.

Vehicle: 1999 Infiniti I30

Customer concern: Rough idle

Tools Used:

  • Snap-on Verus Scan Tool
  • Snap-on Vantage Pro Lab Scope
  • ATS EMisfire Detector software and pressure transducer

Steps:

  1. Verify complaint & conditions
  2. Visual inspection
  3. Check for codes (even with no MIL)
  4. Check Freeze Frame
  5. Avoid diagnostic disassembly: Design a test to non-invasively trim the list of possibilities
  6. Use a tailpipe transducer to identify the misfiring cylinder
  7. Replace Bank 1 injectors

A customer brought their 1999 Infiniti I30 in with a complaint of a rough idle. On the initial test drive, we noted that the idle was consistently rough, though worse in gear. Also, the vehicle ran fine under all off-idle conditions.

Visual inspection

There’s not much to visually check, but it’s a good discipline to avoid missing the obvious. In this case we looked for missing or broken hoses, modifications and other obvious faults. Motor mounts were checked as well, since a fault could transmit vibrations that feel like misfires. All was well, so it was time to choose a diagnostic path.

The wrong path

Given the age and pattern failures on these 3.0L Nissan engines, it was tempting to start directly testing ignition coils and fuel injectors. Both have very high failure rates. But what’s also common on these engines as they age is the damage caused by touching anything. Wires and connectors are brittle, coil boots are often swollen, breather and vacuum lines are hard, etc.

The strategy for every drivability complaint should be to test as much as possible without disassembly, but it’s especially important on applications like this where it’s easy to break things during diagnosis. We’ve all had customers say ‘no’ to a repair estimate, and if we’ve disassembled (and broken) anything during diagnosis, we’re often buying parts just to get it back to the condition it was in when it arrived.

The key points here are to:

  • Diagnose the vehicle quickly enough that the diagnosis itself is profitable (does not exceed the diagnostic time sold).
  • Diagnose it non-invasively so that we’re not performing reassembly for free if the customer declines the repair.

Check for codes

We always check for codes, regardless of MIL status. In this case it was a great idea because there was a P0300. The MIL wasn’t on because the test must have passed at least the last three trips, which turns off the MIL. A P0300 means that the PCM knows there’s a misfire but can’t quite figure out what cylinder it is (and in fact it may be all of them). To get a direction, we mined the PIDs in the freeze frame for clues.

Freeze frame data

The freeze frame PIDs show that the code set at idle, which makes sense based on our verification test drive. But we also noticed that the combined bank 1 fuel trim was -35 percent. That’s the pivotal clue! But was fuel trim a cause or an effect?

Normally, a misfire causes extra oxygen to flow past the O2 sensor, resulting in a lean signal which the PCM reacts to by adding fuel. Therefore, negative fuel trim makes it much more likely that fuel trim is causing the misfire than reacting to it. (See Fig. 1)

Possible causes

What could cause the PCM to subtract fuel from only one bank during a misfire?

  • A rich biased O2 sensor may cause inappropriately negative fuel trim on one bank, resulting in a lean misfire.
  • There could actually be too much fuel going to Bank 1, but -35 percent fuel trim is still considered "in fuel control," so the engine should be running well. However, uneven fueling on Bank 1 could explain it.

That’s about it. You may think that another sensor fault (i.e. TPS) could convince the PCM it’s under load so there would be too much fuel, resulting in negative fuel trim. But that would affect both banks, so it can’t explain this fault.

What about other single-bank faults like ignition or engine mechanical? Both ignition and mechanical faults can cause misfiring on only one bank, but they both result in fuel trim shifting positive in reaction to the unburned oxygen from the misfire, so we ruled those out as well. Maybe you’ve seen negative fuel trim when a converter is clogged, which ties right into the misfire complaint. However, a clogged exhaust would get worse with engine speed, and this misfire was worse at idle. So we were still left with the two possibilities: O2 sensor bias or uneven Bank 1 fueling?

Non-invasively testing for O2 sensor performance

The best order for testing possible causes is always to test the easiest one first. Most vehicles have fast enough data refresh rates that the scan tool can be used to verify O2 sensor performance. And with some clever test combinations, sensor bias can also be ruled out.

In the example below the front sensors are on the left, and the rear sensors are on the right. Bank 2 is the top row, and bank 1 is the bottom row. Graphing all available sensors helps you diagnose by comparison. For example, without any further testing it’s obvious that bank 1 was misfiring consistently. Each misfire pushes 21 percent oxygen by the front O2 sensor, causing a sudden downward spike. (See Fig. 2)

So all O2 sensors worked, and the only thing we could see wrong with the Bank 1 front sensor is too much activity, which was explained by the misfiring. Also, the Bank 1 rear sensor was pretty flat under 100 mV, so we knew that the bank was running lean overall. Could an O2 sensor bias cause this? Yes, the -35 percent fuel trim could be required to pull the signal into range, causing the lean misfire. So a more invasive test was needed to check for bias.

Invasively testing for O2 sensor bias

Just before frame 1000 in the example, propane was added to the intake through a breather hose. You can see that all four sensors reacted by shifting rich. This proved that the front sensors had a similar upper range (960 mV for Bank 2 and 990 mV for Bank 1). But even with the propane, the front Bank 1 sensor couldn’t hold a rich signal because of the misfire and the extremely negative fuel trim. This certainly wasn’t a rich bias, since it wouldn’t stick rich even when forced.

Injector lab scope test

That’s one possible cause eliminated – time to check injectors. But the rear bank injectors on this engine are inaccessible. The intake could be removed, but then you’d be back to the possibility of reassembling and maybe replacing hoses and gaskets for free. Instead, the 1, 3 and 5 control circuits were back probed at the subharness connector near the intake manifold.

The Snap-on Vantage Pro captures (Fig. 3) show the injector control voltage for injectors 1 and 3. Both had good open circuit voltage, both were pulled to near ground, and both had similar inductive kicks. These injector control circuits don’t look "normal" compared to most vehicles, but they are normal for this engine, and all three looked identical (injector 5 is not shown here, but results were identical).

Note: Nissan fans will say we should have used the scan tool injector-kill function to perform a balance test. In fact we did, but with a badly running engine and fuel trim fighting to keep control, the results were not very conclusive.

Injector current flow can also be helpful for finding electrical faults, and even some mechanical faults. So a PDI CA60 inductive low current probe was clamped around the control wires one at a time to obtain the current waveforms below. You could also clamp it around the common power wire to get all three injector current ramps with a single connection, but then you wouldn’t be able to identify which ramp belongs to what injector without using a second channel to track the sequence (i.e. injector voltage).

The Snap-on Vantage Pro captures (Fig. 4) show the current from the same two injectors, and again they looked exactly the same. Both flowed just over 1 amp in just over 2 ms. That’s a pretty short injector duration, but remember that the fuel trim was -35 percent, shortening the injector duration by over one-third. And both had a slight dip at about the 700 mA point in the ramp. This dip appears where the pintle actually opens, so matching results prove that the injectors mechanically opened at the same time. Again, all 3 injectors were tested, but only two results are shown here for comparison.

Stop and check the integrity of the diagnostic path

So all three Bank 1 injectors worked the same electrically, flowed the same current and mechanically worked the same. A failing result would have found the root cause of our problem, but a passing result didn’t prove that there wasn’t an injector fault. After all, we knew through logical elimination that it was a lean misfire caused by the PCM subtracting fuel, and that this condition could only be caused by uneven fueling or a biased O2 sensor. Since the O2 sensor tested OK, we still had to figure out which injector was pouring in fuel.

Using logic and documenting results really makes a difference in cases like this because we didn’t get distracted by passing test results. By knowing what could cause negative fuel trim and a lean misfire, we weren’t fooled when our two prime suspects had good alibis. It was time for a new test.

Locating the misfiring cylinder

Because the misfire code did not identify the misfiring cylinder(s) and this vintage Infiniti did not support misfire PIDs, a tailpipe transducer was used to help narrow the diagnosis. For this test, we used an ATS EMisfire software package with the ATS tailpipe transducer. The idea is that a misfiring cylinder will reduce the velocity of the exhaust (because the gases from that cylinder did not burn and expand). The software knows the firing order, so as long as you set up the trigger properly (usually ignition primary or secondary), the program can match exhaust pulses to each cylinder in the firing order.

You can see that cylinders 3 and 5 were misfiring badly (Fig. 5). Since we were convinced it was an injector fault, we should have replaced injectors 3 and 5, right? No! This is the most important point in this entire case: We knew that whatever cylinders were misfiring were doing so only because the PCM was taking away too much fuel. And that the fuel was being removed in reaction to at least one injector pouring in too much fuel. So the cylinder that wasn’t misfiring was the bad one! 

The repair

Based on the test results, it was clear that injector 1 was leaking fuel into cylinder 1, causing it to run rich, and resulting in negative fuel trim and lean misfiring in cylinders 3 and 5. Without pulling the injectors we couldn’t be 100 percent sure, but we got to a point where we could confidently sell the job, and we hadn’t disassembled a single part. With a 15-year-old vehicle and $800 worth of injectors and labor for the rear bank alone (plus whatever breaks), odds are decent the customer won’t fix it. Having it fully assembled and running at the end of a diagnosis like this is key.

This customer did approve disassembly to figure out where that extra cylinder 1 fuel was coming from, and the graphics that follow show what we found. These are bottom-feed injectors that sit inside of the rail and are bathed in fuel. The plastic spacers at the bottom retain an O-ring that prevents the fuel surrounding each injector from leaking into the intake runner. The spacers on all three injectors were cracked, and the one on injector 1 was broken and missing (See Fig. 6 and Fig. 7). So our diagnosis was correct – extra and uneven fueling on bank 1. But our injector tests were also accurate – there wasn’t anything wrong with them. The fuel was actually getting into cylinder 1 around the injector. All three injectors were replaced based on the cracking, Bank 1 spark plugs were replaced, codes were cleared and a test drive proved that the rough idle misfire was fixed.

The customer was informed that the Bank 2 injectors were probably in about the same condition. Due to the expense, and the ease of access for replacement at a later date, the customer declined the Bank 2 repair. However, the recommendation was a good move both for the customer’s sake and as a good defense for the shop if similar symptoms show up in the future.

A path not taken

If we had risked pulling the COP coils earlier in the diagnosis for plug access, the graphic below shows what we would have found (See Fig. 8). Could you tell what was going on from these plugs? Maybe, but it’s easy to misinterpret them. The black plug was actually from cylinder 1, and it was the only one firing. The misfiring plugs look great, since a lean misfire won’t build up the same deposits as a plug in a rich cylinder. 

Sponsored Recommendations

ZEUS+: The Cutting-Edge Diagnostic Solution for Smart, Fast, and Efficient Auto Repairs

The new ZEUS+ simplifies your diagnostic process and guides you through the right repair, avoiding unnecessary steps along the way. It gives you the software coverage, processing...

Diagnostic Pre- and Post-scan Reports are Solid Gold for Profitability

The following article highlights the significance of pre-scans and post-scans, particularly with Snap-on scan tools, showcasing their efficiency in diagnosing issues and preventing...

Unlock Precision and Certainty: TRITON-D10 Webinar Training for Advanced Vehicle Diagnostics

The TRITON-D10 lets you dig deep into the systems of a vehicle and evaluate performance with comparative data, systematically eliminating the unnecessary to provide you with only...

APOLLO-D9: Trustworthy Diagnostics for Precision Repairs

The APOLLO-D9 provides the diagnostic information and resources you need to get the job done. No more hunting through forums or endlessly searching to find the right answers. ...

Voice Your Opinion!

To join the conversation, and become an exclusive member of Vehicle Service Pros, create an account today!