Allpar Forums banner

Intermittent miss comes and goes, cold or hot, IAC 255

262 Views 6 Replies 2 Participants Last post by  mike5plus5
2003 Caravan 3.8 has me puzzled. Sorry for the long explanation below but it's needed since I've done a lot without a solution yet.
Ongoing problem with intermittent miss/misfire, engine cold or at operating temp.
Sometimes runs fine/ sometimes frequent misses. No misfires are ever recorded by PCM real time nor in misfire history
(via Matco Maximus 3.0 scan tool)
Recently got P0171 (lean condition) due to bad fuel pump (I heard the old whining pump's pitch go lower right before setting code). Replaced old pump (44 psi) and inline filter - new pump is 55psi, within spec (58 +/-5).
No DTC's ever since pump replacement 2 weeks ago but miss issue persist.
LT Fuel trim 12-14 at idle, lowers at off-idle RPM sometimes quickly, usually gradually.
Short term fuel trim is responsive and follows 02 sensor signal well.
Primary O2 sensor switching above/below 450mv well.
*According to scan tool data stream, IAC valve remains at 255 steps (replaced IAC valve/motor with same result)?
Idle speed is within spec ~750 RPM.
TPS sensor voltage follows actual position smoothly.
MAP, IAT, Coolant temp are accurate in-spec.
Cleaned injectors today, ran good at first then intermittent miss returned.(may be coincidental?)
New spark plugs today (.050 gaps checked) = no change.
No vacuum leaks can be found via carb cleaner sprayed around possible leak sites (intake manifold, injector o-rings, vacuum lines, etc.
No vacuum leak found via clamping off vacuum supply lines from intake manifold to accessories (brake booster, EGR, EVAP, etc.)
No vacuum leak found using smoke machine other than smoke coming out of intake duct at air cleaner.
Any suggestions very welcome!
See less See more
1 - 7 of 7 Posts
Welcome to Allpar. When you pulled the old fuel pump out, was there any signs of debris or fuel contamination in the old pump?

A misfire counter wont be active under a 1/4 tank of fuel. Does the fuel gauge work & show above 1/4 tank? If the battery was disconnected or went dead, has it been driven (drive cycle) to pass the OBDII readiness monitors?

Continuous or Non-continuous Monitors
These monitors can be of either type. It’s up to the manufacturer to decide.
  • Misfire
  • Fuel System
  • Comprehensive Component
OBD II Readiness Monitors


Continuous Monitors


Misfire Detection Monitor
"The misfire detection monitor runs during normal engine operating and driving conditions, and is used by the PCM to determine whether the engine is malfunctioning to the point where it is releasing excess pollutants into the atmosphere. The OBD II system detects misfires on most vehicles by monitoring variations in the speed of the crankshaft through the crankshaft position sensor.

A single misfire will cause a subtle change in the speed of the crank. The PCM tracks each misfire, adding them up and averaging them over time to determine if the rate of misfire is high enough to cause the vehicle to exceed the federal emissions limit. Whenever crankshaft rotational velocity varies by just 2 percent, one or more misfire codes are stored in the memory of the PCM. If this happens on two consecutive trips, the check engine light will be commanded on to alert the vehicle operator that a misfire problem is occurring. If the misfire causes the crankshaft speed to vary by more than 10 percent, the check engine light flashes two times a second to warn the vehicle operator that a severe, catalytic converter damaging misfire condition is occurring.

When the misfire detection monitor detects a misfire, the check engine light will flash as the misfire is occurring. But the light will not remain on the first time a misfire problem is detected. It will come on only if the misfire continues during a second drive cycle and will set a P0300 series code.

A P0300 code would indicate a random misfire (probably due to a vacuum leak, open EGR valve, etc.). If the last digit is a number other than zero, it corresponds to the cylinder number that is misfiring. A misfire code P0301, for example, would indicate a misfire problem in cylinder #1. Of course, misfires can be caused by a lot of things. It could be a worn or fouled spark plug, a weak coil, a bad plug wire, a dirty or dead fuel injector, or a loss of compression (burned exhaust valve or leaky head gasket). Further diagnosis is always needed to isolate and identify the root cause".


Intake manifold vacuum leaks will tend to raise idle speed. The IAC should compensate. Is it always pegged full open at 255? This could be a scan tool issue and not an actual problem? Try a different tool. A 'real' 255 should result in a very fast idle.

Typically, fuel trims should be within +/- 10%. Positive values are 'adding' fuel. If misfires are present, fuel trim numbers are invalid anyways as unburned O2 is going past the O2 sensor. This will look 'lean' to the PCM and it will try to enrich the mixture.

Are the spark plugs the correct Champion RE14PLP5 & the wires in good shape? Are the spark plug tips indicating richness by carbon fouling?
See less See more
  • Like
Reactions: 1
According to the receipts I received at purchase the old pump was replaced in 2017 by the previous owner with a cheap AIRTEX E7144M. The intake strainer was not contaminated bad at all. However, I did not see a receipt for an inline fuel filter so that's the first thing I replaced. When I blew out the old filter (in the correct direction of travel) there was a black tint to the fuel, so the filter integrity may have been compromised. This is the main reason for cleaning the injectors yesterday with a professional cleaning machine using BG Fuel Injection & Combustion Chamber Cleaner:


I filled the fuel tank immediately after the new fuel pump install and keep it above 1/4 tank since then.
It has been driven many drive cycles since last battery disconnect but may still need to relearn some more (2 drive cycles so far) for PCM fuel table to fully re-map after injector cleaning yesterday, if applicable.
The engine was missing when I went to the store @ 4:30 this morning returning home within 2 total miles. But then I went to the shop around 10am and it wasn't missing hardly at all. I connected the scan tool to see the LT Fuel trim had come down from ~14 to ~9.9 and IAC also was down to 223 rather then 255. I don't have access to another scan tool today but I can change tools Wednesday morning. Will report any differences in data.
I'll link a video of today's scanner data stream HERE:


Still no P0300 code or variation. Still no misfires being detected and no misfires in hisory.
Installed 6 new Champion RE14PLP5 plugs yesterday gaps checked @.050. All old plugs were in excellent condition with no sign of being run in rich condition nor any carbon build up.
See less See more
You are sure that this is a misfire? Can you 'make' it throw a misfire code by pulling one of the front spark plug wires (#2, #4 or #6) & then driving it?

On the video, I see that the 'target IAC' is 0 steps, yet it measured IAC is 223 steps (about max opening). If you 'blip' the throttle. does the target IAC change?

The 1/1 O2 sensor volts is pegged low (0.1v-0.15v). It should be swinging rapidly between about 0.2v & 0.8v.
In post #1, you seemed to think it was switching OK.

If you pull the PCV hose to create a large vacuum leak (make lean), the engine should race to a fast idle. Does the O2 voltage react?
Conversely, if we spray a shot of carb or t/body solvent into the air horn to make it rich (try not to stall it), does the O2 voltage react at all?

It there still contamination (black tint) at the tank bottom?

Has the upstream.downstream O2 sensors been replaced? OEM (NGK/NTK) is best.
See less See more
The 1/1 O2 sensor volts is pegged low (0.1v-0.15v)
I don't follow you ImperialCrown? I see the 1/2 (post cat) O2 sensor staying around "(0.1v-0.15v)", but the 1/1 O2 sensor (primary on ex manifold) looks to be switching above and below 450mv often, although I would expect a bit quicker response. May just be the scanner slow to update display.
I disconnect plug wire #6 this morning and drove to the store and back. Then later I drove to the shop where I hooked up the scan tool. No misfire seen in real time data 'misfire screen' nor any recorded in history.
Is there a setting to disable misfire detection because I couldn't find anything in the Matco mode selections? I don't remember that function on any OBD2 vehicles I've worked on in the past.
I blipped the throttle, ran it at 3000 RPM for 2 minutes, and then torqued the motor in gear with no change in the IAC steps from 233 or IAC target '0'.
When I pull the brake booster vacuum line the short term trim responds quickly and the long term trim comes down pretty well. Vacuum line reconnected doesn't have as dramatic effect on short term but it does bring the long term back up slowly. The O2 sensor reaction to the vacuum change seems a little lazy to me and I'm wandering if the MAP sensor input has most of the influence over short, then long term fuel trim adjustments.
Reaction to carb cleaner sprayed into intake is very similar to pulling the booster vacuum line, only in reverse. Sorry I couldn't get that on video with only 1 hand to work with and working alone. I wish I had a similar vehicle to compare these characteristics to.
I was planning to clean the fuel tank while I had it out but it looked clean enough to run without any contamination so I left it as is.
Neither O2 sensor has been replaced to my knowledge by the previous owner. There was no parts receipt for these included with other purchase papers, but neither was a receipt for injectors which he told me had been replaced.
I'm low on parts funds at the moment, especially after I pay to have a diffent scan tool hooked to it Wednesday, but I probably can come up with enough to buy a part if I'm sure the part is the problem. The motor has been running good today with very little misses. I feel like it's safe to drive without damaging the motor at this point, if nothing changes (fingers crossed). But I still need to find out why these issues exist. I will be driving to Florida 1400mi roundtrip to visit family on occasion and can't afford to have an issue come up half way there and be at the mercy of whatever repair shop is available in the area.
I'll link the video of scan tool screen from disconnecting the vacuum booster line today:
See less See more
You're right, I wasn't paying attention to which one I was looking at.
Thanks for posting the live video.
The front 1/1 O2 seems to be switching awful slow. The rear 1/2 O2 sensor is pegged low.
If you drive the van to heat up the sensors, does their switching frequency increase? Cold sensors may switch more slowly.
Let's wait & see what the other scan tool does differently, if anything.

The OBDII readiness monitors may show 'not ready'? A battery disconnect will wipe them out and they have to be relearned with the drive cycle.
The PCM needs to see a working fuel gauge (above 1/4 tank), an engine that reaches normal operating temperature through the coolant temperature sensor, vehicle distance through a working speed sensor and a working transaxle range switch to see PRNDL.
Does the 'ck eng' light come on briefly at start up?
If all the OBDII I/M requirements are being met, I would think that the PCM would be the problem.

OBD2 Drive Cycle - Enable OBD2 Emissions Readiness Monitors

EDIT:
See less See more
I usually run the monitor test with the Matco Maximus 3.0 scan tool before I start looking at data but I have the sequence that you linked printed and will follow it first thing in the morning for the next 2 days to be sure before hooking up the scan tool again Tuesday when I go back to the shop.
Yes the Check Engine light comes on with key on/engine off then goes out immediately after starting.
Yes I'm anxious to see if the different scan tool see's something that this Matco is not or sees something different data.
Will post back when I have more info.
Thanks for your help IC!
  • Like
Reactions: 1
1 - 7 of 7 Posts
Top